The incidence of port-site metastases might be reduced

M A. Reymond,1 Ch. Wittekind, 2 A. Jung,2 W. Hohenberger,1 Th. Kirchner, 2 F. Kockerling1

1 Department of Surgery, University of Erlangen, P.O. Box 3560, 91023 Erlangen, Germany2 Department of Pathology, University of Erlangen, P.O. Box 3560, 91023 Erlangen, Germany

Received: 26 October 1996/Accepted: 14 February 1997

AbstractBackground:Laparoscopic resection for cancer is contro-versial and port-site metastases are not infrequent. Themechanisms of occurrence of port-site metastases remainunclear. Animal experiments have suggested a role for car-bon dioxide (CO2), but port-site metastases also occur afterthoracoscopy, where no CO2 is used. The aim of this studywas to define the role of CO2 in the seeding of tumor cellsin the human patient.Methods: CO2, instruments, trocars, suction device, andperitoneal washing were examined during 12 staging lapa-roscopies for pancreatic cancer. The presence, viability, andbiological significance of cells were investigated using con-ventional cytology, polymerase chain reaction (PCR), andrestriction fragment length polymorphisms (RFLPs) to de-tect the presence of a mutant k-ras gene as a genetic markerof cancer cells.Results:Cytology exam of peritoneal washing, instruments,the suction device, and trocars revealed many cells. Tumorcells were detected in 6/12 peritoneal, in 4/12 trocars and4/11 instruments washings, but not in 12 CO2 samples. TheDNA content of CO2 was very low—as assessed by PCR.Mutant DNA was detected by RFLP in four out of 12 aero-sols. Six aerosols did not contain any DNA. Two aerosolswere borderline.Conclusions:During staging laparoscopy for pancreaticcancer in humans, CO2 contains only very low levels offree-floating tumor cells, even in the presence of massiveperitoneal contamination. These results suggest that the in-cidence of port-site metastases might be reduced if me-chanical contamination of the port sites with instruments orwith the specimen can be avoided.

Key words: Neoplasms, staging — Laparoscopy, adverseeffects — Pancreatic neoplasms, surgery

Recent technological advances now allow surgeons to treat

abdominal malignancies via the laparoscope. Although suchtreatment is accepted for palliative operations, laparoscopyfor curative resection remains controversial. Items of con-cern are questionable radicality, a lack of long-term results,and port-site metastases. The last-mentioned complication,occurring after a curative laparoscope-assisted colonic re-section, was first reported in 1993 [1] by Alexander, andnumerous case reports have since followed.

The first experimental animal study on port-site metas-tases was published in 1995 [12]. Clinical conclusions aredifficult to draw from the results obtained so far.

The present study was undertaken to determine howmany cells are present in the CO2 during a clinical staginglaparoscopy for pancreatic cancer, whether these cells areviable, and whether they include tumor cells. The staginglaparoscopy for pancreatic cancer is an excellent model forinvestigating these questions, because over 90% of suchcancers show mutation on codon 12 or 13 of the k-ras gene[4], thus permitting detection of a small number of cancercells.

Materials and methods

Tumors and cell lines

Samples were obtained from 12 patients who underwent a staging lapa-roscopy for pancreatic cancer at the universities of Erlangen, Wu¨rzburg,Regensburg, and Ulm between 15.1 and 8.5.96. Eight tumors could beresected. Biopsies were performed in cases 8, 10, 11, and 12. The diagnosisof pancreatic adenocarcinoma was based on histopathological analysis inall cases. Since neither the indication nor the procedure was modified bythe analysis, no informed consent was required. The cell line of humanpancreatic carcinoma (PC44) was a kind gift from H. Heinmo¨ller from theUniversity of Regensburg; it showed a mutation on codon 12. K-ras nega-tive control cell lines (T-ALL Jurkat) were obtained from our laboratory.

Sampling

Twenty milliliters of peritoneal washing was sampled at the beginning ofthe procedure. The CO2 was delivered during all the duration of the pro-cedure through an artificial leak into a closed Bu¨lau bottle, which was filledwith 400 ml phosphate-buffered saline (PBS). The CO2 was allowed toCorrespondence to:F. Kockerling

Surg Endosc (1997) 11: 902–906

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

diffuse through the medium, so all its contents were captured in the bottle;89.3 ± 3.8% of CO2 was collected. The pH was maintained around 7.4using titration with NaOH. By the end of the procedure, trocars and in-struments were rinsed separately with 50 ml PBS per sample. The proce-dure was followed by immediate centrifugation of the samples at 400g for10 min at 4°C. The collecting system was washed and sterilized betweenthe procedures according to surgical standards. After three cases, the analy-sis of a standard laparoscopic cholecystectomy excluded a possible con-tamination.

Cytology

After vital staining with trypan blue, the viability of the cells was assessedby conventional microscopy. An aliquot of 1.5 ml was centrifugated at400g for 5 min at room temperature. After Giemsa staining, the cytologysamples were analyzed by a senior pathologist. Cell counting was per-formed after fixation and adapted to the initial sample volume.

DNA preparation

The cell pellet was resuspended into 1 ml PBS and centrifugated at 400gfor 5 min at room temperature. Then the pellet was resuspended into 380ml Higushi buffer (50 mM KCl, 20 mM Tris-HCl, pH 8.3, 2.5 mM MgCl2,0.1 mg/ml gelatine, 0.45% (v/v) NP-40, 0.45% [v/v] Tween 20), and 20mlproteinase K (Boehringer, Mannheim, Germany), and incubated at 56°Covernight with continuous shaking at 1,200 rpm. After denaturation at95°C for 5 min, the DNA was stored at −20°C until analysis. Because ofthe presence of inhibitors the samples had to be purified using a QIAampDNA purification kit (Qiagen, Hilden, Germany) according to the manu-facturer’s protocol.

K-ras PCR

A nonradioactive method was used, involving PCR amplification of K-rasfirst exon sequences (157 pairs of bases) [19]. Oligonucleotide primerswere as follows:

K-ras 58: ACT GAA TAT AAA CTT GTG GTA GTT GGA CCTK-ras 38: TCA AAG AAT GGT CCT GGA CC

Primers were purchased from MWG Biotech (Ebersberg, Germany); 1mlof high molecular weight DNA was amplified in a volume of 50ml con-taining 1 × PCR reaction buffer, 4ml MgCl2 (25 mM), 1 ml primers (50pmol/ml), 1 ml NTP (10 mM), and 0.2ml 1U TaqDNA polymerase (PerkinElmer). The reaction was overlaid with 50ml mineral oil (Sigma ChemicalCo., St. Louis, MO). DNA was amplified with a Hybaid Omnigene 3Thermal Cycler using following parameters:

3 cycles with 95°C, 0.5 min/42°C, 0.5 min/72°C, 1 min35 cycles with 95°C, 0.5 min/55°C, 0.5 min/72°C, 1 min

The PCR products were run on 2% (w/v) 0.5 × TBE ethidium-bromide-containing (1mg/ml) agarose gels, and the result was photographed on anultraviolet transilluminator. The PCR allowed us to detect less than 10 cellsin PBS, as assessed with the mutant cell line (Fig. 1). The 40-cycles-second-PCR was performed with the PCR products of a first 40-cyclesPCR, after 1:1,000 dilution. The negative controls remained negative.

Restriction enzyme analysis

The PCR was followed by restriction to detect either endogenous restric-tion polymorphisms, which can be found in the PC44 pancreas carcinomacell line. For the codon 12, the wild gene was digested byMvaI (Boe-hringer Mannheim) into a main fragment of 114 pb; the mutant gene wasdigested into a 143-pb main fragment. For codon 13, the mutant gene wascut by HphI (New England Biolabs) at a length of 114 pb; the wild genewas not restricted. The digestions were incubated at 37°C for 3 h, followingthe instructions of the supplier. An electrophoresis using 3% (w/v) 0.5 ×TBE Metaphore gel containing 1mg/ml ethidium bromide was then per-formed, and the result was photographed on an ultraviolet transilluminator.

Statistical analysis

To rule out the null hypothesis that the cellular contamination was equal inall samples, the Kruskal-Wallis test for equality of populations was used.Data were not normally distributed. Tests were two-tailed.

Results

Cellular content of CO2

Figure 2 shows the characteristics of the various items. Theperitoneal washing sample was the most cellular (5.3 × 105

± 4.1 × 105 cells), although the suction device (7.7 × 104 ±1.2 × 105), instruments (4.9 × 104 ± 4.6 × 104), and trocars(2.9 × 104 ± 1.8 × 104) also bore many cells. The CO2contained almost no nucleated cells (3.4 ± 12.4) but didcontain carbonized material and some cell fragments. Thedifference between the cell contents of the different itemswas statistically highly significant (p 4 0.0001).

Cell viability

The viability of the cells found on the instruments and tro-cars was estimated to be over 90% with trypan blue andconventional microscopy. No cells could be seen in CO2samples.

Biological significance of the cells

Table 1 shows the results of conventional cytology for thevarious items: no tumor cell was detected in the retrievedCO2 samples, but many suspect cells were detected on theinstruments (Fig. 3) and trocars (Fig. 4). A more preciseanalysis was enabled by molecular biological methods. Itwas possible to amplify the k-ras gene in six out of 12aerosols (data not shown). An RFLP analysis of four of thesix positive PCR products was possible and showed thepresence of tumor cells (Fig. 5). The remaining two aerosolscontained insufficient DNA for RFLP analysis. A secondPCR excluded the presence of DNA in the other six aero-sols, and thus the presence of tumor cells. Table 2 shows theresults of the gene analysis for the various items.

Fig. 1. Polymerase chain reaction (35 cycles) of the k-ras gene, with serialdilutions in phosphate-buffered saline solution (PBS), showing a sensitiv-ity of less than 10 cells.

903

Discussion

Our results show that during a staging laparoscopy for pan-creatic cancer—a procedure for which port-site metastaseshave been described [21]—the risk of seeding tumor cellsthrough CO2 remains very low. On the other hand, a highcellular contamination of the instruments and the trocars—including sometimes tumor cells—is documented.

These results are consistent with the clinical data avail-able. Some experienced surgeons have performed hundredof colectomies without any port-site metastases [12] beingseen. Beside two other publications [8, 13], Downey re-cently reported on a series of 21 port-site metastases afterthoracoscopy where no CO2 is usually used [11]. Such trou-bling complications have also been reported after medias-tinoscopy [26]. Incision site metastases occurring after fineneedle biopsies are well known [9, 24, 27].

Since 1995, many experimental studies on laparoscopyand cancer have appeared. Compared with control animals(anesthesia alone), tumor implantation at trocar sites is en-hanced by CO2 in the hamster [20] and the rat [7, 10, 16, 18]when cells are injected intraperitoneally, and there is a defi-nite dose-response relationship between tumor implantationand the number of cells in the inoculum [20]. This does not

happen when a solid [6] or retroperitoneal [23] tumor modelis used, which is similar to the clinical situation in earlytumor stages.

Interestingly, laparoscopy involving the use of air en-hances intraperitoneal tumor growth more markedly thanCO2 in laparoscopy [17]. The influence of helium in lapa-roscopy is presently being debated [10, 18]. These resultssuggest both a gas-independent mechanism and an inhibi-tory effect of CO2 at high concentrations, and this deservesfurther investigation.

After laparotomy, intraabdominal recurrences occurpreferentially in the laparotomy wound or at the anastomo-sis [25]. Intraabdominal recurrences after laparoscopy occuronly at serosal lesions (port sites [16, 20] or liver surface[20]), which underscores the importance of an intact peri-toneal barrier for preventing the implantation of free cancercells in the peritoneal cavity [14].

Although CO2 laparoscopy enhances tumor implanta-tion when compared with anesthesia alone, it is associatedwith less experimental tumor growth than laparotomy whencancer cells are injected under the renal capsule [7], or intothe pancreas after manipulation of the tumor [23], or sub-cutaneously [3, 5] or even intraperitoneally [17].

Using a radioactive cell line in pigs, Allardyce has re-cently shown that although CO2 may increase wound-site

Fig. 2. Cell content of the various items, showing significant differencesbetween the subgroups (Kruskal-Wallis test for equality of populations,p< 0.0001).

Table 1. Results of the cytologya

PatientPeritonealwashing Instruments Trocars

Suctiondevice CO2

1 N N N N No cells2 S T S S No cells3 T S S N Normal4 S — S — No cells5 T N N — No cells6 N N N N Normal7 N S N — No cells8 T N S — No cells9 N N N S Normal

10 T N T N No cells11 N S N N No cells12 N N N — No cells

a N, normal cells; S, suspect cells; T, tumor cells; —, no sample.

Fig. 3. Patient 2: Typical cytology of the instruments, showing manysuspect cells (Giemsa staining, magnification 200×).Fig. 4. Patient 8: Cluster of suspect cells found on a trocar (Giemsa stain-ing, magnification 200×).

904

implantation, the major variable influencing tumor cell de-position is whether or not the port is used by the surgeon [2].Using filters, Hewett demonstrated that the movement ofcells throughout the peritoneal cavity during laparoscopy isvia contaminated instruments, with local contamination ofthe port by dispersion within water vapor remaining a pos-sibility [15]. Using in vitro and in vivo models, Whelanshowed that trocar-site recurrence is unlikely to result fromaerosolization of tumor cells [28].

To explain the experimental findings, pressure gradientmust be assumed to play a limited role in tumor dissemi-nation, but only when tumor cells are already present in theperitoneal cavity. This effect seems to be dose dependent

and relatively gas independent. Although tumor growth isenhanced through CO2 laparoscopy when compared withanesthesia alone, there is less tumor growth after CO2 lap-aroscopy than after conventional laparotomy, as the resultsobtained in five independent animal studies [3, 5, 7, 17, 23]have shown. If this finding is confirmed in the clinical set-ting, it might be an argument for using laparoscopy in cu-rative surgery in selected indications.

In summary, only very low levels of free-floating tumorcells can be detected in CO2 during staging laparoscopiesfor pancreatic cancer using CO2 at pressures between 10and 15 mmHg. Mechanical inoculation of the abdominalwall by the surgical specimen or an instrument is signifi-cantly more likely to occur, since they carry tumor cells.This must be the case in thoracoscopy and mediastinoscopy,where no CO2 is used. Proper prevention measures—including a good surgical technique and proper indications[12, 22]—could reduce the incidence of port-site metasta-ses.

Acknowledgments.This study was funded by the universities of Genevaand Erlangen. We gratefully acknowledge the help of K. Fuchs (Wu¨r-zburg), R. Kunz (Ulm), and K. Jauch (Regensburg) for the recruitment ofpatients and the technical help of S. Kastl, J. Riese, and T. Brabletz.

References

1. Alexander RJ, Jaques BC, Mitchell KG (1993) Laparoscopically as-sisted colectomy and wound recurrence [letter; comment]. Lancet 341:249–250

2. Allardyce R, Morreau P, Bagshaw P (1996) Tumor cell distributionfollowing laparoscopic colectomy in a porcine model. Dis Colon Rec-tum 39: S47–S52

3. Allendorf JD, Bessler M, Kayton ML, et al. (1995) Increased tumorestablishment and growth after laparotomy vs laparoscopy in a murinemodel. Arch Surg 130: 649–653

4. Bos JL (1989) Detection of ras oncogenes using PCR. In: Erlich H (ed)DNA amplification using the PCR. Stockton Press, New York, pp225–233

5. Bouvy ND, Marquet RL, Hamming JF, Jeekel J, Bonjer HJ (1996)Laparoscopic surgery in the rat. Beneficial effect on body weight andtumor take. Surg Endosc 10: 490–494

6. Bouvy ND, Marquet RL, Jeekel H, Bonjer HJ (1996) Impact ofgas(less) laparoscopy and laparotomy on peritoneal tumour growth andabdominal wall metastases. Surg Endosc 10: 1618

7. Bouvy ND, Marquet RL, Lambert SWJ, Jeekel J, Bonjer HJ (1996)Laparoscopic bowel resection in the rat: earlier restoration of IGF-1and less tumor growth. Surg Endosc 10: 567 (Abstract)

8. Collard JM, Reymond MA (1996) Video-assisted thoracic surgery(V.A.T.S.) for cancer: risk of parietal seeding and of early local re-currence. Int Surg 81: 343–346

9. Dick R, Heard BE, Hinson KEW (1974) Aspiration needle biopsy ofthoracic lesions: an assessment of 227 biopsies. Br J Dis Chest 68:86–93

10. Dorrance HR, Oein K, O’Dwyer PJ (1996) Laparoscopy promotestumour growth in an animal model. Surg Endosc 10: 559 (Abstract)

11. Downey RJ, McCormack P, LoCicero III J, and the Video-AssistedThoracic Surgery Study Group (1996) Dissemination of malignanciesfollowing video-assisted thorac surgery. J Cardiovasc Thorac Surg111: 954–960

12. Franklin ME, Rosenthal D, Abrego-Medina D, et al. (1996) Prospec-tive comparison of open vs. laparoscopic colon surgery for carcinoma.Dis Colon Rectum 39: S35–S46

13. Fry WA, Sidiqqui A, Pensler JM (1995) Thoracoscopic implantationof cancer with a fatal outcome. Ann Thorac Surg 59: 42–45

14. Goldstein DS, Lu ML, Hattori T, Ratliff TL, Loughlin KR, KavoussiLR (1993) Inhibition of peritoneal tumor-cell implantation: model forlaparoscopic cancer surgery. J Endourol 7: 237–241

15. Hewett PJ, Thomas WM, King G, Eaton M (1996) Intraabdominal cell

Fig. 5. Four aerosols contained mutant DNA; two aerosols were border-line; six aerosols did not contain any DNA. U4 uncut PCR product; 124 restricted withMvAI (codon 12); 134 restricted withHphI (codon 13).

Table 2. Contents of CO2: amplification and restriction fragment lengthpolymorphisms (RFLP)

PatientPCR(k-ras gene)

RFLPa

(codons 12 & 13)

1 Present —2 Present Mutant3 Present Mutant4 Absent —5 Absent —6 Present —7 Present Mutant8 Absent —9 Present Mutant

10 Absent —11 Absent —12 Absent —

a — 4 no reaction possible in the absence of DNA; mutant4 presence oftumor cells.

905

movement during abdominal carbon dioxide insufflation and laparos-copy. Dis Colon Rectum 39: S62–S66

16. Hubens G, Pauwels M, Hubens A, Vermeulen P, Van Marck E,Eyskens E (1996) The influence of pneumoperitoneum on the perito-neal implantation of free intraperitoneal colon cancer cells. Surg En-dosc 10: 181 (Abstract)

17. Jacobi CA, Ordermann J, Bo¨hm B, Zieren HU, Volk HD, Muller JM(1996) Increased tumor growth after laparotomy and laparoscopy withair versus CO2. Surg Endosc 10: 551 (Abstract)

18. Jacobi CA, Sabat R, Bo¨hm B, Zieren HU, Volk HD, Muller JM (1996)Pneumoperitoneum with CO2 stimulates malignant tumor growth.Surg Endosc 10: 551

19. Jiang W, Kahn SM, Guillem JG, Lu SH, Weinstein IB (1989) Rapiddetection of ras oncogenes in human tumors; applications to colon,esophageal, and gastric cancer. Oncogene 4: 923–928

20. Jones DB, Guo LW, Reinhard MK, et al. (1995) Impact of pneumo-peritoneum on trocar site implantation of colon cancer in hamstermodel. Dis Colon Rectum 38: 1182–1188

21. Jorgensen JO, McCall JL, Morris DL (1995) Port site seeding afterlaparoscopic ultrasonographic staging of pancreatic carcinoma [letter].Surgery 117: 118–119

22. Kockerling F, Reymond MA, Schneider C, Hohenberger W (1997)Mistakes and hazards in oncological laparoscopic surgery. Chirurg 68:215–224

23. Mutter D, Hajri A, Tassetti C, Solis-Caxaj C, Aprahamian M, Mares-caux J (1996) Experimental pancreatic tumor growth and spread afterlaparoscopy versus laparotomy in the rat. Surg Endosc 10: 490–494(Abstract)

24. Nankhonya JM, Zakhour HD (1991) Malignant seeding of needle as-piration tract: a rare complication. Br J Dermatol 1243: 285–286

25. Skipper D, Jeffrey M, Cooper AJ, Alexander P, Taylor I (1989) En-hanced growth of tumour cells in healing colonic anastomosis andlaparotomy wounds. Int J Colorectal Dis 4: 172–177

26. Sullivan WD, Passamonte PM (1982) Mediastinoscopy incision sitemetastasis: response to radiation therapy. South Med J 72: 1428

27. Voravud N, Shin DM, Dekmezian RH, Dimery I, Lee JS, Hong TW(1992) Implantation metastasis of carcinoma after percutaneous fine-needle aspiration biopsy. Chest 43: 1533–1540

28. Whelan RL, Sellers GJ, Allendorf JD, et al. (1996) Trocar site recur-rence is unlikely to result from aerosolization of tumor cells. Dis ColonRectum 39: S7–S13

906

The phagocytosis activity during conventional and laparoscopicoperations in the rat

A preliminary study

C. N. Gutt, P. Heinz, W. Kaps, V. Paolucci

Department of General Surgery, Johann-Wolfgang-Goethe-University, Theodor-Stern-Kai 7, D-60590 Frankfurt/Main, Germany

Received: 8 December 1996/Accepted: 20 February 1997

AbstractBackground:Numerous experimental and clinical investi-gations indicate that the mononuclear phagocyte system(MPS) has a relevant function in terms of physiologicaldefense against tumor metastasis and bacterial infection.Consequently, a point of major interest is the influence ofsurgical techniques on the MPS function.Method:The model investigation examines the phagocyto-sis activity of the rat’s MPS during conventional fundopli-cation (group 1,n 4 10), laparoscopic fundoplication usinga pneumoperitoneum (group 2,n 4 10), and gasless lapa-roscopic fundoplication (group 3,n 4 10). The MPS func-tion is evaluated by an intravascular carbon clearance test(G. Biozzi).Results:The fastest carbon elimination half-life was foundin group 3. By way of contrast, there was a significantincrease of carbon half-life in group 2 (p < 0.005). Evengroup 1 caused less MPS depression (p < 0.1) than group 2.Conclusion:Gasless laparoscopic procedures have a favor-able effect on phagocytosis activity. The CO2 pneumoperi-toneum seems to be the main reason for a decreased antigenelimination in laparoscopic treatments.

Key words: Laparoscopy — Pneumoperitoneum —Gasless laparoscopy — Mononuclear phagocytes system —Rat model

The human and animal organism’s answer to an injury ischaracterized by an early rise of the serum level for stresshormones and a drop in the immunoresponse conveyed bythe cells. The stress reaction’s force and duration are con-sidered to be proportional to the severity of the injury. Thereis a correlation between a reduced perioperative cell-conveyed immunoresponse and an increased risk of postop-

erative infection and metastatic tumor spreading [6, 9, 10,11, 13].

The investigation of the mononuclear phagocyte system(MPS)’s phagocytosis activity is determined by means ofthe Carbon-Clearance test which, itself, is carried out in amodified form, i.e., following the method as described byBiozzi et al. [3] and Lemperle et al. [7]. In the course of thistest, the degree to which carbon particles are eliminatedfrom the circulating blood by local macrophages of the MPSis determined. This phagocytosis activity depends on thekind and extent of an operative manipulation and can bejudged to be a measure of interference with the organismcaused by surgical interventions. Our investigation com-pares the phagocytosis activity during conventional opera-tions (group 1) with that of laparoscopic surgery using apneumoperitoneum (group 2) and gasless laparoscopic sur-gery without the use of a pneumoperitoneum (group 3)(Table 1).

Method

Laparoscopy in the rat

A surgical table for small animals (KAPS Company, Asslar, Germany) isused consisting of an operation platform and a base plate with a flexiblesupporting arm. In a supinated position, the narcotized animal lies on theraised operation platform which is set over the base plate. By slightlymoving the operation platform via two micrometer screws with fixed lapa-roscope, the surgeon or his assistant can finely adjust the picture segment.The surgical instruments consist of a 4-mm arthroscope and micro-alligatorforceps and microscissors (Aesculap, Tuttlingen, Germany). All instru-ments, as well as the suture material, can easily be inserted into the abdo-men via 3-mm synthetic ports [2, 5].

Experimental procedure

Before the operation starts, the rats are injected via the v. dorsalis peniswith 0.1 ml/100 kg of ‘‘fount india’’ shellac-free ink (Pelikan Company,Correspondence to:C. N. Gutt

Surg Endosc (1997) 11: 899–901

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Hannover, Germany) which, in order to obtain a better colloid stabilization,is mixed with gelatine. Periodically, 200ml blood is taken from the ani-mals. After the hemolysis in 2 ml of a 0.1% Na2 CO3 solution, the mix-ture’s extinction at a wavelength of 640 nm is determined as a blank valueagainst Na2 CO3/blood (before ink injection). The extinction depends onthe ink particles’ quantity (particle size approx. 250 Å)/volume and con-sequently represents a measure for the ink concentration in the peripheralblood. Throughout that intervention period, the ink’s half-life in the ani-mal’s circulating blood can be determined by exploratory blood takings.The bigger the half-life, the longer the ink remains in the blood and thelower is the MPS’s phagocytosis activity under the corresponding opera-tion conditions. At first, the rats of the first group of animals (group 1,n4 10) are subject to a conventional laparotomy; then a fundoplication afterNissen is made. Throughout the operation, and in defined intervals (3, 15,30, and 40 min after the ink injection), 200ml of blood is taken from theanimal’s retrobulbar plexus venosi and analyzed. In the second group ofanimals (group 2,n 4 10), the fundoplication is carried out under anendoscopic operation with a pneumoperitoneum of 7 mmHg. The ratsbelonging to the third group of animals (group 3,n 4 10), as well, aresubject to an endoscopic fundoplication; however, a pneumoperitoneum isrenounced and, instead, the abdomen is stretched out by a specially con-structed wire bow. Still narcotized, all animals are killed by carbon mon-oxide.

Statistics

The mean carbon half-life and the standard deviation (SD) of the collecteddata were recorded. Carbon half-life was calculated by using simple linearregression (Pearson-regression). To calculatep values a Kruskal-Wallistest was used.

Results

The lowest elimination half-life of ink particles in the labo-ratory animal’s bloodstream (t1/2 4 12.86 min) could beascertained in group 3 (gasless laparoscopic). Compared tothis group, the elimination half-life is significantly in-creased (t1/2 4 21.91 min; p value < 0.005) in group 2(laparoscopic with pneumoperitoneum). In group 1 (con-ventional laparotomy) the ink particles’ elimination half-liferanged below the one of group 2 (t1/2 4 16.1 min;p value< 0.1) (Fig. 1).

Discussion

Repeatedly, literature has described the phenomenon of theimplantation of tumor cells during laparoscopies [4, 10, 13].Why this problem occurs and how often are still unclear.Yet there are hypotheses according to which suboptimaltechnique and instrumentation are said to play a decisivepart in that matter [1, 12]. Under certain conditions, even thepneumoperitoneum seems to support the intraperitoneal dis-semination of tumor cells [8].

The mononuclear phagocyte system (MPS) representsan important component of the endogenous defense against

circulating tumor cells. A decreased perioperative, cell-conveyed immunoresponse correlates not only with an in-creased metastatic tumor spreading risk but also with post-operative infections; the phagocytotic activity of the MPSconsequently seems to be of significance when judging theetiology of implantation metastases during laparoscopic in-terventions.

The Carbon-Clearance test is a well-established methodof measuring the phagocytosic activity of the mononuclearphagocytes system (MPS). In conformity with the methoddescribed by Biozzi et al. [3] and Lemperle et al. [7], it isapplied in a modified form which offers the advantage ofmeasuring in vivo whole the extent of a hematogen antigenelimination. Since the phagocytosis activity depends on thetype and extent of an operation it, therefore, can be consid-ered a measure of the disturbance of the organism by anysuch interventions. Therefore, a highly standardized pro-ceeding, as well as the choice of an anesthesia method notinfluencing the test procedure, is a basic condition. From atechnical point of view, the rat fundoplication is a simpleoperation which, without any problems, can be carried outboth conventionally and laparoscopically [5].

As shown by the results of this investigation, the fastestway to eliminate carbon particles from the bloodstream bylocal macrophages of the MPS is via gasless laparoscopic(group 3). Compared with the conventional method (group1), the minor disturbance of the organism seems to be theminimally invasive access. The use of a pneumoperitoneum(group 2) has an even stronger influence on the antigenelimination from the circulating blood than the choice ofsurgical access. Although, in case of the rat, with a pneu-moperitoneum of 7 mmHg there is no long-term decrease inthe cardiac output, the intraperitoneal pressure seems tocause a reduced circulation speed of carbon particles in theblood. Further impacts that may prejudice the organism,apart from the circulatory influences of the CO2 pneumo-peritoneum, are not excluded. The results of the presentinvestigation show that a pneumoperitoneum in the rat is aninvasive arrangement leading to a significant reduction ofthe phagocytosis activity of the MPS.

Table 1. Nissen fundoplication was performed in three animal groups

Group 1 (n 4 10)Conventional laparotomyGroup 2 (n 4 10)Laparoscopy with pneumoperitoneumGroup 3 (n 4 10)Gasless laparoscopy

Fig. 1. Average carbon elimination half-life of the experimental groups,graphically described.

900

References

1. Allendorf JDF, Bessler M, Kayton ML, Whelan R, Treat MR, Nowy-grod R (1995) Tumor growth after laparotomy or laparoscopy. SurgEndosc 9: 49–52

2. Berguer R, Gutt CN (1994) Laparoscopic colon surgery in a rat model.A preliminary report. Surg Endosc 8: 1195–1197

3. Biozzi G, Benacerraf B, Halpern BN (1953) Quantitative study of thegranulopectic activity of the reticulo-endothelial system. II. Br J ExpPathol 34: 441–457

4. Childers JM, Aqua KA, Surwit EA, Hallum AV, Hatch KD (1994)Abdominal-wall tumor implantation after laparoscopy for malignantconditions. Obstet Gynecol 5: 765–769

5. Gutt CN, Berguer R, Stiegmann G (1993) Laparoskopie an der Ratte:Beschreibung einer neuen Technik. Zentralbl Chir 118: 631–634

6. Keller SE, Weiss JM, Schleifer SJ, Miller NE, Stein M (1983) Stress-induced suppression of immunity in adrenalectomized rats. Science221: 1301–1304

7. Lemperle G (1972) Der Funktionszustand des retikuloendothelialenSystems bei chirurgischen Erkrankungen. Habilitationsschrift Frank-furt/M15–17, 38–42

8. Mouiel J, Gugenheim J, Toouli J, Crafa F, Cursio R, Chastanet S(1995) Port-site recurrence of cancer associated with laparoscopic di-agnosis and resection: the European experience. Semin Laparosc Surg2(3): 167–175

9. Pollock RE, Lotzova E (1987) Surgical-stress-related suppression ofnatural killer cell activity: a possible role in tumor metastasis. NatImmunol Cell Growth Regul 6: 269–278

10. Ramos JM, Gupta S, Anthone GJ, Ortega AE, Simons AJ, Beart RWJr (1994) Laparoscopy and colon cancer. Is the port site at risk? Apreliminary report. Arch Surg 9: 897–899

11. Saba TM, Antikatzides TG (1976) Decreased resistance to intravenoustumor cell challenge during periods of reticuloendothelial depressionfollowing surgery. Br J Cancer 34: 381–386

12. Treat RM, Bessler M, Whelan RL (1995) Mechanisms to reduce in-cidence of tumor implantation during minimal access procedures forcolon cancer. Semin Laparosc Surg 2(3): 176–178

13. Walsh DC, Wattchow DA, Wilson TG (1993) Subcutaneous metasta-ses after laparoscopic resection of malignancy. Aust N Z J Surg 7:563–565

901

Cetrimide overdose

The intraperitoneal use of cetrimide 1% [6] has recentlybeen advocated as an adjunct to the treatment of hydatiddisease of the liver [4]. We would like to report the case ofa 2-year-old boy who died following such a treatment re-gime.

A 2-year-old Moroccan boy was admitted with generalmalaise and fever. Clinical examination revealed a mass inthe liver. A diagnosis of chronic granulomatous disease wasmade but concomitant hydatid disease of the liver could notbe excluded. We decided to puncture the mass in the liverunder laparoscopic control, after immersing the liver in ce-trimide 1%, a scolicidal agent, as has been advocated byKhoury et al. [4]. General anesthesia with endotracheal in-tubation was induced with intravenous thiopentone, sufen-tanyl, and vecuronium, and it was maintained with halo-thane 0.6–1% in oxygen-enriched air and with additionaldoses of sufentanyl and vecuronium. Monitoring includedECG, pulse oximetry, nasal temperature, intra-arterial andcentral venous pressure, and end-tidal CO2. A 5-mm trocarwas inserted through the infraumbilical fold using an opentechnique. Pneumoperitoneum was induced with CO2 at aninitial pressure of 5 mmHg and later at 8 mmHg (initial flowrate 2 l/min, later 5 l/min). No surface abnormalities of theliver were seen. The patient was positioned head down witha three-quarter right lateral tilt. The epigastrium was thenpunctured with a 14-gauge Abbocath and cetrimide 1% wasinstilled until the liver was completely submerged, whichrequired 1 l of the solution. No fluid was obtained on ul-trasound-guided laparoscopic puncture of the mass in theliver. After about 1 h, the instilled cetrimide solution wasaspirated and 800 ml was easily retrieved. Before removingthe laparoscope and closing the trocar wound, the peritonealcavity was rinsed three times with 1 l of normal saline. Thepatient remained stable during the procedure, which lasted13⁄4 h. No unexpected physiological changes were noteddespite extensive monitoring. However, after withdrawal ofhalothane, the patient failed to regain consciousness, and itwas noted that the pupils were dilated and did not respondto light. The patient was apneic with a general flaccid pa-ralysis. Although supplementary doses of sufentanyl andvecuronium had not been given for more than 1 h, neostig-mine, atropine, and naloxone were administered. However,this did not result in any improvement in the clinical con-dition of the child. Arterial blood analysis revealed a mild

respiratory and a severe metabolic acidosis (pH 7.05, pCO253 mmHg, base excess −15 mmol/l, pO2 157 mmHg, withan FiO2 of 40%). Despite the adjustment of the ventilatorand the administration of sodium bicarbonate, the metabolicacidosis continued to increase over the ensuing hours. Asmethaemoglobinemia formation is a known complication ofthe use of cetrimide, a blood sample was taken to measurethis and 15 mg methylene blue was given intravenously [1,4]. The methemoglobin level, however, later proved to bewithin normal limits (less than 0.6%). The laboratory de-tected mild hemolysis, but neither macroscopic hematureianor hemoglobinuria was observed. Both may be associatedwith absorption of cetrimide [2, 3]. As the patient remainedunconscious with a flaccid paralysis and fixed dilated pu-pils, he was transferred to the intensive care unit, where, 90min after the end of surgery, he became cardiovascularlyunstable and subsequently died 1 h later, despite vigorousresuscitation measures. Postmortem examination failed toreveal a cause for the patient’s death.

A number of complications have been described in as-sociation with internal use of cetrimide, including coma [3],fixed dilated pupils [2, 3], flaccid paralysis [2, 3], metabolicacidosis [2, 5], cardiovascular collapse [2, 5], cardiac arrest[2], methemoglobinemia [1, 4], hemolysis [2], hemoglobin-uria, and hematuria [2]. None of these reported complica-tions, however, has been lethal or has resulted in long-termmorbidity.

We were not able to confirm cetrimide absorption byplasma bromide analysis, as has been suggested [5]. Asbromide ions (Br−) are distributed over 25% of the bodyweight (i.e., 4 l in our patient), absorption of 100 ml ce-trimide 1% (i.e., 60 mg/kg in our patient, at least twice thereported lethal dose [3]) will elevate plasma Br− to a maxi-mum value of 0.75 mmol/l. This is below the detection limitof the commonly used gold-chloride-analysis technique forbromide. As a result of our experience, we consider thatcetrimide is a dangerous and potentially lethal agent forinternal application. It should certainly not be used intra-peritoneally as has been recently advocated [4]. Even treat-ment of the hydatid cyst cavity with cetramide may result inlife-threatening complications [3]. We conclude that the useof cetrimide should be restricted to external applicationonly.

References

1. Baraka A, Yamut F, Walid N (1980) Cetrimide induced methaemoglo-binaemia after surgical excision of hydatid cyst. Lancet II: 88–89Correspondence to:N. M. A. Bax

Surg Endosc (1997) 11: 967–968

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

2. Gode GR, Jayalakshmi TS, Kalla GN (1975) Accidental intravenousinjection of cetrimide. Anaesthesia 30: 508–510

3. Klouche K, Charlotte N, Kaaki M, Be´raud JJ (1994) Coma and hemo-lysis after cetrimide washout of epidural hydatid cyst. Intensive CareMed 20: 613

4. Khoury G, Jabbour-Khoury A, Bikhazi K (1996) Results of laparoscop-ic treatment of hydatid cysts of the liver. Surg Endosc 10: 57–59

5. Momblano P, Pradere B, Jarrige N, Concina D, Bloom E (1984) Meta-bolic acidosis induced by cetrimonium bromide. Lancet II: 1045

6. Reynolds JEF (1993) Martindale The extra pharmacopoeia. 30th ed.The Pharmaceutical Press, London, p. 787

N. M. A. Bax1

D. C. van der Zee1

N. M. Turner 2

C. M. A. Rademaker3

R. J. B. J. Gemke4

1 Department of Pediatric SurgeryUniversity Children’s Hospital WilhelminaP.O. Box 180093501 A UtrechtThe Netherlands

2 Department of AnesthesiologyUniversity Children’s Hospital WilhelminaP.O. 180093501 CA UtrechtThe Netherlands

3 Department of PharmacyUniversity Children’s Hospital WilhelminaP.O. 180093501 CA UtrechtThe Netherlands

4 Department of Intensive CareUniversity Children’s Hospital WilhelminaP.O. 180093501 CA UtrechtThe Netherlands

968

Acute appendicitis—a clear-cut case in men, a guessing game in youngwomen

A prospective study on the role of laparoscopy

P. J. Borgstein, R. V. Gordijn, Q. A. J. Eijsbouts, M. A. Cuesta

Department of Surgery, Academic Hospital, Vrije Universiteit, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands

Received: 18 October 1996/Accepted: 2 April 1997

AbstractBackground:The aggressive surgical approach to patientssuspected of having acute appendicitis for fear of perfora-tion, and the inaccuracy of available diagnostic methodslead to an unacceptably high negative appendicectomy rate,especially in young women, in whom gynecological disor-ders frequently mimic appendicitis. Our objectives were todetermine the value of diagnostic laparoscopy in women ofchild-bearing age to reduce the number of negative laparot-omies and establish the correct diagnosis to allow promptand appropriate treatment.Methods:161 consecutive adult female patients under 50years of age with a clinical diagnosis of acute appendicitisunderwent diagnostic laparoscopy prior to the planned ap-pendicectomy. If an inflamed appendix was found, appen-dicectomy was usually done through a muscle-splitting Mc-Burney incision. Other diagnoses were treated accordingly.A normal appendix was not removed. Results were com-pared to a group of 42 similar patients in whom the lapa-roscopy was omitted for various reasons, to 23 postmeno-pausal women, and to all 137 male adults, directly operatedby the McBurney approach.Results:After laparoscopy, 55% of the patients requiredappendicectomy for appendicitis while in 23% a gyneco-logical diagnosis was made in spite of previous examinationby a gynecologist. Fourteen percent had a negative laparos-copy. There were no false-negative results. The negativeappendicectomy rate after laparoscopy was 5% due to twofalse positives and eight laparoscopy failures. In the groupof fertile females who escaped laparoscopy the negativeappendicectomy rate was 38%. The respective rates forpostmenopausal women and men were 4% and 8%.Conclusions:All women of child-bearing age suspected of

having acute appendicitis should undergo diagnostic lapa-roscopy prior to the planned appendicectomy, regardless ofthe certainty of the preoperative diagnosis. This is currentlythe only way to reduce the negative appendicectomy rateand establish a correct diagnosis allowing prompt and ap-propriate treatment. In male patients and postmenopausalwomen one may proceed directly to emergency appendi-cectomy.

Key words: Acute appendicitis — Diagnostic laparoscopy

Acute appendicitis is the most common surgical emergencyin the western world and demands accurate early diagnosisto prevent progression to perforation, with its associatedhigh morbidity and mortality, and which can result in latecomplications such as adhesions and infertility in youngwomen [16, 23]. The widely accepted surgical approach,appendicectomy through a muscle-splitting McBurney inci-sion, leads to a high rate of normal appendices being re-moved under the suspicion of an acute appendicitis. Thereported negative appendicectomy rate for men varies from7% to 15%, whereas that for women of child-bearing agelies between 22% and 47% [4, 6].

It has long been obvious that more accurate methods ofdiagnosing acute appendicitis, besides clinical suspicion,are required to prevent potentially harmful, unnecessarysurgery especially in young women. Several methods havebeen described to reduce the diagnostic error rate of acuteappendicitis [13], but only three have been shown to be ofvalue in clinical practice—namely, ultrasonography, com-puter-assisted diagnosis, and diagnostic laparoscopy.

Ultrasonography has been used since 1986 in diagnos-ing acute appendicitis and has been shown, in large series,to have a sensitivity of 75% to 89% and a specificity of 86%to 100% [20]. Advantages are the noninvasive techniqueand the possibility of diagnosing alternative abdominal con-

Presented at the 4th International Congress of the European Association forEndoscopic Surgery, Trondheim, Norway, June 1996Correspondence to:M. A. Cuesta

Surg Endosc (1997) 11: 923–927

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

ditions such as mesenteric adenitis, terminal ileitis, andsome gynecological disorders. The main disadvantage isthat ultrasonography is strongly operator-dependent. An ex-perienced radiologist must be available 24 h a day toachieve a clinically useful sensitivity [1] (option not avail-able in our hospital during the study).

Structured data collected from history-taking and physi-cal examination can be entered into previously programmedcomputers to produce a diagnostic probability for acute ap-pendicitis or nonspecific abdominal pain [8, 10]. Althoughearly studies reported a decline in the negative appendicec-tomy rate, initial enthusiasm has been dampened by variouslimitations and programming difficulties.

Laparoscopy has long been used by gynecologists todiagnose lower abdominal pain, but prior to 1990 it did notachieve widespread attention as a diagnostic aid in acuteappendicitis because the failure rates were unacceptablyhigh (up to 28%) due to difficulties in visualizing the ap-pendix [3, 22]. The lack of specially designed equipmentand general expertise unfortunately made this diagnosticlaparoscopy more a theoretical tool used by a few, withoutany practical adoption in most surgical departments. Sincethe era of laparoscopic surgery, however, the significantincrease in experience with laparoscopic techniques has ledto several reports on the value of laparoscopy as a methodof reducing the negative laparotomy rate in women of afertile age. The risks involved are extremely low, whileimportant advantages are the ability to directly view theappendix and diagnose other possible causes of the acuteabdomen [7].

Based on our own experience in diagnostic and thera-peutic laparoscopic surgery, a prospective study was startedin April 1991 with the following objectives:

1. To determine the value of diagnostic laparoscopy in alladult female patients planned to undergo laparotomy forsuspected acute appendicitis

2. To compare results with appendicectomy in male pa-tients treated directly through a muscle-splitting McBur-ney incision

3. To introduce the technique of laparoscopy to all resi-dents and surgeons in our General Surgical departmentand yet maintain the classical McBurney appendicec-tomy so crucial for surgical training

Patients and methods

From April 1991 to September 1995, a prospective study was done on allconsecutive adult patients admitted with the clinical signs of acute appen-dicitis and planned to undergo laparotomy. The total of 363 patients weredivided into three groups:Group IGroup IIGroup III

Women 16 to 50 years of age (203 patients)Women older than 50 years (23 patients)Men 16 years and older (137 patients)

The decision to subdivide the female patients into two groups was based onthe relatively high frequency of gynecological disorders found in women ofchild-bearing age. Fifty years was the arbitrary limit chosen between pre-and postmenopause.

Diagnostic workup included history taking, complete physical exami-nation, and routine blood and urine analysis. Protocol dictated that once thesurgeon decided that there was an indication for urgent appendicectomy, allfemale patients were also seen by a gynecologist, who performed special-ized physical examination, cervical culture, and transvaginal ultrasonog-

raphy to exclude obvious gynecological diseases. Patients were first seenby residents with subsequent supervision by different fellows of Gynecol-ogy (nine) and Surgery (seven). If the diagnosis of acute appendicitis wasupheld, a diagnostic laparoscopy was done. Unfortunately not all patientsactually underwent diagnostic laparoscopy for several reasons. Therefore,group I contains subgroup Ia (42 patients) in whom a diagnostic laparos-copy could not be performed and patients underwent directly an appendi-cectomy and subgroup Ib of 161 patients who underwent the diagnosticlaparoscopy prior to the appendicectomy.

All male patients with a clinical suspicion of acute appendicitis under-went appendicectomy, without previous diagnostic laparoscopy, through amuscle-splitting McBurney incision (Fig. 1).

Diagnostic laparoscopy was performed under general anesthesia. Pro-phylactic antibiotics (metronidazole 500 mg) were administered intrave-nously and the urinary bladder was catheterized. A two-port approach wasgenerally used (Fig. 2), infraumbilical and suprapubic, allowing adequateinspection of the lower abdomen. Occasionally an extra 5-mm trocar wasrequired in the right abdomen to dissect a retrocecal appendix or com-pletely visualize the upper abdomen. If an acutely inflamed appendix wasfound, or if it could not be excluded, appendicectomy was performedthrough a muscle-splitting McBurney incision or laparoscopically depend-ing on the surgeons experience. Obvious gynecological disorders weretreated by the gynecologist appropriately. A visually normal appendix wasnot removed.

Results

Group I

Group I consisted of 203 female patients from 16 to 50years (median age 24) all suspected of suffering from acuteappendicitis.

Fig. 1. Scheme of the protocol.

924

Subgroup Ia

In 42 women diagnostic laparoscopy was omitted for vari-ous reasons: 31 during the first study year due to the duty-surgeon’s inexperience, eight instrument failures, and threeadvanced pregnancies (over 26 weeks). In all 42 patients amuscle-splitting incision was directly performed. Only 26(62%) were found to have acute appendicitis, in 16 (38%) anormal appendix was removed, as confirmed by pathologi-cal examination.

Subgroup Ib

Some 161 women underwent diagnostic laparoscopy priorto the planned appendicectomy, including eight who werebetween 8 and 26 weeks pregnant. (Table 1) In 88 (55%)patients acute appendicitis was diagnosed by laparoscopy.Seventy-five subsequently underwent appendicectomythrough a muscle-splitting incision; in 11 the appendix wasremoved laparoscopically. One case required a median lap-arotomy because of a perforated appendicular mass, and inanother it was decided to treat the appendicular mass con-servatively. There were two false-positive results due tomisinterpretation of the visual aspect of the appendix. His-tology showed no signs of acute inflammation.

In 65 (40%) women the appendix was determined to benormal at laparoscopy and it was not removed (Table 1)except in two cases with appendicular faecoliths. In 37 ofthese patients (23% of subgroup Ib) a gynecological disor-der was diagnosed (Table 2); four were operated laparo-scopically, four required a Pfannenstiel incision and 29could be treated conservatively.

Diverse diagnoses were found in five patients: a small

liver laceration from a forgotten accident treated by lapa-roscopy, partial omental necrosis resected throughminilaparotomy, bowel adhesions taken down laparoscopi-cally, and two other patients had faecoliths in the appendix,this requiring appendicectomy.

Twenty-three patients had a negative laparoscopy. Duringtheir hospital stay, four of these were diagnosed as havinggastroenteritis and four had urinary tract infections or stones.Five patients later suffered repeated attacks of nonspecific ab-dominal pain. Only ten patients (6% of subgroup Ib) haveremained undiagnosed at completion of the follow-up.

There were no false-negative laparoscopies. None of the63 patients in whom a visually normal appendix was left in situwere readmitted for acute appendicitis during the follow-up.

In eight patients (5%) laparoscopy failed because offailure to insufflate the abdomen (three cases) or inability tocompletely visualize the appendix due to adhesions or aretrocecal position (five cases). All eight consequently un-derwent appendicectomy through a muscle-splitting Mc-Burney incision. In only four was the appendix found to beacutely inflamed.

Only one patient suffered a direct complication due tothe laparoscopy. Bleeding from a trocar site required ab-dominal wall exploration, which was performed through themuscle-splitting McBurney incision.

Group II

Group II consisted of 23 postmenopausal women 50 to 86years old (median age 54). During the learning period, 13patients escaped laparoscopy and directly underwent appen-dicectomy. Acute appendicitis was diagnosed in 22 patients(96%). The remaining patient had a perforated peptic ulcerrequiring an additional vertical incision.

Group III

Group III comprised 137 male adults, median age 23 years.All were operated on through the muscle-splitting McBur-ney approach for suspected acute appendicitis. An inflamed

Fig. 2. Placement of the trocars in diagnostic laparoscopy.

Table 1. Group Ib; diagnostic laparoscopy in women 16–50 years

Laparoscopic diagnosis Number %

Acute appendicitis 88 (55%)Normal appendix

Gynecological disorder 37 (23%)Diverse diagnoses 5 (3%)Negative laparoscopy 23 (14%)

Failed laparoscopy 8 (5%)

Total 161

Table 2. Group Ib; gynaecological disorders found at laparoscopy inwomen 16–50 years (37/161 patients)

Pelvic hemoperitoneum 15Pelvic inflammatory disease 13Ovarian/cyst torsion 3Endometriosis 4Unknown pregnancy/ectopic 2

925

appendix was found in 126 patients (92%). In 11 patients(11%) a normal appendix was removed.

Operation time

The mean operating time for diagnostic laparoscopy was 25min while that for open appendicectomy was 45 min.

Hospital stay

The mean length of hospital stay was 1.7 days whendiagnostic laparoscopy was the only procedure. If appendi-cectomy was performed the stay was similar whether theappendix was inflamed or not—respectively, 3.6 and 3.4days.

Discussion

The main purpose of this prospective study was to deter-mine the value of laparoscopy in establishing the diagnosisin young female patients suspected of having an acute ap-pendicitis, and to reduce the negative laparotomy rate.Acute appendicitis is probably the only surgical diseasewhere a diagnostic accuracy of only 70% to 75% is stillaccepted. Many methods have been described in an attemptto reduce the negative appendicectomy rate—close obser-vation, specific laboratory tests, ultrasonography, computer-aided diagnosis, scoring systems, barium enema studies,computed tomography, radioisotope imaging, and evenperitoneal cytology [11, 18]. However, in spite of severalpromising initial reports, none of these has achieved wide-spread clinical application.

The greatest concern is clearly in women of a fertile agein whom gynecological disorders frequently masquerade asacute appendicitis. Negative appendicectomy rates as highas 40% are no longer acceptable. Since the early 1980s, agroup of authors has repeatedly pointed out the specificproblem of the wide spectrum of diagnosis of acute lowerabdominal pain in child-bearing age females in contrast tothe simplicity of diagnosis in male patients. They suggestedthe use of diagnotic laparoscopy in young females withacute lower abdominal pain [18, 26]. Patients with an atypi-cal presentation for appendicitis underwent laparoscopy,while straightforward cases directly proceeded to appendi-cectomy. In the latter group 25% still had a normal appendixremoved, while in the former only one-third required ap-pendicectomy. The authors also stressed the capacity oflaparoscopy to diagnose other abdominal conditions, found

in 50% of their patients. Diagnosing those problems facili-tated prompt and appropriate care in a majority of thesepatients. Despite this, a small group of 17% of patientsremained undiagnosed. Since the widespread application ofsurgical laparoscopy in 1990, there have appeared a fewprospective randomized studies on diagnostic laparoscopyin young female patients. Jadallah et al. [14, 17] have com-pared the diagnostic value of laparoscopy with the directMcBurney approach in female patients with clinical diag-nosis of acute appendicitis. While both groups had the samepercentage of 12 acute appendicitis (around 60%), after lap-aroscopy 38% of the patients did not require appendicec-tomy, and in all these patients a gynecological diagnosiswas made.

Our present study shows that in 161 women of child-bearing age, suspected of having acute appendicitis, only55% (88 patients) required appendicectomy after laparos-copy. In 23% (37 patients) a gynecological disorder wasdiagnosed in spite of prior gynecological examination. Only6% (10 patients) ultimately remained undiagnosed. Whencompared to the subgroup of 42 women in whom laparos-copy was omitted, the negative appendicectomy rate wasconsequently reduced from 38% to 5% following diagnosticlaparoscopy. (In only eight cases the removed appendix wasnot acutely inflamed) (Table 3).

In sharp contrast, 96% of 23 women over 50 years and92% of 137 men did indeed have acute appendicitis at op-eration, with respective negative appendicectomy rates of4% and 8%.

While there were two false-positive laparoscopic diag-noses of appendicitis, there have been no false-negativeresults after leaving a visually normal appendix in situ, evenin the absence of other pathology. In this context, Wang etal. [24] recently published an interesting report on micro-scopical markers of inflammation found in apparently nor-mal appendices. Traditionally, a normal appendix found atoperation is nonetheless removed for two reasons: fear ofmissing an appendicitis or progression to it, and because aMcBurney incision scar traditionally means the patient hasundergone appendicectomy. These concepts may now bechanged, and some authors make a plea for conservation ofa normal appendix to reduce postoperative morbidity andbecause the appendix may be required in the future for otherpurposes [25].

Finally, the heated debate on whether laparoscopic ap-pendicectomy should become the new gold standard stillcontinues. In a recent overview of published reports byHeinzelmann et al. [12], there were no clear benefits oflaparoscopic appendicectomy for operation time, hospitalstay, or complication rate. Only postoperative comfort and

Table 3. Acute appendicitis; comparison between all three groups (363 patients)

Patient group (No.)Acuteappendicitis (No.)

Negativeappendicectomy (No.)

Group Ia Women 16–50 yearswithout laparoscopy

(42) 62% (26) 38% (16)

Group Ib Women 16–50 yearswith laparoscopy

(161) 56% (90) 5% (8)

Group II Women >50 years (23) 96% (22) 4% (1)Group III Men >16 years (137) 92% (126) 8% (11)

926

cosmetics were subjectively better. Undoubtedly, there willbe advantages in certain categories of patients, but futurestudies are needed to define the exact role for laparoscopicappendicectomy.

In conclusion, we are convinced that with the presentexperience and modern techniques at our disposal, allwomen of child-bearing age suspected of having acute ap-pendicitis should undergo diagnostic laparoscopy in order toreduce the rate of negative appendicectomy, but also toestablish the correct diagnosis allowing prompt and appro-priate treatment. Laparoscopy, however, must not be used asan alternative to good clinical judgment.

References

1. Editorial (1987) A sound approach to the diagnosis of acute appendi-citis. Lancet 1(8526): 198–200

2. Addiss DG, Shaffer N, Fowler BS, Tauxe RV (1990) The epidemiol-ogy of appendicitis and appendectomy in the United States. Am JEpidemiol 132(5): 910–925

3. Anderson JL, Bridgewater FH (1981) Laparoscopy in the diagnosis ofacute lower abdominal pain. Aust N Z J Surg 51(5): 462–464

4. Andersson RE, Hugander A, Thulin AJ (1992) Diagnostic accuracyand perforation rate in appendicitis: association with age and sex of thepatient and with appendicectomy rate. Eur J Surg 158(1): 37–41

5. Berry J Jr, Malt RA (1984) Appendicitis near its centenary. Ann Surg200(5): 567–575

6. Chang FC, Hogle HH, Welling DR (1973) The fate of the negativeappendix. Am J Surg 126(6): 752–754

7. Cuesta MA, Borgstein PJ, Meijer S (1993) Laparoscopy in the diag-nosis and treatment of acute abdominal conditions. Clinical review.Eur J Surg 159(9): 455–456

8. De Dombal FT, Horrocks JC (1978) Use of receiver operating char-acteristic (ROC) curves to evaluate computer confidence threshold andclinical performance in the diagnosis appendicitis. Methods Inf Med17(3): 157–161

9. Diehl JT, Eisenstat MS, Gillinov S, Rao D (1981) The role of perito-neoscopy in the diagnosis of acute abdominal conditions. Cleve Clin Q48(3): 325–330

10. Edwards FH, Davies RS (1984) Use of a Bayesian algorithm in thecomputer-assisted diagnosis of appendicitis. Surg Gynecol Obstet158(3): 219–222

11. Eriksson S (1996) Acute appendicitis—ways to improve diagnosticaccuracy (clinical review). Eur J Surg 162: 435–442

12. Heinzelmann M, Simmen HP, Cummins AS, Largiader F (1995) Islaparoscopic appendectomy the new ‘gold standard’? Arch Surg130(7): 782–785

13. Hoffmann J, Rasmussen OO (1989) Aids in the diagnosis of acuteappendicitis [see comments]. Br J Surg 76(8): 774–779

14. Jadallah FA, Abdul-Ghani AA, Tibblin S (1994) Diagnostic laparos-copy reduces unnecessary appendicectomy in fertile women. Eur JSurg 160: 41–45

15. Lewis FR, Holcroft JW, Boey J, Dunphy JE (1975) Appendicitis: acritical review of diagnosis and treatment in 1000 cases. Arch Surg110: 677

16. Mueller BA, Daling JR, Moore DE, Weiss NS, Spadoni LR, StadelBV, Soules MR (1986) Appendectomy and the risk of tubal infertility.N Engl J Med 315(24): 1506–1508

17. Olsen JB, Myren CJ, Haahr PE (1993) Randomized study of the valueof laparoscopy before appendicectomy [see comments]. Br J Surg80(7): 922–923

18. Paterson-Brown S, Eckersley JR, Sim AJ, Dudley HA (1986) Lapa-roscopy as an adjunct to decision making in the ‘acute abdomen’. BrJ Surg 73(12): 1022–1024

19. Pieper R, Kager L, Nasman P (1982) Acute appendicitis: a clinicalstudy of 1018 cases of emergency appendectomy. Acta Chir Scand148(1): 51–62

20. Puylaert JB, Rutgers PH, Lalisang RI, de Vries BC, van der Werf SD,Dorr JP, Blok RA (1987) A prospective study of ultrasonography inthe diagnosis of appendicitis. N Engl J Med 317(11): 666–669

21. Silberman VA (1981) Appendectomy in a large metropolitan hospital.Retrospective analysis of 1,013 cases. Am J Surg 142(5): 615–618

22. Spirtos NM, Eisenkop SM, Spirtos TW, Poliakin RI, Hibbard LT(1987) Laparoscopy—a diagnostic aid in cases of suspected appendi-citis. Its use in women of reproductive age. Am J Obstet Gynecol156(1): 904

23. Tanphiphat C, Chittmittrapap S, Prasopsunti K (1987) Adhesive smallbowel obstruction. A review of 321 cases in a Thai hospital. Am J Surg154(3): 283–287

24. Wang Y, Reen DJ, Puri P (1996) Is a histologically normal appendixfollowing emergency appendicectomy always normal? Lancet347(9008): 1076–1079

25. Wheeler RA, Malone PS (1991) Use of the appendix in reconstructivesurgery: a case against incidental appendicectomy. Br J Surg 78(11):1283–1285

26. Whitworth CM, Whitworth PW, Sanfillipo J, Polk HC Jr (1988) Valueof diagnostic laparoscopy in young women with possible appendicitis.Surg Gynecol Obstet 167(3): 187–190

927

Hemodynamics during laparoscopic extra- andintraperitoneal insufflation

An experimental study

J. J. G. Bannenberg,1 B. M. P. Rademaker,2 F. M. J. A. Froeling,3 D. W. Meijer 1

1 Department of Surgery, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands2 Department of Anesthesiology, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands3 Red Cross Hospital, The Hague, The Netherlands

AbstractBackground:Total extraperitoneal laparoscopic surgery isan alternative to the laparoscopic transperitoneal route;however, its effects on hemodynamics have not been ad-equately studied. This experimental study compared the ef-fects of intraperitoneal insufflation and extraperitoneal in-sufflation on hemodynamics and oxygen transport.Methods:Sixteen pigs were randomly assigned for intraper-itoneal insufflation or extraperitoneal insufflation with 15mmHg carbon dioxide. Hemodynamic and oxygen transportparameters were taken during an hour of insufflation andanalyzed for statistical differences.Results:During extraperitoneal CO2 pneumoperitoneumcentral venous filling pressures (central venous pressure,pulmonary capillary wedge pressure and mean pulmonaryarterial pressure) and end-tidal CO2 increased slower but toa similar magnitude in comparison to intraperitoneal insuf-flation. Cardiac output and indices of oxygen consumptionand oxygen delivery were equally affected by both types ofinsufflation. Arterial CO2 pressure increased significantlymore during intraperitoneal insufflation.Conclusion:The data from this study suggest that extraperi-toneal insufflation might result in less cardiovascular im-pairment than intraperitoneal insufflation.

Key words: Laparoscopic surgery — Pneumoperitoneum— Hemodynamics — Extraperitoneal — Intraperitoneal

Laparoscopic procedures on extraperitoneally located ana-tomical structures—such as nephrectomy, adrenalectomy,lumbar sympathectomy, para-aortic lymph node sampling,

inguinal hernia repair, and bladder neck colposuspension—have been introduced into the surgical spectrum for clinicaluse [1–5, 10, 12]. The directness of the extraperitoneal ap-proach and contraindications for transperitoneal endoscopicsurgery—such as obesity, inadequate bowel preparation,and intraperitoneal adhesions—are factors that might makethe extraperitoneal approach more attractive. This type oflaparoscopic surgery may, however, be technically moredemanding than its transperitoneal counterpart because ofthe limited view, the restricted working space, and the ab-sence of familiar landmarks. This might result in prolongedprocedures affecting the circulatory basis of the patient.

Pneumoperitoneum for laparoscopic surgery has beenassociated with hemodynamic changes. Clinical studies dur-ing intraperitoneal insufflation with carbon dioxide showedthat arterial blood pressure increases and cardiac output de-creases [8, 13]. Few data exist on the hemodynamic effectsof laparoscopy in an extraperitoneally created cavity. If anextraperitoneal laparoscopic procedure results in more cir-culatory depression, its potential usefulness in humanswould be limited. In this experimental study we comparedthe effects of extraperitoneal and intraperitoneal laparoscop-ic insufflation on the hemodynamic parameters in a porcinemodel.

Materials and methods

Sixteen pigs (27–35 kg) were used in this study. Anesthesia was inducedwith an i.m. injection of azaperon (12 mg/kg) and atropine (1 mg). Afterendotracheal intubation the lungs of the animals were mechanically ven-tilated at a rate of 12 breaths/min with a tidal volume of 10 ml/kg, with amixture of oxygen in air (FIO2 4 0.4). Anesthesia was maintained with0.7–1.0% halothane (inspired concentration) and the infusion regimen con-sisted of lactated Ringer’s solution at a rate of 4 ml/kg/h during the ex-periment. The right internal jugular vein was exposed and a flow-directedpulmonary arterial catheter (Baxter, Americ Edwards Laboratories, Irvine,CA, U.S.A.) was inserted and floated into the pulmonary artery. A Wallace16-gauge catheter was placed in the right brachial artery for arterial pres-sure measurements and arterial blood-gas sampling.

Correspondence to:J. J. G. Bannenberg, Department of Surgery, Aca-demic Medical Centre, University of Amsterdam, Meibergdreef 9, 1105AZ Amsterdam, The Netherlands

Surg Endosc (1997) 11: 911–914

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

On a random basis eight pigs were assigned to extraperitoneal insuf-flation and eight to intraperitoneal insufflation with carbon dioxide.Through a modified open Hasson technique, as described by Horattas [7],a disposable 10-mm trocar was placed, extraperitoneally or intraperitoneal-ly, and connected to a pressure-controlled carbon dioxide insufflator (Elec-tronic Laparoflator 263400-20, Storz-Endoskop, Switzerland) set at 15mmHg pressure.

We attempted to create an extraperitoneal cavity of approximately 1 lof CO2 gas. From previous animal experiments we noted that this volumeis sufficient to create a cavity large enough to be able to perform endo-scopic retroperitoneal nephrectomy, para-aortic lymph-node sampling, orlumbar sympathectomy [1–3]. On average, the amount of gas necessary tocreate a 15-mmHg intraabdominal pressure in these animals was 5 l CO2.

Hemodynamic measurements were recorded using disposable transduc-ers (Gould, U.S.A.). Measurements were made for cardiac output (mean offour determinations by thermodilution, using room-temperature normalsaline injectate (5 ml), Edwards Laboratories, Santa Ana, CA, U.S.A.),heart rate, mean arterial blood pressure, central venous pressure, meanpulmonary arterial pressure, pulmonary capillary wedge pressure, and end-tidal carbon dioxide. Arterial and central venous blood samples were takenfor gas analyses of arterial oxygen, arterial carbon dioxide pressure, andmixed venous oxygen saturation (ABL II, Radiometer, Copenhagen, Den-mark).

After preoperative workup the pigs were positioned in the supine po-sition and allowed a stabilization period of 20 min before control mea-surements were taken. After starting the insufflation, measurements weremade as follows; 1 min after insufflation with carbon dioxide, 5 min, 10min, 15 min, 30 min, 45 min, and 60 min after insufflation. After the lastsample was taken the pneumoperitoneum was desufflated and a controlsample was taken after 10 min of desufflation. Blood samples were drawnduring each measurement.

Results are expressed as mean ± SD. Data was analyzed with two-wayanalysis of variance for repeated measures. When indicated, differencesbetween means were analysed using pairedt-tests with Bonferroni correc-tion for multiple comparisons. Blood-gas measurements were analyzedwith the Mann-Whitney U test;p values of <0.05 were considered statis-tically significant.

Results

Hemodynamic changes with both insufflation methods areshown in Fig. 1. Heart rate did not change significantlyduring either of the insufflation methods. Although meanarterial blood pressure and cardiac output did increase sig-nificantly during intraperitoneal insufflation, but not duringextraperitoneal insufflation, the differences between bothmethods were not significant.

Filling pressures such as central venous pressure, mean

pulmonary arterial pressure, and pulmonary capillary wedgepressure increased significantly during the first minutes ofintraperitoneal insufflation, reaching a plateau after ap-proximately 10 min. These pressures showed a moregradual increase during extraperitoneal insufflation, reach-ing a plateau after 45 min. Differences in this respect weresignificant between both methods.

The gas exchange parameters are shown in Fig. 2. Thereis a rapid increase of end-tidal CO2 during the first fewminutes of intraperitoneal insufflation, reaching a plateauafter 10 min. End-tidal CO2 during extraperitoneal insuffla-tion increases at a slower pace and the magnitude of theincrease is significantly lower than with intraperitoneal in-sufflation.

During both insufflation methods arterial CO2 pressureincreases, but the increase during intraperitoneal insuffla-tion is significantly larger in magnitude than with extraperi-toneal insufflation. pH decreased similarly in both groups.The arterial oxygen pressure and central venous oxygensaturation did not change significantly during either of theinsufflation procedures.

Discussion

Extraperitoneal laparoscopic surgery is rapidly becoming anestablished route for surgical procedures. The effects ofextraperitoneal insufflated carbon dioxide on gas exchangehave been studied previously; however, the effects on he-modynamics are not yet clear. Our data concerning hemo-dynamic changes suggests that extraperitoneal insufflationwith carbon dioxide is associated with hemodynamics simi-lar to those observed during intraperitoneal insufflation. Ex-traperitoneal insufflation is associated with less rapid in-creases in central venous filling pressures compared to in-traperitoneal insufflation.

The mechanisms that are responsible for the hemody-namic changes during laparoscopic surgery appear to bemultifactorial. Both pressure and pharmacological effects ofthe insufflated gas may affect hemodynamics. Pressure gra-dients will affect venous return. Increases in venous returnsecondary to increased intraperitoneal pressure have beenreported to augment cardiac output. In general, however,

Fig. 1. Hemodynamic variablesduring extraperitoneal (–h–) andintraperitoneal (–s–) carbondioxide insufflation. Each data pointrepresents the mean ± SD (n 4 8).

912

venous return is expected to decrease during intraperitonealinsufflation, as indicated by decreases in cardiac perfor-mance. This apparent contradiction may be explained by atime-dependent phenomenon. Initially, blood will besqueezed out of the abdominal cavity to the heart, causing atransient increase in cardiac output. Secondly, sustained in-creases in intraabdominal pressure will ultimately impedevenous return and depress cardiac output. In addition to thepressure effects, pharmacological effects of the insufflationgas will affect hemodynamics during laparoscopy. The ab-sorption of CO2 from the insufflation cavity, resulting inmild hypercarbia, is associated with sympathetic stimula-tion, which may increase heart rate, blood pressure, sys-temic vascular resistance, and cardiac output. On the otherhand, severe acidemia and hypercapnia may depress heartperformance, secondary to decreases in myocardial ino-tropy.

We did observe significant increases in cardiac outputand blood pressure during intraperitoneal insufflation, butnot during extraperitoneal insufflation. Heart rate did notincrease with either method. A clinical study [15] recentlyconfirmed this data. However, in this nonrandomized study,measurements were only taken just before and at the end ofthe insufflation period. Furthermore, cardiac output andcentral venous filling pressures were not measured. Ourresults concerning cardiac output are in contrast to mosthuman studies, which report moderate decreases in cardiacoutput during intraperitoneal insufflation. Studies per-formed in pigs may give different hemodynamic resultscompared to studies in humans. There are no data on humanstudies measuring cardiac output and central filling pres-sures during extraperitoneal laparoscopy. We could notdemonstrate a significant difference between the extra- andthe intraperitoneal approach with respect to cardiac output,blood pressure, and heart rate. Indeed, central venous oxy-gen saturation, which reflects global oxygen delivery to thetissues, was not affected by either approach. Therefore, itseems reasonable to assume that cardiac performance isaffected similarly by both the extra- and the intraperitonealapproach.

Central venous filling pressures increased during bothintra- and extraperitoneal insufflation. However, during in-traperitoneal insufflation, central filling pressures increased

faster than during extraperitoneal insufflation. Maximumpressures were reached on average 30 min earlier during theintraperitoneal approach. A possible explanation for thisphenomenon might be that pressures are transmitted morequickly into the thorax during intraperitoneal insufflation.The soft tissues that cover the extraperitoneal cavity mayserve as a buffer and may therefore delay the transmissionof the pressure into the thorax. In theory, this might beadvantageous for patients with limited cardiac reserve.However, whether the differences between both methodswith respect to the central venous filling pressures are ofany clinical consequence remains to be determined. Theduration of an extraperitoneal approach is on average longerthan 45 min. Therefore, it is to be expected that during anextraperitoneal procedure central venous filling pressuresare elevated throughout a significant part of the procedure.

Gas-exchange parameters showed an earlier and stron-ger tendency toward the development of respiratory acidosisin the intraperitoneal insufflation group. These findings arein agreement with others that showed, in an experimentalstudy in dogs, that if the insufflated gas is limited to theretroperitoneal space, the absorption of CO2 appears to bereduced compared to intraperitoneal insufflation [14]. Thesefindings contrast with a study that showed more markedCO2 diffusion into the body during extraperitoneal than dur-ing intraperitoneal CO2 insufflation [11]. However, in thatstudy three different operations were evaluated, whichmight have influenced the results. Two mechanisms con-tribute to the development of hypercarbia during laparosco-py: absorption from the intra- or extraperitoneal cavity andincreased physiologic dead space ventilation. Lister showedin an experimental study on pigs that CO2 absorption fromthe abdominal cavity depends on the level of intraabdominalpressure [9]. He hypothesized that by increasing the intra-abdominal pressure from 0 to 10 mmHg a progressivelylarger area of the peritoneum is exposed to CO2, resulting inan increase in the diffusion area. Increasing intraperitonealinsufflation pressures above 10 mmHg will not result in anyfurther increase in the diffusion area, causing a plateau inCO2 absorption. Nevertheless, a further increase in arterialCO2 pressure may occur as a result of increases in physi-ologic dead space ventilation. The greater absorptive capac-ity of the peritoneal membrane may also be reflected in a

Fig. 2. Gas exchange variablesduring extraperitoneal (–h–) andintraperitoneal (–s–) carbondioxide insufflation. Each data pointrepresents the mean ± SD (n 4 8).

913

greater systemic absorption of CO2 [6]. Our results indicatethat the extraperitoneal insufflation is associated with lowerarterial CO2 pressure values compared to intraperitonealinsufflation. A possible explanation might be that duringextraperitoneal insufflation a relatively small cavity wascreated, resulting in a small diffusion area for CO2.

A possible criticism of the study protocol is that the twodifferent experiments were not compared in the same ani-mal. Performing both intra- and extraperitoneal measure-ments in the same pig would only be feasible in the half ofthe animals where the intraperitoneal experiment was car-ried out first. The opposite order of experiments is, how-ever, impossible. The creation of an extraperitoneal areacauses the peritoneum to detach from the fascia of the ab-dominal muscles. This would have biased the intraperitone-al measurements. For this reason the randomized model wasselected in which to perform the comparative measure-ments.

In conclusion, our results indicate that extraperitonealinsufflation with CO2 for laparoscopy is associated withhemodynamic changes similar to those observed during in-traperitoneal insufflation. Intraperitoneal insufflation is as-sociated with more rapid changes in central venous fillingpressures. Gas-exchange variables indicate a stronger ten-dency toward the development of respiratory acidosis dur-ing conventional intraperitoneal insufflation.

The data from this study suggests that extraperitonealinsufflation might result in less cardiovascular impairmentthan intraperitoneal insufflation.

References

1. Bannenberg JJ, Hourlay P, Meijer DW, Vangertruyden G (1995) Ret-roperitoneal endoscopic lumbar sympathectomy: laboratory and clini-cal experience. Endosc Surg Allied Technol 3: 16–20

2. Bannenberg JJ, Meijer DW, Klopper PJ (1994) The prone position.Using gravity for a clear view. Surg Endosc 8: 1115–1116

3. Bannenberg JJ, Meijer DW, Klopper PJ (1995) Extraperitoneal lapa-roscopic paraaortic lymph node sampling in prone position: develop-ment of a technique. Laparoendosc Surg 5: 41–46

4. Brunt L, Molmenti E, Kerbl K, Soper N, Stone A, Clayman R (1993)Retroperitoneal endoscopic adrenalectomy: an experimental study.Surg Laparosc Endosc 3: 300–306

5. Clayman R (1993) Retroperitoneoscopy. In: Clayman R, McDougall E(eds) Laparoscopic urology. Quality Medical, St Louis, pp 383–394

6. Collins J (1981) Inert gas exchange of subcutaneous and intraperito-neal gaspockets in piglets. Respir Physiol 46: 391–404

7. Horattas M, Rosser R (1993) A new and simple approach to openlaparoscopy. Surg Gynecol Obst 176: 287–289

8. Joris J, Noirot D, Legrand M, Jacquet N, Lamy M (1993) Hemody-namic changes during laparoscopic cholecystectomy. Anesth Analg76: 1067–1071

9. Lister D, Rudston-Brown B, Wriner B, McEwin J, Chan M, Walley K(1994) Carbon dioxide absorption is not linearly related to intraperi-toneal carbon dioxide insufflation in pigs. Anesthesiology 80: 129–136

10. Mandressi A, Buizza C, Antonelli D, Belloni M, Chisensa S, Zaroli A,Bernasconi S (1993) Retro-extraperitoneal laparoscopic approach toexcise retroperitoneal organs: kidney and adrenal gland. MinimallyInvasive Ther 2: 213–220

11. Mullet C, Viale J, Sagnard P, Miellet C, Ruynat L, Counioux H, MotinJ, Boulez J, Dargent D, Annat G (1993) Pulmonary CO2 eliminationduring surgical procedures using intra- or extraperitoneal CO2 insuf-flation. Anesth Analg 76: 622–626

12. Raboy A, Hakim LS, Ferzli G, Antario JM, Albert PS (1993) Extra-peritoneal endoscopic vesicourethral suspension. Laparoendosc Surg3: 505–508

13. Westerband A, Water JM VD, Amzallag M, Lebowitz PW, Charda-voyne R, Abou-Taleb A, Wang X, Wise L (1992) Cardiovascularchanges during laparoscopic cholecystectomy. Surg Gynecol Obst175: 535–538

14. Wolf JJ, Carrier S, Stoller ML (1995) Intraperitoneal versus extraperi-toneal insufflation of carbon dioxide as for laparoscopy. J Endourol 9:63–66

15. Wright DM, Serpell MG, Baxter JN, O’Dwyer PJ (1995) Effects ofextraperitoneal carbon dioxide insufflation on intraoperative blood gasand hemodynamic changes. Surg Endosc 9: 1169–1172

914

News and notices

New Address for the European Association forEndoscopic Surgery (E.A.E.S.)

Effective January 1, 1997, the new correspondence, telephone, and faxnumbers of the E.A.E.S. office are:

E.A.E.S. Office, c/o Mrs. Ria Palmen

Luchthavenweg 81Unit 1.425657 EA EindhovenThe Netherlandsor: P.O. Box 3355500 AH VeldhovenThe NetherlandsTel: +31 40 2525288Fax: +31 40 2523102

Volunteer Surgeons NeededNorthwestern Nicaragua LaparoscopicSurgery Teaching Program,Leon, Nicaragua

Volunteer surgeons are needed to tutor laparoscopic cholecystectomy forthis non-profit collaboration between the Nicaraguan Ministry of Health,the National Autonomous University of Nicaragua, and MedicalTraining Worldwide. The program consists of tutoring general surgeonswho have already undergone a basic laparoscopic cholecystectomycourse. Medical Training Worldwide will provide donated equipmentand supplies when needed.

For further information, please contact:

Medical Training WorldwideRamon Berguer, MD, ChairmanTel: 707-423-5192Fax: 707-423-7578e-mail: berguer.r@martinez.va.gov

Fellowship in Minimally Invasive SurgeryGeorge Washington Medical CenterWashington, DC USA

A one-year fellowship is being offered at the George WashingtonUniversity Medical Center. Interested candidates will be exposed to abroad range of endosurgical Education and Research Center. Activeparticipation in clinical and basic science research projects is alsoencouraged.

For further information, please contact:

Carole Smith202-994-8425

or, send curriculum vitae to:

Dr. Jonathan M. SackierDirector of Endosurgical Education and ResearchGeorge Washington University Medical CenterDepartment of Surgery2150 Pennsylvania Avenue, N.W.6B-417Washington, DC 20037, USA

Fellowships in Minimally Invasive SurgeryThe University of Pittsburgh Medical CenterPittsburgh, PA, USA

One year fellowships in advanced minimally invasive surgery in bothgeneral and thoracic surgery are being offered at the University ofPittsburgh Medical Center beginning on July 1, 1997. Requirementsinclude completion of residency training programs in the desired area.The fellowships include a competitive salary and travel allowance.

Interested candidates should send a letter of inquiry with curriculum vi-tae to:

Philip R. Schauer, MD (General Surgery) orJames Luketich, MD (Thoracic Surgery)The University of Pittsburgh Medical Center3471 Fifth AvenueSuite 300Pittsburgh, PA 15213-3221

Fellowships in Laparoscopic SurgeryStaten Island University HospitalStaten Island, NY USA

A one year fellowship, to start July 1, 1997, in advanced laparoscopicsurgery is being offered at Staten Island University Hospital. Theselected fellow will be exposed to many advanced general laparoscopicsurgeries including: hiatal hernia repair, splenectomy, adrenalectomy,bowel resection, and others. Participation in research projects will beencouraged.

For further information, please contact:

Barbara ColemanCoordinator, Surgical residency programTel: 718-226-9508

Essentials of Laparoscopic SurgerySurgical Skills UnitUniversity of DundeeScotland, UK

Under the direction of Professor A. Cuschieri the Surgical Skills Unit isoffering a three-day practical course designed for surgeons who wish toundertake the procedures such as laparoscopic cholecystectomy. Thisintensely practical program develops the necessary operating skills,emphasizes safe practice, and highlights the common pitfalls anddifficulties encountered when starting out. Each workshop has amaximum of 18 participants who will learn both camera andinstrument-manipulation skills in a purpose-built skills laboratory.During the course there is a live demonstration of a laparoscopiccholecystectomy. The unit has a large library of operative videos editedby Professor Cuschieri, and the latest books on endoscopic surgery areon display in our Resource area. Course fee including lunch and coursematerials is $860.

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 971–973

For further details and a brochure please contact:

Julie Struthers, Unit Co-ordinatorSurgical Skills UnitNinewells Hospital and Medical SchoolDundee DD1 9SYTel: +44 382 645857Fax: +44 382 646042

Advanced Endoscopic SkillsSurgical Skills UnitUniversity of DundeeScotland, UK

Each month Professor Cuschieri Surgical Skills Unit offers a 41⁄2 daycourse in Advanced Endoscopic Skills. The course is intensely practicalwith ‘‘hands on’’ experience on a range of simulated models. The programis designed for experienced endoscopic surgeons and covers advanceddissection techniques, extracorporeal knotting techniques, needle control,suturing, internal tying technique, stapling, and anastomotic technique.Individual workstations and a maximum course number of 10 participantsallows for personal tuition. The unit offers an extensive collection of sur-gical videos and the latest books and publications on endoscopic surgery.In addition, participating surgeons will have the opportunity to see liveadvanced laparoscopic and/or thoracoscopic procedures conducted by Pro-fessor Cuschieri and his team. The course is endorsed by SAGES. Coursefee including lunch and course materials is $1850.

For further details and a brochure please contact:

Julie Struthers, Unit Co-ordinatorSurgical Skills UnitNinewells Hospital and Medical SchoolDundee DD1 9SYTel: +44 382 645857Fax: +44 382 646042

The Practical Aspects of Laparoscopic FundoplicationSurgical Skills UnitUniversity of DundeeScotland, UK

A three-day course, led by Professor Cuschieri, designed for experiencedlaparoscopists wishing to include fundoplication in their practice. Thecourse covers the technical details of total and partial fundoplication usingsmall group format and personal tuition on detailed simulated models.There will be an opportunity to observe one of these procedures live duringthe course. Maximum course number is six. Course fee including lunch is$1850.

For further details and a brochure please contact:

Julie Struthers, Unit Co-ordinatorSurgical Skills UnitNinewells Hospital and Medical SchoolDundee DD1 9SYTel: +44 382 645857Fax: +44 382 646042

Courses at George Washington University EndosurgicalEducational and Research Center

George Washington University Endosurgical Educational and ResearchCenter is proud to offer a wide range of surgical endoscopy courses. Thesecourses include advanced laparoscopic skills such as Nissen fundoplica-tion, colon resection, common bile duct exploration, suturing, as well assubspecialty courses. Individual surgeons needs can be met with privatetuition. The Washington D.C. area is a marvelous destination to visit for

recreational pursuits which can be arranged by the facility to suit yourpersonal agenda.

For further details please contact:

Carole Smith:Department of Surgery2150 Pennsylvania Avenue NW6BWashington, DC 20037, USATel: (202) 994-8425

Courses at the Royal Adelaide Centre forEndoscopic Surgery

Basic and Advanced Laparoscopic Skills Courses are conducted by theRoyal Adelaide Centre for Endoscopic Surgery on a regular basis. Thecourses are limited to six places to maximize skill development and tuition.Basic courses are conducted over two days for trainees and surgeons seek-ing an introduction to laparoscopic cholecystectomy. Animal viscera insimulators is used to develop practical skills. Advanced courses are con-ducted over four days for surgeons already experienced in laparoscopiccholecystectomy who wish to undertake more advanced procedures. Awide range of procedures are included, although practical sessions can betailored to one or two procedures at the participants request. Practical skillsare developed using training simulators and anaesthetised pigs.

Course fees: $A300 ($US225) for the basic course and $A1,600($US1,200) for the advanced course.

For further details and brochure, please contact:

Dr. D. I. Watson or Professor G. G. JamiesonThe Royal Adelaide Centre for Endoscopic SurgeryDepartment of SurgeryRoyal Adelaide HospitalAdelaide SA 5000 AustraliaTel: +61 8 224 5516Fax: +61 8 232 3471

Advanced Laparoscopic Suturing and SurgicalSkills Courses

MOET InstituteSan Francisco, CA, USA

Courses are offered year-round by individual arrangement. The MOETInstitute is accredited by the Accreditation Council for Continuing MedicalEducation (ACCME) to provide continuing medical education for physi-cians and designates these CME activities for 20–40 credit hours in Cat-egory 1 of the Physician’s Recognition Award of the American MedicalAssociation. These programs are also endorsed by the Society of Gastro-intestinal Endoscopic Surgeons (SAGES).

For further information, please contact:

Wanda Toy, Program AdministratorMicrosurgery & Operative Endoscopy Training (MOET) Institute153 States StreetSan Francisco, CA 94114, USATel: (415) 626-3400Fax: (415) 626-3444

Courses at WISEWashington Institute for Surgical EndoscopyWashington, DC, USA

The Washington Institute of Surgical Endoscopy is pleased to offer thefollowing courses:

972

Laparoscopic antireflux and hiatal hernia surgery (July 14–15, 1997); Lap-aroscopic management of the common bile duct and difficult cholecystec-tomy (May 15–16, August 11–12, November 10–11, 1997); Laparoscopiccolon and rectal surgery (June 20–21, September 15–16, December 4–5,1997). Also, courses for operating room nurses and technicians will be runon a monthly basis and personal instruction and preceptorship is available.

For further information, please call:

Carole SmithWashington Institute of Surgical Endoscopy2150 Pennsylvania Avenue, N.W.Washington, DC 20037Tel: 202-994-9425

Call for AbstractsSociety of American Gastrointestinal EndoscopicSurgeons (SAGES) 1998 Annual MeetingApril 1–4, 1998Seattle, WA, USA

Abstract deadlines: Oral and Poster abstracts: September 12, 1997Video Submissions: September 18, 1997

For further information, or to obtain an abstract form, please contact:

SAGES Program CommitteeSociety of American Gastrointestinal Endoscopic SurgeonsSuite #30002716 Ocean Park BoulevardLos Angeles, CA 90405Tel: (310) 314-2404Fax: (310) 314-2585e-mail: SAGESMail@AOL.com

European Course on Laparoscopic Surgery(English language) November 18–21, 1997Brussels, Belgium

Course director: G.B. Cadiere

For further information, please contact:

Administrative SecretariatConference Services s.a.Avenue de l’Observatoire, 3 bte 17B-1180 Bruxelles

Tel: (32 2) 375 16 48Fax: (32 2) 375 32 99

Courses Offered at the University of MinnesotaMinneapolis, Minnesota, USA

September 17, 1997: Fourth Annual Conference,Molecular Biology of Colorectal Cancer

September 17, 1997: Sixth Annual Conference,Endorectal Ultrasonography

September 18–20, 1997: Sixtieth Annual Conference,Principles of Colon and Rectal Surgery

For further information, please contact:

Continuing Medical EducationUniversity of Minnesota615 Washington Avenue SE, Suite 107Minneapolis, MN 55414Tel: 800-776-8636Fax: (612) 626-7766

9th International MeetingSociety for Minimally Invasive Therapy

July 14–16, 1997Kyoto, Japan

Scientific program to include: Plenary, Parallel, Poster, and Video sessions.Host Chairman: Professor Osamu Yoshida, Department of Urology, KyotoUniversity, 54 Shogoin Kawahara-sho, Sakyo, Kyoto 606, Japan. Phone: +8175 751-3328, Fax: +81 75 751-3740. This meeting coincides with the GionFestival in Kyoto, one of the greatest festivals in Japan.

For further information, please contact:

Secretariat of SMIT 9th Annual International Meetingc/o Academic Conference Planning383 Murakami-choFushimika, Kyoto 612 JapanTel: +81 75 611-2008Fax: +81 75 603-3816

Colorectal Disease in 1998February 19–21, 1998Fort Lauderdale, FL, USA

Symposium Director: Steven D. Wexner, MD

Cleveland Clinic Florida presents its ninth annual postgraduate course.Provides an intensive, in-depth, analytical review of all aspects of colo-rectal disease, including laparoscopy; colorectal carcinoma screening andgenetics, inflammatory bowel disease; and pouch surgery. There will be areview of both basic and advanced principles of diagnosis and managementof disease. Video techniques will be shown as well. The faculty is inter-nationally represented and includes leading experts in the field. Simulta-neous Spanish and Italian translation is available.

For more information, please contact:

Cleveland Clinic FloridaDepartment of Education2950 West Cypress Creek RoadFort Lauderdale, FL 33309-1743Tel: 800-359-6101, ext. 6066Fax: 954-978-5539

6th World Congress of Endoscopy Surgery ‘‘Roma 98’’6th International Congress of European Associationfor Endoscopic Surgery

June 3–6, 1998Rome, Italy

The program will include: the latest, original high quality research; sym-posia; plenary lectures; abstract presentations (video, oral, and posters);EAES and SAGES postgraduate courses, OMED postgraduate course ontherapeutic endoscopy; working team reports; educational center and learn-ing corner; meeting of the International Society of Nurses and Associates;original and non original scientific reports; and a world expo of newtechnology in surgery.

For further information, please contact:

Congress Secretariat: Studio EGAViale Tiziano, 1900196 Rome, ItalyTel: +39 6 322-1806Fax: +39 6 324-0143

973

Technique

A new fine probe for electric cautery during endoscopic surgery

Y. Yanagita,1,2 T. Asao,1 R. Katoh,1,2 M. Takahashi,1,2 Y. Nagamachi1

1 First Department of Surgery, Gunma University, 3-39-15 Showa-machi Maebashi 371, Japan2 Department of Surgery, Gunma Cancer Center, Ota 373, Japan

Abstract. The present paper introduces a new fine probefor electric cautery (1.65 mm in diameter, 22 cm long) thatcan be connected to a conventional cylindrical hand-controlled cautery holder, which is monopolar and widelyused in general surgery. When cautery was required, a 14-gauge intravenous catheter was inserted at an appropriatesite under the guidance of a videoscope. After removing thesteel inner needle, the extra tube was used as the fine sur-gical port for the cautery probe. The position of insertioncould be altered according to the operating field. Cauterywas performed by conventional methods. There was nobleeding or air leakage at the site of puncture during or aftersurgery. The puncture wound was closed without any su-tures. Based on these results, the new fine probe for cauterycan reduce the number of surgical ports required for instru-ments during video-assisted surgery, thus improving theease and safety of endoscopic surgery.

Key words: Endoscopic surgery — Cautery probe — Cau-tery holder

During endoscopic surgery, all surgical manipulation mustbe performed through a limited number of fixed surgicalports, which contributes to the difficulty associated withlaparo- or thoracoscopic surgery [1, 2]. Therefore, we de-veloped a new cautery probe that is sufficiently fine to allowit to enter through the extra tube of a 14-gauge intravenouscatheter. The use of this new probe was evaluated duringlaparo/thoracoscopic surgery.

Materials and methods

The cautery probes were 1.65 mm in diameter and 22 cm long. With theexception of the tip, the whole shaft was insulated with a contractive tubepurchased from Ohm Co. (Tokyo, Japan). Three different tip shapes were

developed (Spatula, L-hook, and Needle-shaped). The probes were con-nected to a conventional cylindrical hand-controlled cautery holder fromValleylab Inc. (Colorado, USA). When cautery was required, a 14-gaugeintravenous catheter was inserted at an appropriate site under videoscopicguidance. After removing the steel inner needle, the extra tube was used asthe surgical port for the fine cautery probe (Fig. 1).

Results and discussion

The probe was used during eight endoscopic surgical pro-cedures (two cholecystectomies, two colectomies, and fourCorrespondence to:T. Asao

Fig. 1. Surgical techniques using fine cautery probe.

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 961–962

thoracoscopic operations). No complications associatedwith the probe were observed in any of the patients. Thecautery could be manipulated as during conventional sur-gery. The position of insertion could be altered according tothe operating field. Insertion had to be repeated severaltimes in order to allow suitable manipulation during thora-coscopic surgery due to the narrow space between the ribs.No bleeding or air leakage was observed at the puncture siteduring or after surgery. The puncture wound was closedwithout any sutures and scar formation was not induced.Based on these results, the new fine probe for cautery can

reduce the number of surgical ports required for instrumentsduring video-assisted surgery, thus improving the ease andsafety of endoscopic surgery.

References

1. Rohlf S (1995) Electrosurgical safety considerations for minimally in-vasive surgery. Minim Invasive Surg Nurs 9: 26–29

2. Waxman K, Birkett DH, Sackier JM, Este-McDonald J, Duquette J(1994) Clinical and laboratory evaluation of an electrosurgical laparo-scopic trocar. Surg Endosc 8: 1076–1079

962

Sonographic blood flow measurements in malignant breast tumors

A potential new prognostic factor

C. Sohn, F. Beldermann, G. Bastert

Department for Prenatal and Gynecological Ultrasound Diagnosis and Treatment, Clinic of OB/GYN, University of Heidelberg, Heidelberg, Germany

Received: 16 August 1996/Accepted: 27 January 1997

AbstractBackground:The aim of this study was to find a possiblerelationship between the biological behavior of carcinomasof the breast and sonographically detectable blood flow af-ter first studies showed a correlation between blood flowand prognostic factors.Method: 259 patients with ductal invasive breast cancerwere examined using MEM (i.e., the Maximum EntropyMethod), a new sonographic blood flow measurement tech-nique capable of discerning considerably slower blood flowvelocities than Doppler sonography. Due to the lack of ob-jective methods for quantifying the blood flow, the findingswere divided into three classes dependent upon the visualcolor information obtained. The blood flow was correlatedwith the size of the tumor, lymph node and receptor status,ploidy and S-phase fraction.Results:Most of the patients with small tumors, withoutlymph node metastases, with positive receptors, with a dip-loid genome, and with a low S-phase fraction belonged tothe group with the lowest blood flow.Conclusion:The close relationship between the establishedprognostic factors and the sonographic blood flow measure-ments using MEM might be indicative of a new preopera-tive prognostic factor; this must, however, be confirmed bylarger studies.

Key words: Ultrasound blood flow — MEM technique —Prognostic factors

During sonographic determination of the blood flow withintumors of the breast we noticed that there are malignanttumors in which the carcinomas display a very rapid bloodflow (in the form of numerous colored areas) and carcino-mas with much less blood flow (in the form of merely

individual colored points) [15, 16]. This appeared to corre-late with the established prognostic factors.

This study was conducted in order to confirm this initialimpression; 259 patients with ductal invasive breast cancerwere examined sonographically on the preoperative day andthe sonographic data was compared with the postoperativefindings. The utilization of a new ultrasound technique, ca-pable of distinctly more sensitive blood flow diagnosis thanconventional Doppler sonography, appears to be decisivefor this study.

Materials and methods

Sonographic technique

Ultrasound system: CSA manufactured by Acoustic Imaging, Phoenix,USA

Color technique: MEM (Maximum Entropy Method)

The maximum entropy method is a nonlinear method of spectral analysisinitially developed by J.P. Burg for the US Naval Forces for the location ofboats. A modification of the algorithms used in the color Doppler results inan improved spectral resolution and noise perception. MEM forestalls thesignals from becoming imperceivable due to the noise by optimizing theseparation of the flow information and the disturbing noise. MEM is thuseven capable of detecting signals which are weaker than the backgroundnoise. The detection of extremely slow blood flow velocities, which areimperceivable to the Doppler technique, is thus possible.

Slowest discernible blood flow velocities (phantom measurements):

MEM: 0.2 mm/sColor Doppler: 10–20 mm/s

The MEM technique is thus capable of discerning the slowest of bloodflow velocities, as present in the minutest of blood vessels.

The following problem does, however, arise: To date the color infor-mation concerning the blood flow has been quantified using an integratedpulsed Doppler, which searched for the individual colored points. The thusobtained Doppler spectra were analyzed and the usual Doppler parameterscould be calculated. The MEM technique is, however, capable of detectingconsiderably slower blood flow velocities than the pulsed Doppler canregister; hence, this type of quantification is no longer possible. One thusmerely has the possibility of directly registering the color information andevaluating it.

This led to the following division of the tumor blood flow velocitiesinto three classes:

Correspondence to:C. Sohn, Universita¨ts-Frauenklinik, Voss Str. 9,D-69115 Heidelberg, Germany

Surg Endosc (1997) 11: 957–960

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Class I: Merely individual colored points within the tumor or on its edgeare discernible.

Class II: One colored area is discernible within the tumorClass III: Numerous colored areas are discernible within the tumor.

This is a subjective classification. Thus, we tested its reproducibility inadvance: Two physicians were presented with five examples, as colordepictions, from each class. These 15 examples were all correctly classifiedby both colleagues; hence, we could assume a good reproducibility [16].

Patients

Some 259 patients with ductal invasive breast cancer were included in thestudy. Only patients with ductal invasive carcinomas were included inorder to obtain a homogenous group as far as histology is concerned. Allpatients were examined on the preoperative day and their inclusion in thestudy was confirmed postoperatively by the histologic findings of an in-vasive ductal carcinoma. The size of the tumor, the lymph node and thereceptor status, the ploidy, and the S-phase fraction were registered to-gether with the histology of the tumor.

The examinations were all carried out by the same person.For statistical analysis the Kruskal-Wallis Test and Fisher’s Exact Test

were used.

Results

1. The classification of the 259 patients into the three bloodflow classes:Class I: 106 (40.9%)Class II: 60 (23.1%)Class III: 93 (36%)

2. The correlation of the T1 tumors with the blood flowclasses: There were 49 T1 tumors in class I (46.2% of thetumors in class I), 36 in class II (60% of the tumors inclass II), and 30 in class III (32.2% of the tumors in classIII) (Figs. 1–3).

3. The correlation of the lymph node status with the bloodflow classes: More than 14 axillary lymph nodes wereremoved from all patients and histologically examined;125 patients had no lymph node metastases; 134 patientshad metastases in a lymph node; 72 (58%) of the 125patients without lymph node metastases belonged toblood flow class I, 21 (17%) to class II, and 32 (25%) toclass III; 34 (25%) of the 134 patients with metastases ina lymph node were classified as class I, 39 (29%) as classII, and 61 (46%) as class III (p 4 0.022).

4. The correlation of the receptor status with the blood flowclasses: Information concerning the receptor status wasonly available for 203 patients examined by the samelaboratory. These patients were classified as having botha positive estrogen and progesterone receptor status, onlyone positive receptor status, or as both receptors havinga negative status.● Both receptors were found to be positive in 73 patients:

33 (47%) of these belonged to blood flow class I, 14(19%) to class II, and 25 (34%) to class III.

● In 44 patients only one receptor (estrogen or proges-terone receptor) was positive: 15 (34%) of these pa-tients belonged to class I, 10 (23%) to class II, and 19(43%) to class III.

● Both receptors were negative in 63 patients: 20 (32%)of them belonged to class I, six (9%) to class II, and 37(59%) to class III (p 4 0.034).

5. The correlation of the ploidy with the blood flow classes:The ploidy was only known in 200 patients examined bythe same laboratory; 79 patients had a diploid genome:38 (48%) of these belonged to class I, 19 (24%) to classII, and 22 (28%) to class III; 121 patients had an aneu-ploid genome in the tumor: 37 (31%) of them were inclass I, 33 (27%) in class II, and 51 (42%) in class III (p4 0.035).

6. The correlation of the S-phase fraction with the blood

Fig. 1. Blood flow within a carcinoma of the breast corresponding toclass I.

Fig. 2. Blood flow within a carcinoma of the breast corresponding toclass II.

Fig. 3. Blood flow within a carcinoma of the breast corresponding to classIII.

958

flow classes (n 4 243): 243 patients were divided intotwo groups with an S-phase fraction greater or smallerthan 5. The S-phase was less than 5 in 137 cases; 86(63%) of them belonged to class I, 24 (17%) to class II,and 27 (20%) to class III.

The S-phase fraction was greater than 5 in the remaining106 patients: 21 (20%) of them belonged to blood flow classI, 22 (21%) to class II, and 63 (59%) to class III (p 40.022).

Discussion

The results presented show that the intensity of the bloodflow within a carcinoma of the breast (determined using anew and extremely sensitive color method) clearly corre-lates with the established prognostic factors. This can beimpressively demonstrated by the correlation with thelymph node status: Almost 60% of all patients with a poorblood supply within the carcinoma of the breast were free ofaxillary metastases of the lymph nodes, whereas only 25%of the patients with a good blood supply within the tumorwere free of metastases in a lymph node. The correlation ofa low blood flow with the ploidy, S-phase fraction, andreceptor status—and thus a favorable prognosis—is alsoevident: a low S-phase fraction was mostly evident in theblood flow class I, whereas tumors with an aneuploid ge-nome mostly had a lot of color pixels or areas. The detectionof estrogen and progesterone receptors also correlated withless blood flow within the tumor, a lack of receptor corre-lating with a good blood supply [7, 15, 16]. These results arenot so clear as shown in our first studies on this field, butshow a correlation between color pixels and prognostic fac-tors in malignant breast tumors [15, 16].

The prerequisite for this diagnosis appears to be theutilization of a new sonographic color technique—the MEMcolor technique. Thanks to its modified color algorithms itis capable of detecting distinctly slower blood flow veloc-ities than conventional color Doppler sonography can. It isthus possible to demonstrate blood flow within notablysmaller vessels than was previously the case. The theoreticreason for the examination of such slow blood flow veloc-ities is based on the fact that the neovascularization of ma-lignant tumors leads to the development of the minutest ofblood vessels without a tunica. These minute vessels have avery low vascular resistance and thus an extremely slowblood flow velocity. The MEM technique appears to becapable of detecting such blood flow velocities [16].

The subjective classification into the three describedblood flow classes is undoubtedly problematic. Althoughwe could demonstrate that this type of classification is re-

producible, it is imperative that further objectivization beattained. To this end we are currently testing an automaticmathematic classification system, according to the classesdescribed above, which places the color pixels in relation-ship to the black and white surface [4].

The study presented here illustrates a new field of uti-lization of color-coded sonographic blood flow measure-ments. To date this type of diagnosis has been used withinthe framework of diagnosis of a tumor’s nature [1–3, 5, 6,8,–11, 13, 14, 16], which does, however, strike us as beingless promising than the utilization described here. B scansonography is capable of determining the nature of a tumorwith a relatively high scan accuracy (85% to 90%); bloodflow diagnosis can further increase this accuracy. An abso-lute accuracy in diagnosis can, however, not be attained;this is true for all scanning methods [1, 2, 6, 12].

Conclusion

It has to be critically evaluated whether or not the number ofoperations of benign findings in the breast can be reducedusing this new method. But as we mentioned above, abso-lute accuracy in diagnosis can and will not be attained in thenear future. Thus, it makes sense not only to use this suit-able highly sensitive ultrasound technique for diagnosis butalso to evaluate the prognosis of malignant neoplasms.

Our results present preliminary and still somewhat un-

Table 1. Lymph node statusa

LN negative: CL 1: 72 Pat 58%n 4 125 CL 2: 21 Pat 17%

CL 3: 32 Pat 25%LN positive: CL 1: 34 Pat 25%n 4 134 CL 2: 39 Pat 29%

CL 3: 61 Pat 46%

a The correlation of the lymph node status with the blood flow classes:n 4259.

Table 2. Receptor statusa

ER and PR positive: CL 1: 34Pat (47%)n 4 73 CL 2: 14Pat (19%)

CL 3: 25Pat (34%)ER or PR positive: CL 1: 15Pat (34%)n 4 44 CL 2: 10Pat (23%)

CL 3: 19Pat (43%)ER and PR negative: CL 1: 20Pat (32%)n 4 63 CL 2: 6Pat (9%)

CL 3: 37Pat (59%)

a The correlation of the receptor status with the blood flow classes:n 4180.

Table 3. Ploidya

Aneuploid genome: CL 1: 37 Pat 31%n 4 121 CL 2: 33 Pat 27%

CL 3: 51 Pat 42%Diploid genome: CL 1: 38 Pat 48%n 4 79 CL 2: 19 Pat 24%

CL 3: 22 Pat 28%

a The correlation of the ploidy with the blood flow classes:n 4 200.

Table 4. S-Phasea

S-phase < 5%: CL 1: 86 Pat 63%n 4 137 CL 2: 24 Pat 17%

CL 3: 27 Pat 20%S-phase > 5%: CL 1: 21 Pat 20%n 4 106 CL 2: 22 Pat 21%

CL 3: 63 Pat 59%

a The correlation of the S-phase with the blood flow classes:n 4 243.

959

certain indications concerning a new prognostic method forcarcinomas of the breast [16]. Long-term studies on largegroups of patients must confirm these preliminary results.

This examination opens an entirely new field in theprognostic evaluation of malignant breast cancer—whenpreoperative assessment of the prognosis thus appears to bepossible for the first time.

Should our first results be confirmed the following ques-tion arises: Could this new preoperative prognosis assess-ment perhaps even influence the surgical procedure?

References

1. Bamber JC, Sambrook M, Minasian H, Hill CR (1983) Doppler studyof blood flow in breast cancer. In: Jellins J, Kobayashi T (eds) Ultra-sonic examination of the breast. John Wiley, pp 371–378

2. Blohmer J (1991) Mammasonographie. In: Sohn Ch, Holzgreve (eds).Ultraschall in Gyna¨kologie und Geburtshilfe. Chapman & Hall, Buch,Druck

3. Burns PN, Virjee JM, Gowland M, et al (1983) The origin of dopplershift signals from breast tumors. In: Jellins J, Kobayashi T (eds) Ul-trasonics examinations of the Breast. John Wiley, pp 379–384

4. Delorme S, Anton H-W, Knopp MV, Betsch B, Trost U, JunkermannI, Fournier DV, Van Kaick G (1991) Vaskularisation des Mam-makarzinoms: Quantitative und morphologische Beurteilung mittelsfarbcodierter Dopplersonographie. Abstract 423 Ultraschall Klin Prax6: S.219

5. Heilenkötter U, Jagella P (1993) Farbdopplersonographie exstirpa-tionsbedu¨rftiger Mammatumoren—Darstellung einer Untersuchungs-methode. Geburtshilfe Frauenheilkd 53: 247–252

6. Jellins J (1988) Combining imaging and vascularity assessment ofbreast lasions. Ultrasound Med Biol 14: 121–130

7. Kaufmann M, Minckwitz GV, Finn HP, Schmid H, Goerttler K,Bastert G ( ) Combination of grading and new biological factors(S-phase fraction and epidermal growth factor receptor) can predictrelapse and survival in patients with node-negative primary breastcancer. Onkologie 17: 166–172

8. Madjar H, Sauerbrei W, Münch S, Prömpeler H, Schillinger H (1990)Methodenanalyse zur Doppleruntersuchung der weiblichen Brust. Ul-traschall Med 4: 196–201

9. Madjar H, Jellins J, Schillinger H, Hillemanns HG (1986) Differen-zierung von Mammakarzinomen durch CW-Doppler-Ultraschall. Ul-traschall Med 7: 183–184

10. Madjar H, Pro¨mpeler H, Wilhelm CH (1991) Doppler zur Diagnostikund Therapie von Brusterkrankungen. Abstract 425 Ultraschall KlinPrax 6: S.220

11. Minasian H, Bamber JC (1982) A preliminary assessment of an ultra-sonic doppler method for the study of blood flow in human breastcancer. Ultrasound Med Biol 8: 357–364

12. Sohn CH, Stolz W (1991) Dopplersonographische Durchblutungsmes-sung von Brusttumoren. Abstract 424 Ultraschall Klin Prax 6: S.219

13. Sohn CH, Grischke EM, Wallwiener D, Kaufmann M, Fournier DV,Bastert G (1992) Die sonographische Durchblutungsdiagnostik gut-und bosartiger Brusttumoren. Gerbutshilfe Frauenheilkd 52: 397–403

14. Sohn CH, Stolz W, Grischke EM, Wallwiener D, Bastert G, FournierDV (1992) Die dopplersonographische Untersuchung von Mammatu-moren mit Hilfe der Farbdopplersonographie, der Duplexsonographieund des CW-Dopplers. Zentralbl Gyna¨kol 114: 249–253

15. Sohn CH, Grischke EM, Stolz W, Bastert G (1993) Untersuchungenzum Zusammenhang zwischen dem Grad der Durchblutung und dembiologischen Verhalten von Mammatumoren. Ultraschall Klin Prax 8:11–14

16. Sohn CH, Thiel C, Baudendistel A, Kaufmann M, Bastert G (1996) Isthe sonographic degree of blood supply a new prognostic factor?Breast 5: 110–112

960

Hormone-cytokine response

Pneumoperitoneum vs abdominal wall-lifting in laparoscopic cholecystectomy

T. Yoshida,1 E. Kobayashi,1,2 Y. Suminaga,1 H. Yamauchi,1 T. Kai,1 N. Toyama,1 H. Kiyozaki,1 A. Fujimura, 2

M. Miyata 1

1 Department of Surgery, Omiya Medical Center, Jichi Medical School, 3311-1 Kawati-gun, Tochigi 329-04, Japan2 Department of Clinical Pharmacology, Jichi Medical School, 3311-1 Kawati-gun, Tochigi 329-04, Japan

Received: 10 November 1996/Accepted: 19 February 1997

AbstractBackground:Changes in blood hormone and cytokine wereinvestigated in patients who underwent laparoscopic chole-cystectomy via insufflation (CO2 group) vs those who hadabdominal wall-lifting (Air group).Methods:Seventeen female patients with cholecystolithiasiswere randomly divided into two groups. Peripheral bloodsamples were obtained during perioperative period, andplasma hormone levels (ACTH, cortisol) and serum cyto-kine levels (TNFa, IL-1b, IL-6, IL-10) were measured.Results:The number of circulating lymphocytes signifi-cantly decreased at 1 h after surgery in both groups, but thedecrease in the CO2 group was significantly smaller thanthat in the Air group. There was no significant difference inhormone elevation between groups. Serum concentrationsof IL-6 and IL-10 in the Air group were significantly higherthan in the CO2 group.Conclusions:CO2 insufflation may reduce cytokine produc-tion in laparoscopic cholecystectomy.

Key words: Laparoscopic cholocystectomy — Pneumo-peritoneum — Wall-lifting — Hormone — Cytokine

In recent years, laparoscopic cholecystectomy (LC) has be-come accepted world-wide because it is believed to be lesstraumatic than open cholecystectomy. A number of reportshave shown that LC reduces postoperative pain and lengthof hospital stay and minimally impairs respiratory function,muscle performance, and immune function [7, 16, 19, 21].There is a possibility that protection of the peritoneal cavityfrom exposure to air might be important to limit surgicaldamage due to the reduced production of cytokines aftersurgery [13, 22]. In this study, we compared hormonal and

cytokine production in patients who underwent laparoscopiccholecystectomy using the pneumoperitoneum method andabdominal wall-lifting method. The former procedure wasperformed by insufflation with carbon dioxide to protect theperitoneum from exposure to air, while in the latter, air wasfreely circulated. The results indicate that the pneumoperi-toneum method produced less cytokines, while no signifi-cant difference in physiological conditions was observedbetween the two groups. The mechanisms of minimizationof surgical damage by laparoscopic surgery are discussed.

Materials and methods

Patients

Seventeen female patients (35 to 65 years old; 49.0[2.25], mean[SEM])with cholecystolithiasis without choledocholithiasis were selected. Theywere diagnosed by ultrasonographic examination and drip infusion chol-angiography. Endoscopic retrograde cholangiography was also done in thecases where common bile duct was not clear from former examinations.Patients with liver dysfunction, acute inflammation, metabolic diseases,and cardiopulmonary diseases were excluded from this study. The subjectswere randomly selected to undergo laparoscopic cholecystectomy using thecarbon dioxide insufflation method (CO2 group) or the abdominal wall-lifting method [17] (Air group). The patients gave their informed consentafter the protocol was approved by the appropriated Review Board. Selec-tion took place on the day before the surgery, using numbered and sealedenvelopes.

Operation

All patients were anesthetized according to the same protocol; after fastingfor 8 h, general anesthesia was achieved using thiopental sodium, vecuro-nium bromide, and sevoflurane. After the operation, pentazocine was givenfor pain relief if necessary. No blood products were required. Fluid re-placement during and 24 h after surgery was similar in each group.

In both groups, laparoscopic cholecystectomy was performed using10-mm and 5-mm ports, as performed routinely [4]. Intraoperative chol-angiography was performed on all patients. In the CO2 group, the intra-abdominal pressure was kept under 8 mmHg during the operation as usu-ally performed. In the Air group, a port for the laparoscope was initiallyinserted into the abdominal cavity through open minimal laparotomy; twoKirschner wires were placed through the skin, one in the upper portion ofCorrespondence to:E. Kobayashi

Surg Endosc (1997) 11: 907–910

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

the umbilicus and the other in the midclavicular line at the right costalmargin. The ends of those wires were pulled by means of a lifting handleand a lifting bar to provide a good view of the surgical side [17]. Onesurgeon (Y. S.), who is proficient in the technique for both methods, per-formed all procedures.

Blood samples and analyses

Peripheral blood samples were drawn at 5 points: at 6A.M. on the day ofthe operation (Preop.), during the operation (Intraop.), 1 h after the opera-tion (Postop.), and at 6A.M. on the 1st and 3rd postoperative days (1POD,3POD). Each sample was divided into three tubes. One of the bloodsamples in the EDTA tube was used for counting peripheral white bloodcells (WBC) and lymphocytes using an auto-analysis system. The plasmaobtained from the blood was stored at −20°C until the measurement ofadrenocorticotropic hormone (ACTH) and cortisol. Serum samples wereused for measuring C-reactive protein (CRP), interleukin-1b (IL-1b), in-terleukin-6 (IL-6), interleukin-10 (IL-10), and tumor necrosis factora(TNFa). Cortisol and ACTH levels were determined using immunoradio-metric assay (IRMA) [1, 2]. CRP was measured by turbidimetric immu-noassay [5]. Concentrations of serum IL-6, IL-10, and TNFa were deter-mined using an enzyme-linked immunosorbent assay [8]. IL-1b was alsodetermined using IRMA.

Statistical analysis

Data are expressed as mean (standard error of the mean [SEM]) for eachstudy group. Statistical analysis was done by the Mann-Whitney’s U test,and the significant difference was set at less than 5%.

Results

The clinical parameters of the patients are summarized inTable 1. There was no significant difference in those valuesbetween the CO2 and Air groups.

The changes in the circulating lymphocyte count afterthe operation are shown as percentages: [number of periph-eral lymphocytes in postoperative course/number of lym-phocytes in preoperative condition] ×100 (%) and are sum-marized in Fig. 1. The lymphocyte counts declined tran-siently in both groups. However, the rate of decrease in theAir group was significantly large at Postop. and 1POD (CO2vs Air group: 82.1% [3.6] vs 69.2 [4.03]p 4 0.027, and100.1 [3.9] vs 82.9 [2.72]p 4 0.008 at Postop. and 1POD,respectively; mean [SEM]).

The changes in the level of serum CRP are shown in Fig.2. The serum CRP level in the Air group was significantly

higher than in the CO2 group on 1POD and 3POD (0.94mg/dl [0.26] vs 3.92 [1.12]p 4 0.027, and 1.7 [0.43] vs 6.8[1.44] p 4 0.012, at 1POD and 3POD, respectively).

The changes in the level of plasma hormones after sur-gery are shown in Fig. 3. Both ACTH and cortisol increasedduring and immediately after surgery, then decreased to thepreoperative level on 1POD. There was no significant dif-ference in the levels of these two hormones between theCO2 and the Air groups.

No definite changes in serum IL-1b and TNFa concen-tration were detected in either group. Serum IL-6 and IL-10were detected in the peripheral blood and the changes inserum concentrations of these cytokines are shown in Fig. 4.The concentration of IL-6 in the Air group significantlyincreased postoperatively, in comparison with the CO2group (6.64 pg/ml [1.38] vs 22.1 [7.38]p 4 0.0039 and10.4 [2.75] vs 71.4 [39.33]p 4 0.0021, at Postop. and

Fig. 1. Change of peripheral lymphocyte number in CO2 vs Air groupsover time. The decrease of lymphocyte number in the Air group wassignificantly larger at 1 h and 1 day after operation than in the CO2 group.j CO2 group;h Air group; *p < 0.05, **p < 0.01 (Mann-Whitney U test).†: Lymphocyte number of preoperative sample was equaled to 100%.

Fig. 2. The changes in serum CRP after laparoscopic cholecystectomy.The serum CRP level in the air group was significantly higher at 1 and3POD than those in the CO2 group.s: CO2 group,j: Air group; *p <0.05(Mann-Whitney U test). †: The minimal detectable value of CRP was0.3 mg/dl.

Table 1. Patient characteristics and clinical parametersa

Mean [SEM]

CO2 group(n 4 9)

Air group(n 4 8) p 4b

Age (years) 47.7 [3.7] 50.5 [2.7] 0.700Height (cm) 158.0 [1.0] 155.3 [1.6] 0.498Weight (kg) 54.7 [2.6] 61.6 [3.4] 0.149Operation time (min) 114.4 [9.3] 121.3 [8.8] 0.562Anesthesia time (min) 198.9 [10.7] 202.5 [5.3] 0.923Pentazocine request (times) 1.1 [0.4] 1.4 [0.5] 0.689Time to full diet (days) 1.7 [0.1] 1.8 [0.2] 0.804Hospital stay (days) 6.0 [0.4] 6.0 [0.1] 0.920

a There was no significant difference in all characters and parametersb Mann-Whitney U test

908

1POD, respectively). Serum IL-10 concentration in the Airgroup was also significantly higher than in the CO2 groupon 1POD (2.23 pg/ml [0.86] vs 6.27 [1.44]p 4 0.031).

Discussion

LC has rapidly become the standard operation for cholecys-tolithiasis around the world. Recent reports have demon-strated that LC is less traumatic than open cholecystectomyfrom the standpoint of physiological, immunological, andinflammatory response [9–11, 16, 19]. It has been suggestedthat the reduction of surgical damage might be due to thesmaller incision required [9, 10]. There is also a possibilitythat protecting the peritoneal cavity from exposure to airmight minimize the inflammatory response [13]. We com-pared two methods for performing LC: the pneumoperito-neum method, which utilizes insufflation with carbon diox-ide, and the abdominal wall-lifting procedure, in which aircirculates from the ports. The former has been considered to

provide a good view of the gallbladder and the latter has themerit of causing smaller changes in cardiopulmonary func-tion during the operation [15]. No significant difference wasnoted postoperatively between the two groups in physi-ological factors such as postoperative pain and length ofhospital stay. However, the changes in the number of cir-culating lymphocytes, CRP, IL-6, and IL-10 were signifi-cantly larger in the Air group than in the CO2 group.

Surgical damage impairs both quality and quantity ofperipheral lymphocytes, and this impairment causes immu-nosuppression after surgery [14, 18]. It has been reportedthat surgical treatment triggers the release of inflammatorycytokines from peritoneal macrophages and the synthesis ofacute-phase proteins, and that these cytokines might be in-dicative of the degree of surgical damage [6]. Recent ex-perimental studies have shown that air contamination of theperitoneal cavity regulates early inflammatory responsesand induces immunosuppression and tumor growth [3, 20].In another paper, we compared the changes in circulatinglymphocytes, hormones, and cytokines that occur in human

Fig. 3. Hormonal changes after laparoscopic cholecystectomy. Changes of serum ACTH and cortisol showed a similar pattern in both groups.s: CO2

group,j: Air group. †: The minimal detectable value of ACTH was 5 pg/ml.

Fig. 4. The changes of serum IL-6 and IL-10 after laparoscopic cholecystectomy.s: CO2 group,j: Air group; *p < 0.05, **p < 0.01 (Mann-WhitneyU test). The concentrations of IL-6 and IL-10 in the Air group showed higher level groups showed higher level at 1 day after surgery than those in the CO2

group. †: The minimal detectable value of IL-6 and IL-10 were 0.4 pg/ml and 0.5 pg/ml, respectively.

909

subjects following major and minor surgery and showedthat the rate of decrease in the number of peripheral lym-phocytes and the increase of IL-6 and IL-10 were well cor-related with the degree of surgical damage [12]. In thispaper, we demonstrated that the abdominal wall-liftingmethod causes a more productive response of those cyto-kines and greater changes in circulating lymphocytes thanthe pneumoperitoneum method. The present study is thefirst clinical observation to compare the hormone and cy-tokine response to air exposure vs CO2 insufflation of theperitoneal cavity.

Acknowledgments.We thank Prof. I. Sakurabayashi (Department of Clini-cal Laboratory, Jichi Medical School, Omiya Medical Center) for technicalassistance and advice. This work was partially supported by a Grant ofPublic Trust Fund for the Promotion of Surgery.

References

1. Addison GM, Hales CN (1971) Two-site assay of human growth hor-mone. Horm Metab Res 3: 59–60

2. Addison GM, Hales CN, Woodhead JS, et al. (1971) Immunoradio-metric assay of parathyroid hormone. J Endocrinol 49: 521–530

3. Allendorf JDF, Bessler M, Kayton ML, et al. (1995) Increased tumorestablishment and growth after laparotomy vs laparoscopy in a murinemodel. Arch Surg 130: 649–653

4. Bailey RW, Zucker KA, Flowers JL, et al. (1991) Laparoscopic cho-lecystectomy. Ann Surg 214: 531–540

5. Bergstrom K, Lefvert AK (1980) An automated turbidimetric immu-noassay for plasma proteins. Scand J Clin Lab Invest 40: 637–640

6. Cruickshank AM, Fraser WD, Burns HJG, et al. (1990) Response ofserum interleukin-6 in patients undergoing elective surgery of varyingseverity. Clin Sci 79: 161–165

7. Delaunay L, Bonnet F, Cherqui D, et al. (1995) Laparoscopic chole-cystectomy minimally impairs postoperative cardiorespiratory andmuscle performance. Br J Surg 82: 373–376

8. Engvall E, Perlmann P (1972) Enzyme-linked immunosorbent assay,ELISA.III. Quantitation of specific antibodies by enzyme-labeled anti-immunoglobulin in antigen-coated tubes. J Immunol 109: 129–135

9. Glaser F, Sannwald GA, Buhr HJ, et al. (1995) General stress responseto conventional and laparoscopic cholecystectomy. Ann Surg 221:372–380

10. Griffith JP, Everitt NJ, Lancaster F, et al. (1995) Influence of laparo-scopic and conventional cholecystectomy upon cell-mediated immu-nity. Br J Surg 82: 677–680

11. Kloosterman T, von Blomberg ME, Borgstein P, et al. (1994) Unim-paired immune functions after laparoscopic cholecystectomy. Surgery115: 424–428

12. Kobayashi E, Yamauchi H (1997) Interleukin-6 and a delay of neu-trophil apoptosis after major surgery. Arch Surg 132 (in press)

13. Kobayashi E, Yoshida T, Yamauchi H, et al. (1995) Immune functionin the patients undergoing open vs laparoscopic cholecystectomy.Arch Surg 130: 676

14. Lennard TWJ, Shenton BK, Borzotta A, et al. (1985) The influence ofsurgical operations on components of the human immune system. Br JSurg 72: 771–776

15. McDermott JP, Regan MC, Page R, et al. (1995) Cardiorespiratoryeffects of laparoscopy with and without gas insufflation. Arch Surg130: 984–988

16. McMahon AJ, Russell IT, Baxter JN, et al. (1994) Laparoscopic versusminilaparotomy cholecystectomy: a randomised trial. Lancet 343:134–138

17. Nagai H, Kondo Y, Yasuda T, et al. (1993) An abdominal wall-liftmethod of laparoscopic cholecystectomy without peritoneal insuffla-tion. Surg Laparosc Endosc 3: 175–179

18. Park KS, Brody JI, Wallace HA, et al. (1971) Immunosuppressiveeffect of surgery. Lancet :53–55

19. Redmond HP, Watson WG, Houghton T, et al. (1994) Immune func-tion in patients undergoing open vs laparoscopic cholecystectomy.Arch Surg 129: 1240–1246

20. Watson RWG, Redmond HP, McCarthy J, et al. (1995) Exposure ofthe peritoneal cavity to air regulates early inflammatory responses tosurgery in a murine model. Br J Surg 82: 1060–1065

21. Yoshida T, Kobayashi E, Yamauchi H, et al. (1995) Laparoscopiccholecystectomy minimally impairs postoperative cardiorespiratoryand muscle performance. Br J Surg 82: 996–997

22. Yoshida T, Kobayashi E, Miyata M, et al. (1996) Increased tumorestablishment and growth after laparotomy vs laparoscopy in a murinemodel. Arch Surg 131: 219

910

Hand-assisted laparoscopic splenectomy for hydatid cyst

K. E. W. Ballaux, J. M. Himpens, G. Leman, M. R. P. Van den Bossche

Department of Abdominal and Thoracic Surgery, Sint-Blasius Ziekenhuis, Kroonveldlaan 50, 9200 Dendermonde, Belgium

Received: 27 September 1996/Accepted: 19 November 1996

Abstract. Splenic hydatidosis is a rare condition. We per-formed a hand-assisted laparoscopic splenectomy for a largehydatid cyst localized in the center of the spleen. We discussthe advantages of the ‘‘helping hand.’’

Key words: Hydatidosis — Laparoscopy — Spleen —Splenectomy

Splenic hydatidosis accounts for 0.5–8% of total hydatido-sis, with liver and pulmonary foci accounting for 60–70%and 30–35%, respectively [1, 8, 9].

Case report

A 26-year-old Turkish man was admitted to our Emergency Departmentwith diffuse urticarial wheals the day after taking a copious, spicy meal.Upon admission the patient was nauseous and had a temperature of 38.7°C.Biochemistry was normal except for eosinophilia of 9.9% (760/mm3). X-ray of the abdomen demonstrated two shell-shaped calcifications in the leftsubcostal region. Ultrasonography of the abdomen revealed a 5-cm sep-tated cystic mass in the spleen. A CAT scan showed an enlarged spleencontaining a multilocular and partially calcified unruptured cyst. Volumeof the spleen was estimated at 1,000 cc. Echinococcal radioallergosorbenttest (RAST) was positive.

A hand-assisted laparoscopic splenectomy was performed. Four trocarportals were made (Fig. 1). An 8-cm incision was made in the left lowerquadrant for application of an airtight system allowing introduction ofone’s hand and forearm (Dexterity Glove and Pneumosleeve; DexterityInc., Carpinteria, CA, USA.). This system allowed us to manipulate thespleen and reflect the colon and small intestine in order to expose thesplenic hilum.

The spleen was mobilized gradually: Gastrocolic and colosplenic liga-ments were divided and the hilum was dissected. The splenic artery wasligated first and the whole spleen turned blue. The splenic vein was clippedand divided. Finally the splenodiaphragmatic attachments were released.The spleen was readily removed intact through the minilaparotomy.

The procedure lasted 200 min. Blood loss was estimated at 100 cc. Thepostoperative course was uneventful; the patient left our department on the4th postoperative day.

Discussion

Although the procedure is difficult, advantages of laparo-scopic splenectomy are obvious: less pain, fast recovery ofbowel function, short hospital stay, and improved cosmeticoutcome [4, 5]. The Dexterity Glove and Pneumosleeve(Dexterity Inc.) is a recent adjunct in laparoscopic surgerythat allows hand manipulation with easier exposure of thestructures and better control of minor or major bleedings[3]. In laparoscopic splenectomy, seizing and manipulatingthe spleen is a major problem, especially in the case ofsplenomegaly [6, 7]. The Dexterity Glove and Pneu-mosleeve (Dexterity Inc.) proved to be a valuable tool indealing with this problem [3].

The 8-cm minilaparotomy used for the helping hand wasnecessary to remove the spleen intact since morcellationcould have spread the echinococcus scolices into the ab-dominal cavity. Introduction of the enlarged spleen into abag may prove extremely difficult and can be avoided bythis technique [2].

Correspondence to:J. M. Himpens Fig. 1. Sites of trocar portals and minilaparotomy.

Surg Endosc (1997) 11: 942–943

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Recently some authors have advocated conservativespleen surgery for echinococcus cysts, i.e., cyst enucleationand marsupialization [9]. This possibility was excluded in

this case because the cyst was large compared to the splenicvolume and was localized in the center of the spleen [8](Fig. 2).

References

1. Abi F, El Fares F, Khaiz D, Bouzidi A (1989) Les localisations inha-bituelles du kyste hydatique. A propos de 40 cas. J Chir 126: 307–312

2. Dexter SPL, Martin IG, Alao D, Norfolk DR, McMahon MJ (1996)Laparoscopic splenectomy. The suspended pedicle technique. Surg En-dosc 10: 393–396

3. Gossot D, Meijer D, Bannenberg J, De Wit L, Jakimovicz J (1995) Lasplenectomie laparoscopique revisite´e. Ann Chir 49: 487–489

4. Grossbard ML (1996) Is laparoscopic splenectomy appropriate for themanagement of hematologic diseases? Editorial Surg Endosc 10: 387–388

5. Phillips EH, Carroll BJ, Fallas MJ (1994) Laparoscopic splenectomy.Surg Endosc 8: 931–933

6. Terrosu G, Donini A, Silvestri F, Petri R, Anania G, Barillari G, Bac-carani U, Risaliti A, Bresadola F (1996) Laparoscopic splenectomy inthe management of hematological diseases. Surgical technique and out-come of 17 patients. Surg Endosc 10: 441–444

7. Trias M, Targarona EM, Balague´ C (1996) Laparoscopic splenectomy:an evolving technique. Surg Endosc 10: 389–392

8. Uriarte C, Pomares N, Martin M, Conde A, Alonso N, Bueno M (1991)Splenic hydatidosis. Am J Trop Med Hyg 44: 420–423

9. Vara-Thorbeck R, Rosell J, Ruiz-Morales M (1991) Milzechinokok-kose. Ihre konservative chirurgische Behandelung. Zentralbl Chir 116:1411–1416

Fig. 2. Longitudinal section of the spleen with large hydatid cyst in itscenter.

943

New technology

Technology and principles of tomographic image-guided interventionsand surgery

A. Melzer, A. Schmidt, K. Kipfmu ller, D. Gronemeyer, R. Seibel

MRI, Institute for Diagnostic and Interventional Radiology, Medical Computer Science University Witten/Herdecke, Schulstrasse 10, 45468Mulheim/Ruhr, Germany

Received: 17 October 1996/Accepted: 21 October 1996

Abstract. Image guidance of instruments can be performedby fluoroscopy, ultrasound imaging, computed tomography(CT), magnetic resonance imaging (MRI), and by stereo-tactic navigation. During the last year, there has been sig-nificant progress in MRI instrument guidance in the field ofinterventional radiology. Our first 168 clinical cases ofMRI-guided interventional procedures, e.g., aspiration bi-opsy of neoplasms and tomographic microtherapy with lo-cal interstitial chemo-ablation, confirm the feasibility ofMRI guidance. An expansion of MRI guidance during sur-gical endoscopic procedures is currently under evaluationand the initial results of this development are presented.Tomographic-guided surgery and the implementation ofMRI or CT scanning within the environment of an operatingroom (OR) entail specific technological requirements andOR design considerations. Fast sequences, interventionalprotocols, in-room monitor, as well as MR-compatibleprobes, cannulae, catheters, instruments, endoscopes, andauxiliary equipment that are necessary for this combinedsurgical image-guided approach are described.

Key words: Interventional MRI guidance — Tomographicmicrotherapy — Endoscopic surgery — Image-guided sur-gery

In open surgery the experienced operator is able to interpretthe patient’s pathological anatomy below the surface viewwithin the operative field. By virtue of the tactile and forcefeedback, he is able to identify and dissect the correct tissueplanes and avoid damage to vital structures. However, thesurgical open access is invasive and a magnified well-illuminated view of the operative field is difficult.

Endoscopic surgery provides a detailed, close-up, mag-nified view of the operative field. However, the remote

nature of the access leads to manipulative restrictions [3].Precise remote dissection has to be performed by the sur-geon in the presence of decoupling of the visual from themotor axes, reduced tactile and force feedback, and kine-matic restriction imposed by the reduced degrees of free-dom of the instruments. The evaluation of local anatomyand pathology is entirely dependent on visual inspection ofthe image on the video monitor screen. The image is gen-erated only from the light reflected from the surfaces oforgans and tissue.

The surfaces of the operative field in conjunction withunderlying supporting tissues and organs present thevolumeof interest.Tomographic visualization of the completeop-erative tissue volumecan be achieved if telescopic inspec-tion is combined with an additional intraoperative tomo-graphic imaging system. Two main issues then arise. Thefirst concerns the type of imaging technology which bestsuits the operative task. The second relates to the need forand benefit to be derived from a stereotactic navigationsystem that would virtually generate the instrument withinperoperatively gained tomographic data.

CT and MRI carry the greatest potential for the provi-sion of high-quality on-line and real-time intraoperativeto-mographic visualization of the operative tissue volume.This new approach can be defined astomographic therapy,or surgical tomography,or tomographic surgery.In the fol-lowing account, current imaging systems, techniques, andtechnologies forsurgical tomographyare described withspecial emphasis on MRI-guided interventions.

Visualization of the operative field

Modern video-endoscopes and laparoscopes provide an ex-cellent image of the surfaces of the operative field in realtime. However, precise and targeted intraoperative controlof the surgical process and navigation of the tools also re-quire ‘‘on-line’’ visualization of the tools, anatomicalplanes, pathology, and physical-chemical changes withoutany adverse effects. Four of the current diagnostic imagingCorrespondence to:A. Melzer

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 946–956

techniques—fluoroscopy, ultrasound (US), computed to-mography (CT), and magnetic resonance imaging (MRI)—meet these requirements to a varying extent and, further-more, can be deployed during surgery. With the exceptionof fluoroscopy, these imaging techniques have, until re-cently, been used primarily for diagnostic imaging. Tech-nical innovations, particularly in the field of MRI, haveenabled these imaging modalities to be used to image-guideradiological and surgical interventions. In endoscopic sur-gery the provision of a colored surface view by the tele-scope combined with near-real-time tomographic informa-tion on the tissue composing the operative volume can leadto a significant increase in the precision of the therapeuticeffect, together with an enhancement of the safety margin,particularly in anatomically crowded and dangerous areas.

The appropriate intraoperative application of fluoros-copy, US, CT, and MRI to the respective surgical procedurerequires analysis and evaluation of each imaging techniqueas well as further dedicated development.

Fluoroscopy

The fluoroscopic image represents a summation of the x-rayabsorption by all the structures that fall within the radiationbeam. Display of soft tissue and vascular structures andhollow visceral organs is inadequate and requires the use ofcontrast media. Intraoperative fluoroscopy is used routinelyin trauma surgery for the reduction of fractures. Intraopera-tive cholangiography identifies anatomical variations andpathology of the biliary system (Fig. 1). Intravascular inter-ventions are routinely performed with flouroscopic guid-ance. The major advantages of fluoroscopy are real-timeand on-line visualization and optimal spatial resolution. Itsdrawbacks include significant x-ray exposure, administra-tion of allergenic contrast media, difficulties with 3-D ren-dering, and reproducibility of the summated images becausespatial orientation is limited, although the use of biplanarfluoroscopy overcomes this limitation to some extent.

Ultrasound

Ultrasound imaging is based on the attenuation and scatter-ing of the reflected sound waves generated by a transducerarray which emits and receives the reflected waves. A sec-tional image is gathered by the transducer array and dis-played in real time. The image is not tomographic becausethe radiation is unidirectional and the scanner is not movedaround the patient. Thus the US image lacks spatial resolu-tion and does not permit a complete view behind structuresthat produce a ‘‘sound shadow’’ such as gallstones or air-filled compartments (Fig. 2). Penetration of bone is minimaland differentiation of soft-tissue planes is limited. Resolu-tion and penetration depth are strongly dependent on thewave length. The Doppler shift is used to detect the velocityof a moving column of blood, and systems are availabelwhich display in color (e.g., blue for venous and red forarterial blood) flow to and away from the probe (colorDoppler). Ultrasound contrast media (micro air bubbles)facilitate tissue discrimination significantly. The advancesin 3-D US rendering during the past 5 years are promisingbut require further refinement before routine clinical appli-cation. Contact ultrasound of organs with high-resolutionlinear-array probes (7.5–10 Mhz) during laparoscopic sur-gery is now used routinely in many centers during biliarysurgery, for the staging of tumors, during laparoscopic cryo-therapy, and for the localization of islet cell tumors. Exter-nal ultrasound-guided insertion of the primary trocar can-nula has been used in the retroperitoneal approach to endo-scopic adrenalectomy, and this technique of ultrasound-guided port insertion carries considerable benefits. As thecurrent systems are not stationary, interactive instrumentguidance requires manual control or coupling of the instru-ment, e.g., biopsy needle or cryoprobe, to the US scanningtransducer head. Evaluation of the percutaneous access tra-

Fig. 1. Fluoroscopic images are acquired on-line and in real time butinvolve a significant amount of radiation exposure. Display of a nonspa-tially defined summation image of the objects within the x-ray beam isobtained.

Fig. 2. Ultrasound imaging requires acoustic coupling to the scanned tis-sue and is altered by sound shadows. Only the tip of an instrument can bedisplayed.

947

jectory with US guidance is difficult and requires consid-erable expertise, and its efficiacy is thus user dependent.Despite these current limitations, which are likely to beovercome as the technology advances, US scanning pro-vides a noninvasive image guidance system of proven valuefor percutaneous biopsy, puncture, and drainage procedures.More recently, US guidance has been reported during neu-rosurgery and for spinal fusion [24].

Computed tomography

CT provides reproducible high-resolution images with gooddifferentiation of soft tissues. As the x-ray beam tube isrotated around the patient, the gathered data on absorptionin relation to the position of the beam allow the calculationof pixels that are displayed in gray shades. In conjunctionwith the slice thickness, the primary two-dimensional natureof the tomographic images permits spatial orientation and3-D rendering. During helical CT scanning the patient ismoved while the x-ray beam is rotated. The acquired data‘‘spiral’’ greatly facilitates 3-D reconstruction.

Although the patient has to be moved in and out of thescanner and the opening of the CT gantry limits surgicalinterventions, CT is a contact-free means of reproducibleand precise interventional imaging with a resolution < 1.0

mm. The major drawbacks relate to X-ray exposure and toonly monoplanar imaging in the axial and paraxial direc-tions (Fig. 3). Electron beam tomography (EBT, Siemens/Imatron) has the advantage over conventional CT scanningin terms of the scanning time, which is reduced to 30 msecby this machine (Fig. 4). In addition, the gantry opening islarger. However, the current image reconstruction time byEBT is approximately 70 s. Thus the latest conventional CTscanners, e.g., the Somatom S4 ‘‘real time’’ (Siemens), arefaster in terms of the actual image representation.

Magnetic resonance Imaging

In contrast to the other imaging systems, MR images arereconstructed from signals emitted from relaxation of hy-drogen proton spins. The signal intensity depends on thehydrogen proton content and type of the chemical binding inthe respective tissue. Thus the image is not only morpho-logical but also contains information on the physical andchemical characteristics. Aside from superior soft-tissuecontrast, various effects, i.e., temperature changes, tissuenecrosis, and blood flow, can be exploited. Instruments,however, are difficult to display with MRI. Due to the volu-metric nature of the MR images, multiplanar and 3-Dimages can be obtained. The application of fast and

Fig. 3. CT scanning is provided in axial and angulated paraxial direction.A The gantry opening of a CT scanner limits intraoperative imaging anduse of instruments. However, the excellent spatial resolution providesaccurate display of both the instruments. However, the excellent spatialresolution provides accurate display of both the instrument and operativevolume,B, C, CT-guided endoscopic thoracic sympathectomy.

948

ultrafast sequences enables ‘‘breathholding’’ and ‘‘fluoros-copy-like’’ imaging. The MR image acquisition parameterscan be externally controlled via computer interfaces, en-abling an interactive slice orientation to the instrument asthis is moved. Scanning time with MRI is relatively unlim-ited as IDA regulations limit the time of exposure to MRI to7.6 Tesla/hour/day ionizing radiation is not involved, and noside effects have been reported provided precautions aretaken to remove all metal objects and exclude certain cat-egories of patients. The new open magnet designs improveaccess to the patient (Fig. 5A–F).

Three-dimensional imaging

The imaging systems previously described provide cross-sectional uniplanar images. Difficult diagnosis and preop-erative surgical planning are facilitated by 3-D reconstruc-tion. For this purpose, the tomographic data are subjected tofurther computed processing. This includes identificationand selection of the tissue and organs to be reconstructed.The classification of tissues relates to the respective grayshade. The preselected gray shades are displayed in colorcodes. This segmentation is usually performed manually ineach of the tomographic slices. Automated segmentationrequires high-performance workstation computers and er-rors due to artifacts are likely with the current generation ofsoftware. The final step in the reconstruction is the 3-Drendering of the respective color-coded segments to thevolumetric model that can be displayed in various angles orcuts (Fig. 6A). For diagnostic applications, the time re-quired for the 3-D reconstruction of the cross-sectional im-ages is not important as it is for the conduct of the surgicaloperations. The intraoperative utilization of 3-D reconstruc-tion of the pathological anatomy requires fast processing ofthe image data in near real time.

In the future 3-D images can be used intraoperatively bymeans of theregistration process whereby the images areprojected on to the patient’s corresponding anatomic region.A video image of the patient or a laser surface scan isaligned with the computer-derived surface of the rendered3-D reconstruction. Prominent anatomical landmarks servefor the video optic matching of the two images. Subse-

quently, the 3-D data are used to display the anatomicalinternal structures by superimposing them onto the patient’sskin (Fig. 6B). In this fashion, the precise outline of thetumor and important or vital anatomical structures whichmust be preserved during surgery are displayed throughoutthe operation—providing a safety road map to the surgeon.

Stereotactic navigation

One of the principles of image-guided surgery is based onstereotactic methods which have been used traditionally byneurosurgeons. The trajectory of the insertion path is evalu-ated initially by conventional radiological skull x-rays orCT or MRI scans. The stereotactic procedure is then per-formed by means of a frame mounted on the patient’s headthat allows advancement of a probe along the line of thetrajectory to the predetermined distance. The instrument isadvanced and positioned without direct control. Only pre-determined distances and angles verify the correct position.The concept of stereotactic instrument guidance has beenadvanced by developments that involve 3-D rendering of awhole set of CT or MR images and optoelectronic trackingof the instrument. Small infrared LEDs (light-emitting di-odes) are attached to the instrument’s handle and these aretracked by linear cameras. These feed this spatial x,y,z in-formation to the controlling computer, which thus receivesinformation on the instrument’s position and orientation.The system then displays a ‘‘virtual’’ instrument (corre-sponding to the real instrument used by the surgeon) on therespective image of the patient, and thus maneuvering of theinstrument can be controlled with great precision and with-out the need of a stereotactic frame—frameless stereotaxis(Fig. 7). This technology is currently being applied in ENTsurgery, orthopedics, and neurosurgery [14, 20]. However,changes within the real operative volume during surgery arenot detected. Thus either update imaging or an additionalon-line imaging system is required.

Techniques of tomographic-guided interventions

CT image guidance

Intraoperative CT scanning has no significant influence onthe operative environment and the nature of the instruments

Fig. 4. Electron beam tomography (EBT) dispenses with the need for rotating x-ray tubes and provides ultrafast scanning of different planes simultaneouslywithin 30–100 msec, whereby motion artifacts and radiation exposure are reduced significantly.

949

used during the procedure. Selection of patients is also notrestricted. Compatibility with CT guidance of materials forinstruments and devices is governed by their radio-opacity.Large metal objects within the scanning field should beavoided. Titanium, aluminum, and synthetic and ceramicmaterials can be used. Smaller instruments and cathetersmust be sufficiently radio-opaque to be visible during the

procedure and can be doped with gold or platinum for thispurpose.

The major drawback of CT interventional guidance isthe movement of the patient in and out of the gantry open-ing, as this restricts surgical instrument access to the opera-tive area. Thus CT scanning has to be performed intermit-tently during these interventions. This limitation is largely

Fig. 5. New open MRI scanners facilitate access to the patient during scanning and provide multiplanar images with superior soft-tissue contrast, displayof physical and biochemical changes, and angiography without contrast agents. Current designs of open and interventional MRI units:A ACCESS Toshiba,San Francisco, CA, USA.B Magnetom Open, Siemens, Erlangen, Germany.C Signa SP, GE Medical Systems, Milwaukee, Wis, USA.D Signa Profile,GE Medical Systems, Milwaukee, Wis, USA.E Horizon, Picker, Cleveland OH, USA.F Airis, Hitatchi, Tokio, Japan.

950

overcome by combination with a flouroscopic C-arm (hy-brid CT-fluoroscopy) which provides a real-time instrumentguidance and adds coronal or sagittal viewing. Percutaneouslaser decompression of spinal disk herniation, endoscopictreatment of intraspinal scars after disc surgery, endoscopic

sympathectomy, as well as local tumor biopsy and intersti-tial therapy are currently performed under combined CTguidance—fluoroscopic control [8, 21, 22].

MR image guidance

The first MR image-guided procedures were performed in1986 with conventional closed (circular magnets) MR units[9, 13, 19]. The closed configuration of conventional MRunits allows high field strength but restricts access to thepatient. Phillips then developed a combination of a fluoros-copy unit and a high field scanner with 1.5 teslas (T) pro-viding fast and ultrafast imaging. The technical realizationof an open magnet MRI unit with large gap and adequatefield strength (0.2–0.5 T) is governed by physical limita-tions and requires either cost-intensive superconductingcoils or heavyweight resistive or permanent magnets weigh-ing 10–50 tons.

Open MRI units (Fig. 5A–F)

In 1986 the first open imaging unit with a temple-like con-figuration (Access, 0.067 T) was developed by Toshiba(San Francisco CA, USA). The low magnet field reducesartifacts and facilitates application but has limited imagingcapabilities compared to higher field units. Since then, newfast imaging with keyhole sequences and techniques using

Fig. 6. A3-D reconstruction of CT images of intraspinal micro-endoscopy.B Projection of an MR image on the skin of the patient within an openmagnet (F. Jolesz, Brigham and Women’s Hospital, Boston, MA, USA).

Fig. 7. Opto-electronic Instrument Navigation System SPOCS (Aesculap,Tuttlingen Germany).

951

MR-guided procedures have been established [7]. The newmodel of the Access will be equipped with superconductingmagnets and a field strength of 0.35 T.

The new Magnetom OPEN (Siemens, Erlangen, Ger-many), comprises an 0.2–T resistive magnet with an open-ing gap of 440 mm and 260° circumferential horizontalaccess. Fast, good-quality imaging is provided through ac-tive field stabilization and water-cooled gradient coils. In-terventional equipment is available that comprises a LCDin-room monitor console, special knee chair with heel ped-als, foot pedal, cold light source, operative field illumina-tion, and special fast sequence protocols. The ‘‘autodis-play’’ function permits instant view of the images while themeasurement protocol is still running, thereby providingnear-real-time visualization of the instrument during its ad-vancement.

Picker has released an open MR unit (Horizon) with adesign similar to that of the Siemens Magnetom OPEN. Themagnetic field strength is slightly higher (0.23 T) and theopening gap is 5 cm wider. The two-axis patient bed can beremoved. Interventional equipment is under development.

A dedicated MR-therapy system has been realized byGE Medical Systems (Milwaukee, Wisconsin USA) in col-laboration with the group of F. Jolesz, Boston. The verticalopen magnet configuration comprises two opposed super-conducting 0.5-T magnet rings that enable vertical accessthrough the operating gap of 560 mm. The patient’s bed canbe moved through the rings in the longitudinal directiononly. An additional MR observation monitor and an opto-electronic navigation system are incorporated in the uppersection between the magnet rings [23].

Application of vertical or horizontal openMR configuration

The selection of the appropriate opening direction of anMRI unit depends on the nature of the intervention to beperformed.

Two types of procedures can be performed currently:perioperative MRI-assistedand intraoperative MRI-guidedinterventions.

MR-assisted procedures.The horizontal open magnets areideal for perioperative MRI procedures that require frequentmovement of the patient; i.e., surgery is performed outsideof the magnet and scanning is conducted only for intermit-tent assessment of the operative tissue volume. This ap-proach is ideal for procedures in which an additional on-linemeans of visualization is applied, e.g., video-endoscopy.Configurations that allow free access from three sides in ca.260° and two axes of patient’s bed movement ease thechange from MR scanning to the operating table. Horitzon-tal open Siemens scanners have been installed in the neuro-surgical OR at the university hospitals of Heidelberg andErlangen in Germany.

MRI-guided procedures.Percutaneous interventions ben-efit from performance under MR control, preferably bymeans of interactive MR fluoroscopy (Fig. 8). We have

carried out 163 MRI-guided procedures such as aspirationbiopsy, peridural and periarticular corticoid injection, facetjoint treatment, chemical sympathicotomy, and intratumoralethanol infiltration in 131 patients. This experience has con-firmed the feasibility and benefits of MRI guidance.

Procedures where drugs or energy are delivered to aspecific site should be guided either by MR ‘‘fluoroscopy’’or intermittent scanning with temperature-sensitive se-quences (turbogradient echoes). If horizontal access to theoperative volume is possible, the horizontal open-designMRI unit is appropriate. However, the vertical open MRIdesign gives more degrees of freedom and an ergonomicallybetter access to the operative site. However, the transfer ofthe patient from the operating table to the MRI unit is moredifficult because the patient has be to moved in and outthrough the magnet rings. This interferes with the instru-ments, supply lines, and anesthetic tubing. IntraoperativeMRI guidance is also important during the placement ofsurgical instruments or endoscopes within the operativefield, e.g., neurosurgical access to the ventricular system orcystic malformations. In Heidelberg the first MR-guidedtranscranial approach to an intracerebral cystic malforma-tion has been performed successfully. At the Brigham andWomen’s Hospital in Boston, 50 CNS biopsies, sinus sur-gery, and open brain surgery have been performed underMR guidance using the vertical open unit [11].

Prerequisites and techniques of MRI guidance

There are certain specific technological requirements forMRI-guided techniques and these procedures differ consid-erably from other interventional techniques.

Development of MR-compatible instruments, devices,and ancillary equipment as well as optimization and adap-tation of software and hardware of the current MR systemsare the major objectives of the current developmental pro-gram. We have been working on the open 0.067-T Access(Toshiba) since 1987 [6] and in 1995 we started a collabo-ration program with Siemens (Erlangen, Germany) to evalu-ate the feasibility of interventions in the Magnetom OPENsystem [18].

Fig. 8. Intratumoral ethanol ablation of a sacral tumor.

952

Patient selection and patient monitoring.In contrast to CT-or US-guided interventions patient selection for MRI-guided procedures is governed by physical considerations.The following conditions constitute absolute contraindica-tions for MRI guidance:

● Pacemakers● Neurostimulation devices● Known or suspected ferromagnetic aneurysm clips● Large ferromagnetic implants

Patients with known history of ischemic heart disease orwith symptoms of angina and ST changes should not besubjected to MRI-guided procedures because ST and Twaves are distorted by the magnetic field, and for this rea-son, detection of a myocardial ischemic episode during theintervention may be obscured [10].

MRI-compatible anesthetic equipment, optoelectronicpulse/ECG sensors, pulse oximetry, and blood gas analysisare available (Magnetic Resonace Equipment Corporation,Bay Shore, NY, USA; Daum, Schwerin, Germany; and oth-ers). However, a complete and safe intraoperative monitor-ing system for patients during MRI-guided surgery requiresadditional development.

Interventional MR sequences.Adaptation of fast sequencesand the protocols to the nature of the intended procedure isnecessary. Time of acquisition, TR, is the time betweeneach pulse sequence—an image with 256 lines requires 256pulse sequences. Reduction of the lines to a rectangular fieldof view reduces the sequence time. Echo time is the timebetween end of the pulse sequence and the start of signalacquisition and leads to the different weighing of spin echosequences. The motion artifacts are reduced by short echotimes. Standard (SE) spin echo pulse sequences have longacquisition times but provide excellent soft-tissue contrast.T2-weighted spin echo images have long echo times. Gra-dient echo (GE) sequences are faster but more subject toartifacts. The weighting of fast gradient echoes is influencedby the flip angle. Thus 90° enhances the weighting of T1and angles <30° result in T2* contrast. GE sequences aremore sensitive to inhomogeneities of the magnetic field andthus the artifacts of instruments are larger than when SEsequences are used. The current fast sequences are 2-DFLASH (fast low-angle shot) and 2-D FISP (free inductionsteady precession). 3-D sequences acquire a volume (slab)that is segmented into slices. 3-D sequences are more sen-sitive to motion artifacts. The crucial component of fast andultrafast imaging is the quality, strength, and speed of thegradient coils. The ultrafast echo planar imaging (EPI) hasacquisition times in the millisecond range because all 256lines of the image are acquired within one TR time. Atpresent, EPI is applicable only to high field units. Newapproaches to faster imaging are ‘‘keyhole’’ and ‘‘wavelet-encoded’’ sequences. Both are based on the principle ofavoiding complete acquisition of the data set required forthe whole image. Instead acquisition is restricted to the vol-ume of interest where, for example, the interventional in-strument is moved [1, 25].

Although heat generation can be directly displayed bymeans of Turbo flash 2-D echoes, their image quality is, in

low-field systems, still suboptimal. The signal changes de-tected by FISP 2-D or flash 2-D echoes are useful to evalu-ate to extent to localized tissue destruction. Image-guidedtumor ablation requires precise control with standard T1 andT2 sequences.

MRI-compatible materials, cannulae, and catheters.Instru-ments and probes that are inserted within the image planereduce the signal intensity. Magnetic field inhomogenetiesaround the instrument by virtue of their paramagnetic ef-fects decrease the relaxation time of the protons encirclingthe instrument such that their signal becomes undetectablewithin the echo time. Control of the artifact size can beachieved by electric current application [5]. The size of theartifact depends on the material, the applied sequences, andthe angle of the probe within the magnetic field. CurrentMRI-compatible instruments are manufactured from anti-magnetic stainless steel, titanium, and nickel-titanium al-loys. MRI-compatible needles and microtrocar/cannulaewith controllable artifact size and a variable-curvature guid-ing cannula (SmartGuide, Daum Medical, Schwerin) manu-factured from these materials have been developed andtested. SmartGuide consists of a superelastic internal pre-curved cannula that is constrained inside a straight outersleeve. The same concept has been applied to the variablecurvature spatual according to Cuschieri. When the innersuperelastic component is advanced beyond the end of therestraining outer sleeve, ‘‘round the corner’’ tissue passageis achieved (Fig. 9). A further development based on theSmartGuide principle has been applied to catheters for im-age-guided interventions. Intravascular interventions posespecial problems for the imaging of static elements becausethe MR angiography sequences display the blood flow se-lectively. Conventional synthetic catheter materials are notvisible in MR images. Various dotting techniques aimed atgeneration of the appropriate artifact which would allowMRI detection of these catheters are under evaluation.

Video optic equipment, trocars, and instruments for MR-guided endoscopic surgery.The nature of MRI specificallyimposes special requirements for instruments, devices, andelectric equipment that can be used within the proximity of

Fig. 9. MRI-compatible cannula with variable curvature (SmartGuide,Daum Schwerin).

953

the magnetic field. In collaboration with OlympusWinter&Ibe we have developed MRI-compatible rod-lensendoscopes. New light cables with a length of 5 m can beconnected to a powerful cold light source placed outside theMR suite. A novel MR cold light source (Siemens) dis-penses with the need for a ventilator, but its maximal currentpower is only 100 W. Prototypes of MRI-compatible endo-cameras have been released for evaluation by Olympus. Theadaptation of the new in-room MRI monitor as an endos-copy monitor is possible and should provide adequate imagequality because of high resolution and multisync capability.

There is a need for MRI-compatible trocar/cannulae forMR-guided laparoendoscopic surgery. The MRI compat-ibility of trocar/cannulae has been evaluated in vitro (Fig.10A,B). The ferromagnetic characteristic of forged or hard-ened stainless steel excludes the majority of trocars. In ourtests we found the Step system (Innerdyne, Sunnyvale Ca,USA) [17] to be the most MRI-compatible access port. Theinitial puncture with a Veress needle and insufflation can beperformed under MR fluoroscopic control. MR visualiza-tion of balloon dissection procedures [2] can be achieved byfilling the balloon with a 2–3 mmol/l gadolinium/saline con-trast agent. Major redesign and development work is neededfor endoscopic instruments such as probes, hooks, forceps,and scissors to achieve MR compatibility. The substitutionfor the steel components of the endoscopic instruments suchas the hinge complex with superelastic shape-memory Ni-Tiis an obvious solution [15, 16].

Interactive image control.The visualization of instrumentsduring MRI-guided procedures is grossly inferior to thatobtained by CT. The instrument has to be aligned com-pletely with the orientation of the image plane. During MRI-guided work, any movement by the patient during the in-sertion of the instrument (e.g., needle) leads to loss of the

needle artifact within the MRI image. This can be overcomeby changing the slice orientation and position when a rep-resentation of the changed needle path can be achieved. Thisprocess is, however, time consuming and awkward.

The principle of optoelectronic frameless instrumentnavigation can be implemented in the control system of anMRI unit whereby the tracking of an instrument within themagnetic field can be directly transferred to the slice orien-tation. (Fig. 11) Images can then be generated according tothe position in alignment with the instrument or perpendicu-lar to the distal tip of the instrument. The image planeorientation can be changed and redefined without movingthe instrument itself. This ensures constant precise controlof a surgical instrument during advancement. Care must betaken to avoid any bending of the instrument as the distortsthe actual position of the instrument tip [23].

Tip tracking. The volumetric nature of the MRI signal de-tection allows orientation of specific signal frequencies inspace. With the incorporation of a miniature radiofrequencycoil at the tip of an instrument, local proton spin signals canbe acquired. The signal source representing the tip of theinstrument can then be localized within the coordinate sys-tem of the magnetic field and thus within the human body[4, 12]. This technique has been evaluated experimentallyand is referred to as tip tracking. As the instrument ismoved, the position of the signal moves accordingly and canbe tracked with great precision. The current approach to tiptracking by Doumulin, Leung, and Debatin involves theneed for two workstations to compute the three-dimensionalcoordinates of the coil and to display the position as a mov-ing symbol in two planes of an acquired MR image. Thusdirect on-line control is not possible. We have started atip-tracking project that allows direct display of the instru-ment signal in a complete MR image, thereby enabling in-

Fig. 10. A Experimental setup of MR-guided laparoscopy.B MRI representation of an MRI-compatible trocar, laparoscope (Olympus Winter&Ibe,Hamburg, Germany), and a superealistic laparoscopic forceps (ND&C, Fremont, CA, USA).

954

teractive slice orientation in the axial or perpendicularplanes with the instrument.

Conclusion

The goal of safe conduct of minimal-access therapeutic pro-cedures requires continuous reference image data sets andcorrespondence between these and the instrument’s positionin relation to the anatomical structures within the operativetissue volume. Accurate control of each step of the proce-dure must be provided and unexpected changes in the ana-tomical geometry within the operative volume should bedetected on-line.

A step toward these ideals is the utilization of stereo-tactic navigation systems that display a virtual instrumentwithin presurgically gained image data. The relativelystable proportions of the operative volume in sinus and in-tracranial surgery permits appropriate intraoperative local-ization of the instruments position. However, there are stillproblems to be solved during neurosurgery as removal ofcerebrospinal fluid, hemorrhage, and cerebral edema induceshifts in the spatial configuration of the operative volume,and the actual process of tissue removal is not displayed.

During thoracic and abdominal procedures there is acontinuous and significantly greater shift of structures dueto breathing, heart beat, intestinal peristaltic activity, etc.With the current systems, external scanning performed pre-operatively cannot determine the displacement of the un-derlying thoracic or abdominal structures. The gross ana-tomical displacements that occur during surgery could only

be deduced from computer simulation. This situation is in-adequate since despite being informative, it does not guar-antee safety during the actual intervention.

There is a need for intraoperative imaging capable offast upgrading of the original data set of the operative vol-ume during the surgical intervention. The optimal solutionto this problem entails the optoelectronic instrument navi-gation with ‘‘tip tracking’’ as the reference for acquisiton ofaccurate intraoperative image planes (Fig. 11). Accurate on-line imaging of the operating instrument is crucial to theprogress of image-guided surgery. The ideal imaging sys-tem should permit:

● Unrestricted conduct of the interventional or surgical pro-cedure

● Reproducibility● On-line and real-time visualization● Appropriate soft-tissue and bone discrimination● Precise display of the instrument● Volumetric data and spatial references● Facility for multiplanar slicing and interactive selection

of image orientation● Flow and spatiothermal sensitivity

Ultrasound, fluoroscopy, and CT are established in clinicalroutine and have specific indications and advantages. US issimple to apply during procedures where the acoustic cou-pling to the tissue is possible. The high resolution of CT (<mm) and its excellent bone delineation make it the idealimage-guiding system in high-risk areas and when bonystructures are part of the operative volume, i.e., intraspinal

Fig. 11. Schematicrepresentation ofopto-electronic instrumentnavigation in combinationwith tip tracking. Thisapproach provides interactiveMRI imaging in an axial orperpendicular directionrelative to the instrumentmovement.

955

and skull-base microsurgery. The instrument display withinthe image is significantly better with CT than MRI, and theposition of the functional tip is visualized with much greaterprecision. For intravascular procedures, fluoroscopy is stillunrivaled, although recent progress in MR angiography(MRA) has clearly demonstrated the great potential forMRA in the evaluation and treatment of vascular disease[16]. In principle MRI is capable of meeting most, if not all,the above-mentioned requirements, but the instrument dis-play need significant improvement. What is needed is fullintegration of the imaging systems with the technology usedduring surgical operations and with the environment of theoperating room. The selection of the imaging tool dependson the procedure to be performed and requires extensiveclinical evaluation.

Tomographic guidance and endoscopic surgery involvecomplementary technologies. The endoscope permits a col-ored, high-resolution, on-line, and real-time image of theoperative field, i.e., accurate visualization of the surfaceanatomy of the internal organs and tissues. However, it doesnot provide any information on the parenchyma of organs,supporting tissues, and intraparenchymal pathology, bloodvessels, and other vital structures (operative tissue volume).Visualization of the trajectory of the instrument’s pathwaythrough the tissues cannot be provided by endoscopic im-aging. Intraoperative tomographic imaging reveals all theessential components of the parenchymal structures withinan operative volume. As soon as the operative field is ac-cessed and exposed, the patient can be moved out of thescanner, and the operation can be conducted without anyspatial restriction. During the course of the operation,should further tomographic visualization of the operativetissue volume become necessary, the patient is scannedagain to obtain refreshed tomographic images. The opticalnavigation and tip-tracking systems can be used to give abaseline orientation within the operative volume in betweenthe scanning periods.

The majority of imaging systems are currently in handsof radiologists who do not usually perform surgical proce-dures. On the other hand, surgeons who perform the vastmajority of operations and interventions are missing outsince they undertake complex endoscopic operations with-out the benefit of tomographic imaging. The future lies inclose interdisciplinary collaboration between these two spe-cialties with the establishment oftomographic image-guided endosurgery.Even in the absence of such multidis-ciplinary units, as we approach the next millennium, sur-geons have to become proficient with tomographic imageguidance. The safety and precision of their surgical inter-ventions will be enhanced; otherwise, they will lose out tothe radiological surgeonsor surgical radiologistsof thenext century.

References

1. Bornsted A, Busch M, Wendt M, Seibel R, Gro¨nemeyer D (1996)Turbo accelerated asymmetric phase read exchange keyhole (Phreak)imaging as a method for fast and high contrast MR-guided-interventions. MAGMA 4, Supplement ESMRMB ‘96, Nr 260, p 142

2. Coptcoat MJ (1995) Overview of extraperitoneal laparoscopy. EndSurg 3: 1–3

3. Cuschieri A (1992) General principles of laparoscopic surgery. In:Cuschieri A, Buess G, Pe´rissat J (eds) Operative manual of endoscopicsurgery. Springer, Heidelberg pp 172–175

4. Davis CP, McKinnon GC, Debatin JF, von Schulthess GK (1996)Ultra-high-speed MR Imaging. Eur Radiol 6: 297–311

5. Glowinski A, Adam G, Bu¨cker A, Neuerburg J, van Vaals J, Gu¨ntherRW (1996) Catheter visualization for interventional MR by activelycontrolled locally induced field inhomogeneities. Procedings, 51,ISMRM, New York

6. Groenemeyer DHW, Seibel RMM, Busch M, Rothschild PA, Kauf-mann L (1989) Interventionelle Kernspintomographie. In: Groenem-eyer DHW, Seibel RMM (eds) Interventionelle Computertomogra-phie. Ueberreuter, Wien-Berlin, pp 296–313

7. Gronemeyer D, Seibel R, Kaufman L (1991) Low-field design easesMRI-guided biopsies. Diagn Imaging 3: 161–165

8. Groenemeyer DHW, Seibel RMM, Melzer A, Schmidt A (1995) Im-age guided access techniques. End Surg 3: 69–75

9. Hajek PC, Gylys-Morin VM, Stava J, van Sonnenberg E (1987) Lo-calization grid for MR guided biopsy. Radiology 165: 825–826

10. Hughes CW, Bell C (1993) Anesthesia equipment in remote hospitallocation. In: Rogers MC, Covino BJ, Tinker JH, Longecker DE (eds)Principles and practice of anesthesiology. Mosby, St. Louis, MO, pp3–8

11. Jolesz F (1996) MRI guided therapy. MAGMA 4, SupplementESMRMB ‘96, Nr 280, p 150

12. Leung DA, Debatin JF, Wildermuth S, Heske N, Dumoulin CL, Dar-row RD, Hauser M, Davis CP, von Schulthess GK (1995) Real-timebiplanar needle tracking for interventional MR imaging procedures.Radiology 197: 485–488

13. Lufkin R, Teresi L, Hanafee W (1987) New needle for MR-guidedaspiration cytology of the head and neck. AJR Am J Roentgenol 149:380–382

14. Melzer A (1996) Instruments for laparo-endoscopic surgery: conven-tional and intelligent. In: Tuoli D., Gossot J., Hunter J (eds) Endosur-gery

15. Melzer A, Stoeckel D (1995) Performance improvement of surgicalinstrumentation through the use of Ni-Ti materials. ProceedingsSMST-94, Monterey, CA, USA, pp 401–410

16. Melzer A, Buess G, Cuschieri A (1994) Instruments for endoscopicsurgery. In: Cuschieri A, Buess G, Perissat J (eds) Operative manualfor endoscopic surgery, vol 2. Springer, Berlin

17. Melzer A, Kipfmuller K, Groenemeyer D, Seibel R, Bueb G (1995)Ports, trocars/cannulae and other access techniques. Semin Laparosc 2:170–201

18. Melzer A, Schmidt A, Kipfmu¨ller K, Deli M, Stockel D, Seibel R,Gronemeyer D (1996) Prerequisites for magnetic resonance image-guided interventions in endoscopic surgery. Min Inv Ther AlliedTechn 4: 260–266

19. Mueller PR, Starck DD, Simeone JF (1986) MR guided aspirationbiopsy: needle design and clinical trials. Radiology 161: 605–609

20. Nolte LP, Zamorano LJ, Jiang Z, Wang Q, Langlotz F, Berlemann U(1995) Image-guided insertion of transpedicular screws. A laboratoryset-up. Spine 20: 497–500

21. Seibel R (1996) Image guided minimally invasive therapy. Surg End

22. Seibel RMM, Groenemeyer DHW (1994) Technique for CT guidedmicro-endoscopic dissection of the spine. End Surg 3(4): 226–230

23. Silverman SG, Collick BD, Figueira MR, Khorasani R, Adams DF,Newmann RW, Topulos GP, Jolesz FA (1995) Interactive MR-guidedbiopsy in an open configuration MR imaging system. Radiology 197:175–181

24. Wenda K, Degreif J, Hu¨wel N, Kessel G (1995) Die Reposition derfrakturierten Wirbelko¨rperhinterkante unter sonographischer Kon-trolle. Langebecks Arch Chir 95: 477

25. Wendt M, Busch M, Lenz M, Seibel R, Gro¨nemeyer D (1996) Dy-namic tracking algorithm for interventional MRI using wavelet encod-ing in 3D gradient-echo-sequences. MAGMA 4, SupplementESMRMB ‘96, Nr 261, p 143

956

Case reports

Covered metal stent for tumor obstruction of efferent loop recurrenceafter gastrectomy

H. Y. Cheung, S. C. S. Chung

Department of Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong

Received: 13 June 1996/Accepted: 31 July 1996

Abstract. Reports the case of a 60-year-old woman whounderwent R2 total gastrectomy, and subsequent palliationof painful symptom recurrence via a membrane-coveredmetal stent.

Key words: Covered metal stents — Efferent loop recur-rence — Gastrectomy

Dysphagia resulting from jejunal loop obstruction by recur-rent tumor can be a challenging clinical problem. We de-scribe successful palliation in a patient with distressful dys-phagia by implantation of a membrane-covered expandablemetal stent.

Case report

A 60-year-old woman underwent R2 total gastrectomy with Roux-en-Yesophagojejunostomy for carcinoma of the proximal stomach. Pathologicalexamination showed T2N0 disease and the resection margins were free oftumor. She made an uneventful recovery but re-presented 2 years and 4months later with progressive dysphagia. Ultrasound examination showedmultiple liver metastasis. Endoscopy showed a healthy anastomosis but theefferent loop was invaded by tumor 5 cm from the esophagojejunal anas-tomosis. The endoscope cannot be passed through the obstruction. Injec-tion of contrast material down the channel of the endoscope showed ashort, narrowed segment with acute angulation (Fig. 1). The obstructionwas negotiated with a Zebra guide wire (Microvasive, Boston ScientificCorporation, USA) and the narrowed segment was dilated with a 15-mm-diameter balloon catheter. The proximal and distal limits of the obstructedsegment were then marked by injection of 0.5 ml of Lipiodol (LipiodolUltra-Fluid, Guerbet, France). A covered nitinol stent (Covered Ultraflex,Microvasive Boston Scientific Corporation, USA) with a 12-cm-longmembrane-covered segment was then implanted in the following manner.A Savary guide wire (Savary Gillard, Wilson-Cook Medical, Inc., USA) ispassed into the jejunum and the stent is inserted over the guide wire (Fig.2). With the distal extent of the stent in a straight segment of the jejunum,the stent was deployed by removing the crocheted thread that held the stentonto the introducer. Deployment started distally and progressed in a proxi-

mal manner, and the position of the stent can be adjusted by pulling backthe introducer shaft as necessary. The proximal end of the stent was left inthe esophagus above the esophagojejunal anastomosis (Fig. 3). The de-ployment process was monitored by x-ray and an endoscope was insertedalongside the stent. At the end of the procedure injection of contrast downthe endoscope showed free flow (Fig. 4). The patient was able to take solidfood until she succumbed from her disease 6 months later.

Discussion

After gastrectomy for carcinoma of stomach, local recur-rence causing dysphagia occurs in about 20% of patients.Reoperation is seldom justified because of the limited lifeexpectancy. Anastomotic recurrence may be amenable toendoscopic laser therapy if the obstruction if predominantlyintraluminal and exophytic [1, 4, 7, 8]. Tumor invasion ofthe efferent loop is even more difficult to manage as themain bulk of the recurrent tumor is extraluminal or fromextrinsic compression. Dilatation alone will only give veryshort-lived relief. Because the obstructed segment is oftenangulated, implantation of conventional plastic stents ishazardous and is unlikely to be successful because thestraight stents cannot conform to the curvature of the jejunalloop.

The availability of self-expanding metal stents has madeesophageal stenting much safer because the stricture doesnot need to be dilated to a large size and the process of stentinsertion is less traumatic. Perforation, the most feared com-plication of stenting, can largely be avoided. Knyrim [6]reported that no perforation occurred in metal stent place-ment in his randomized study. He concluded that self-expanding metal stents are safer and more cost-effectivethan conventional plastic prostheses. However, tumor in-growth through the instices of the uncovered expandablestent can cause recurrent dysphagia 2–6 months later [3, 5].The second-generation expandable metal stents have mem-brane covers to overcome this problem [9, 10].

At the moment there are several types of commercialexpandable stents available. They differ in the method ofdeployment, the radial force they generate, the mechanismof anchorage, and their ability to conform to a curvature [2].Correspondence to:S. C. S. Chung

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 936–938

These factors must be considered in choosing the right stentfor a particular patient. In our case, because of the acuteangulation in the obstructed segment, the ability to with-stand a sharp bend without buckling is the most importantfactor. The particular stent we used is deployed by pullingon a thread crocheted onto the stent to bind it to the intro-ducer. The process of deployment starts distally, so it ispossible to adjust the position of the stent proximally bypulling on the introducer during the deployment process.

This is important when precise deployment of the stent isnecessary. Nitinol is an alloy of nickel and titanium that hasthermal as well as ‘‘shape’’ memory; in addition, its super-elastic behavior enables the stent to fit the contour of thestricture without buckling. The longer stent lengths (150mm) as compared to Wallstent (110 mm) contributed anadvantage in managing tortuous efferent loop recurrenceand prevent stent migration.

The main drawback of expandable metal stents is their

Fig. 1. Short, narrowed segment with acute angulation (arrow) at the efferent loop.

Fig. 2. The undeployed stent is in place with the proximal and distal limits of stricture marked by Lipiodol.

Fig. 3. The proximal end of the stentis placed in the esophagus above theanastomosis.

Fig. 4. Free flow of contrast into thejejunal loop.

937

high cost. Whether their use is justified in patients withdisseminated malignancy and limited life span remains adifficult financial and ethical problem.

References

1. Buset M, des Marez B, Baize M, Bourgeois N, de Boelpaepe C, deToeuf J, Cremer M (1987) Palliative endoscopic management of ob-structive esophagogastric cancer: laser or prosthesis? Gastrointest En-dosc 33(5): 357–361

2. Chung SCS, Shin FG, Chan ACW, Lau JYW, Sung JY, Li AKC(1996) Physical properties of expandable esophageal stents. Gastro-intest Endosc 43 (4): 332

3. Cwikiel W, Stridbeck H, Tranberg KG, von Holstein CS, HambraeusG, Lillo-Gil R, Willen R (1993) Malignant esophageal strictures: treat-ment with a self-expanding nitinol stent. Radiology 187: 661–665

4. Ell C, Riemann JF, Lux G, Demling L (1986) Palliative laser treatmentof malignant stenoses in the upper gastrointestinal tract. Endoscopy 18(Suppl 1): 21–26

5. Ell C, Hochberger J, May A, Fleig WE, Hahn EG (1994) Coated and

uncoated self-expanding metal stents for malignant stenosis in theupper GI tract: preliminary clinical experience with Wallstents. Am JGastroenterol 89(9): 1496–5001

6. Knyrim K, Wagner HJ, Bethge N, Kermling M, Vakil N (1993) Acontrolled trial of expansible metal stent for palliation of esophagealobstruction due to inoperable cancer. N Engl J Med 329(18): 1302–1307

7. Krasner N, Barr N, Skidmore C, Morris AI (1987) Palliative lasertreatment of malignant dysphagia. Gut 28(7): 792–798

8. Mellow MH, Pinkas H (1985) Endoscopic laser therapy for malignan-cies affecting the esophagus and the gastroesophageal junction. Analy-sis of technical and functional efficacy. Arch Intern Med 145 (8):1443–1446

9. Song HY, Do YS, Han YM, Sung KB, Choi EK, Sohn KH, Kim HR,Kim SH, Min YI (1994) Covered, expandable esophageal metallicstent tubes: experiences in 119 patients. Radiology 193: 689–695

10. Wu WC, Katon RM, Saxon RR, Barton RE, Uchida BT, Keller FS,Rosch J (1994) Silicone-covered self-expanding metallic stent for thepalliation of malignant esophageal obstruction and esophagorespira-tory fistulas: experience in 32 patients and a review of the literature.Gastrointest Endosc 40: 22–33

938

The author replies

I appreciate the comments made by Dr. Deitel and can onlyagree to all of his statements. After sharing the experienceof Dr. W. Clark we have also started to perform the Roux-en-Y gastric bypass with the laparoscopic approach. Withthe laparoscopic instrumentation available today this proce-dure seems to be quite feasible and may be even easier to dothan the loop gastrojejunostomy. The main concern today is,as stated by Dr. Deitel, to avoid any disastrous consequencethat could be related to the laparoscopic technique. Thelong-term outcome must of course be at least as good as theresults seen in open surgery.

The adjustable band has rapidly become very popularsince this procedure is easy to perform with the laparoscopic

technique. There is still very limited long-term experiencewith this new procedure. I agree with Dr. Deitel that weshould be very cautious when applying new techniques andnot repeat mistakes already made by ourselves and others inopen surgery.

H. Lonroth

Department of SurgerySahlgren’s University HospitalS-413 45 GothenburgSweden

Surg Endosc (1997) 11: 966

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

The optimum pneumoperitoneum pressure for laparoscopic surgery inthe rat model

A detailed cardiorespiratory study

R. Berguer, T. Cornelius, M. Dalton

Department of Surgery, University of California Davis, Sacramento, CA, USA

Received: 18 December 1996/Accepted: 13 March 1997

AbstractBackground:There is increasing interest in the rat model oflaparoscopic surgery. This study evaluates the cardiorespi-ratory effects of increasing CO2 pneumoperitoneum (PP) inthe rat.Methods:Nine Sprague-Dawley rats were subjected to CO2PP at pressures of 2, 5, 10, and 15 mmHg or control (no PP)under anesthesia. Catheters were placed in the femoral ar-tery and the jugular vein to measure heart rate (HR), bloodpressure (MAP), and arterial pH, PCO2, PO2, and HCO3. Athermistor probe in the aortic arch measured cardiac output(CO) and blood temperature (BT).Results:CO2 PP had no effect on CO, MAP, or BT at anypressure. CO2 PP greater than 5 mmHg caused significantbradycardia and CO2 PP greater than 10 mmHg causedsignificant respiratory acidosis.Conclusions:CO2 PP pressures above 10 mmHg in ratshould be avoided when performing laparoscopic surgery inthe rat model.

Key words: Rats — Pneumoperitoneum — Cardiovascular— Laparoscopy

Since the first reports of laparoscopy in the rat [5, 8] therehas been increasing interest in using a rodent model forresearch [1–4, 11, 20] and, to a lesser extent, for technicaltraining [16] in laparoscopic surgery. If valid experimentalcomparisons of the tissue effects of laparoscopic and openprocedures are to be made in a rodent model, the animals’cardiorespiratory physiology under anesthesia must be com-parable during the two procedures. In humans and largeanimals it is well known that a 15 mmHg CO2 pneumoperi-

toneum decreases tidal volume and assisted hyperventila-tion is needed to compensate for this effect under anesthe-sia. Survival laparoscopic procedures in the rat also requirea pneumoperitoneum but, due to the difficulty of endotra-cheal intubation, the anesthetized rat must breathe sponta-neously. Under these conditions, laparoscopic surgery inrodents with a standard pneumoperitoneum pressure of 15mmHg presents a clear risk of respiratory depression fromthe combined effects of anesthesia and the pneumoperito-neum which could result in circulatory shock and respira-tory acidosis. Therefore, the purpose of this study is todetermine the maximum pressure for CO2 pneumoperito-neum pressure in the anesthetized and spontaneouslybreathing rat that allows normal cardiac and respiratoryfunction.

Materials and methods

Male Sprague-Dawley rats weighing between 330 and 450 g were studiedunder general anesthesia consisting of 1–2% inhaled isoflurane in a 50%mixture of O2 and N2O delivered at a gas flow rate of 2 l/min via asemiclosed anesthetic circuit and close-fitting masks. We selected thisanesthetic regimen because we have previously demonstrated that thisinhaled anesthetic mixture results in excellent analgesia and muscle relax-ation during laparoscopic and open surgery in rats while causing signifi-cantly less respiratory depression than standard pentobarbital anesthesia[7]. The time from the initial administration of anesthetic until surgicalanesthesia was achieved was recorded as the induction time. Rats wereplaced on an electric warming blanket to maintain their rectal temperaturein the range of 36.5–37.5°C. A polyethylene femoral arterial catheter (out-side diameter 0.965 mm) was placed for blood pressure recording andarterial blood gas sampling. The internal jugular vein was similarly can-nulated for central venous pressure (CVP) measurements and saline infu-sion. No intravenous infusion was given in the present study since rats thatundergo experimental survival laparoscopic procedures receive no main-tenance intravenous fluid. A thermistor probe was placed via the rightcarotid artery into the aortic arch for measurements of cardiac output usinga Cardiomax 2, model 95 cardiac output monitor (Columbus Instruments,Columbus, Ohio). The cardiac output was measured using the thermodi-lution method by injecting 0.1 ml of room temperature saline into theinternal jugular catheter. Each cardiac output measurement was calculated

Correspondence to:R. Berguer, 150 Muir Road (112), Martinez, CA94553, USA

Surg Endosc (1997) 11: 915–918

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

as the average of three successive readings at each time point. Followingplacement of the three catheters, baseline measurements were obtained forheart rate (HR, beats/min), mean arterial pressure (MAP, mmHg), cardiacoutput (CO, ml/min), arterial pH, PCO2, PO2, base excess (BE), HCO3, andblood temperature (BT). Arterial blood gas measurements were made bywithdrawing 0.1 ml of blood from the femoral arterial line through acapillary tube into a 1-ml syringe. The excess blood was then flushed intothe circulation via the arterial catheter with an additional with 0.2 ml ofsaline solution.

The pneumoperitoneum was created by making a 2-mm midabdominalskin incision, placing a standard Veress needle (Ethicon Endosurgery, Cin-cinnati, OH) into the peritoneal cavity, and insufflating CO2 through astandard laparoscopic insufflator (Karl Storz, Germany) set at the mini-mum flow rate. Five rats were studied at pneumoperitoneum pressures of0, 2, 5, and 10 mmHg and four additional rats underwent anesthesia with-out pneumoperitoneum for control purposes. Cardiorespiratory measure-ments were obtained in all animals following catheterization (baseline), 10min after reaching each new level of pneumoperitoneum, and again 10 minafter desufflation. At the end of the experiments the rats were sacrificedwith intravenous KCl. Data for each physiologic variable at each time pointare expressed as the mean value ± SEM for each experimental group.Baseline data were compared to values obtained at each time point and thepneumoperitoneum group was compared to control animals using a two-way repeated measures ANOVA and Tukey’s Honest Significant Differ-ence Post Hoc test. Statistical significance was defined asp < 0.05.

Results

There were no significant differences in animals’ bodyweight, time to induction, and total experimental time be-tween the pneumoperitoneum and control groups. Bloodtemperature demonstrated a global decrease from baselineto desufflation (means of 37.3 and 36.9, respectively,p 40.043) that was not significantly different between groups.

Heart rate decreased significantly with increasing pneu-moperitoneum pressure and was significantly lower than incontrol rats at a pneumoperitoneum pressure of 10 mmHg(Fig. 1A). MAP did not change significantly over time andwas similar between groups (Fig. 1B). Cardiac output wassignificantly decreased in the pneumoperitoneum groupcompared to the control group only at a pressure of 5 mmHgand then demonstrated a significant elevation after desuf-flation compared to baseline values (Fig. 1C). CVP mea-surements in the pneumoperitoneum group ranged from 3.2to 4.0 cmH2O and were not significantly different over time,reflecting a total average of 3.3 ± 0.24 ml of intravenoussaline given during CO measurements in the pneumoperi-toneum group.

Rats demonstrated CO2 retention with increasing pneu-moperitoneum pressure. The arterial pCO2 at 10 mmHgpneumoperitoneum was significantly increased in the pneu-moperitoneum group compared to baseline values and wasdramatically higher than in control animals (Fig. 2A). Ar-terial pCO2 levels did return to baseline values after desuf-flation. Blood pH and BE data mirrored the changes inarterial pCO2 in both groups (Fig. 2B,C). Arterial pO2 levelsdid not change significantly over time and O2 saturation inall rats remained near 100% throughout the experiment(data not shown).

Discussion

The purpose of this study was to compare the cardiorespi-ratory effects of CO2 pneumoperitoneum pressures ranging

Fig. 1. Cardiovascular variables in anesthetized rats subjected to CO2

pneumoperitoneum pressures of 0, 2, 5, and 10 mmHg (n 4 5) or nopneumoperitoneum (Control,n 4 4). The last pressure of 0 mmHg cor-responds to values after desufflation. Individual data points represent themean ± SEM (error bars) for each group. *p < 0.05 compared to controlanimals,[p < 0.05 compared to baseline values at each time point. (a)Heart rate (HR). (b) Mean arterial pressure (MAP). (c) Cardiac output(CO). (–j–, control, –d–, pneumo).

916

from 2 to 10 mmHg in spontaneously breathing anesthetizedrats. The results of this study may assist investigators whouse rodent models of laparoscopic surgery in selecting apneumoperitoneum pressure that allows comparably normalcardiac and respiratory function.

In humans and large animals, studies have reported bothincreases [12] and decreases [13, 15] in cardiac output dur-ing CO2 pneumoperitoneum, depending on the combinedeffects of intraabdominal pressure on preload and afterload[18]. In the present study we measured a slight decrease inCO only at a pneumoperitoneum pressure of 5 mmHg, fol-lowed by a significant increase in CO following desuffla-tion. A decrease in CO at 5 mmHg but not at 10 mmHg ofpneumoperitoneum is probably due to experimental vari-ability, and thus we interpret our data as indicating no im-mediate effect of increasing pneumoperitoneum on CO.These results probably reflect an adequate preload duringthe insufflation as measured by CVP. The significant in-crease in CO following desufflation suggests that animalssubjected to pneumoperitoneum nevertheless incurred a‘‘metabolic debt,’’ perhaps due to the respiratory acidosis.In this regard, it is notable that the anesthetic regimen itselfdid not cause any significant respiratory depression in con-trol animals despite experiment durations of 188.8 ± 14.5min. Our results should be interpreted with some cautionbecause inhalational anesthesia with 1–2% isoflurane in a50% N2O/O2 mixture causes significantly less respiratorydepression than other anesthetics [7, 9].

The respiratory effects of CO2 pneumoperitoneum seenin larger animals and humans are progressive increases inarterial PCO2 and decreases in arterial pH during intraper-itoneal CO2 insufflation at 15 mmHg [6, 10, 13, 14, 17, 18].This hypercarbia usually requires controlled hyperventila-tion to maintain a normal arterial PCO2 under anesthesia.The present study in spontaneously breathing rats demon-strated a similar significant respiratory acidosis during CO2pneumoperitoneum. The marked respiratory acidosis at apneumoperitoneum pressure of 10 mmHg is probably theresult of the inability of spontaneously breathing rodents tocompensate for the combined hydrostatic effects of thepneumoperitoneum and the increased absorbed CO2 load.Although our observations could have been partly due to theeffects of accumulated CO2 over the course of the experi-ment, we believe this is unlikely since the pH and pCO2rapidly returned to baseline values after desufflation.

In previous studies of survival of laparoscopic surgeryin rats we have found that a 2 mmHg CO2 pneumoperito-neum in the rat causes no significant cardiorespiratory al-terations [7] and permits the investigator to perform com-plex intraabdominal procedures [2, 3]. The present studyconfirms the safety of a 2 mmHg pneumoperitoneum whiledemonstrating that pneumoperitoneum pressures of 10mmHg cause significant bradycardia and respiratory acido-sis. These findings are relevant to other laparoscopic re-search studies employing rats or mice where pneumoperi-toneum pressure of 4–6 mmHg [1] and 6–8 mmHg [19] arereported. In one study the authors introduced 6 ml of CO2into the peritoneal cavity of CD-1 mice without reportingthe actual intraperitoneal pressure [21].

In summary, we conclude that CO2 pneumoperitoneumpressures of 10 mmHg or greater should not be used during

Fig. 2. Respiratory variables in anesthetized rats subjected to CO2 pneu-moperitoneum pressures of 0, 2, 5, and 10 mmHg (n 4 5) or no pneumo-peritoneum (Control,n 4 4). The last pressure of 0 mmHg corresponds tovalues after desufflation. Individual data points represent the mean ± SEM(error bars) for each group. *p < 0.05 compared to control animals,[p <0.05 compared to baseline values at each time point. (a) Arterial pCO2. (b)Arterial pH. (c) Arterial base excess (BE). (–j–, control, –d–, pneumo).

917

laparoscopic surgery in the spontaneously breathing rat dueto bradycardia and severe respiratory acidosis.

References

1. Allendorf JD, Bessler M, Kayton ML, Whelan RL, Treat MR, Now-ygrod R (1995) Tumor growth after laparotomy or laparoscopy. Apreliminary study. Surg Endosc 9: 49–52

2. Berguer R, Gutt C, Stiegmann GV (1993) Laparoscopic surgery in therat. Description of a new technique. Surg Endosc 7: 345–347

3. Berguer R, Gutt CN (1994) Laparoscopic colon surgery in a rat model:a preliminary report. Surg Endosc 8: 1195–1197

4. Bouvy ND, Marquet RL, Hamming JF, Jeekel J, Bonjer HG (1996)Laparoscopic surgery in the rat: beneficial effect on body weight andtumor take. Surg Endosc 10: 490–494

5. Crane LH, Martin L (1991) Postcopulatory myometrial activity in therat as seen by video-laparoscopy. Reprod Fertil Dev 3: 685–698

6. Cunningham AJ, Turner J, Rosenbaum S, Rafferty T (1993) Transo-esophageal echocardiographic assessment of haemodynamic functionduring laparoscopic cholecystectomy. Br J Anaesth 70: 621–625

7. Dalton M, Hildreth J, Matsuoka T, Berguer R (1996) Determination ofcardiorespiratory function and the optimum anesthetic regimen duringlaparoscopic surgery in the rat model. Surg Endosc 10: 297–300

8. Filmar S, Gomel V, McComb PF (1987) Operative laparoscopy versusopen abdominal surgery: a comparative study on postoperative adhe-sion formation in the rat model. Fertil Steril 48: 486–489

9. Flecknell PA (1987) Laboratory animal anesthesia. Academic Press,San Diego

10. Graham AJ, Jirsch DW, Barrington KJ, Hayashi AH (1994) Effects ofintraabdominal CO2 insufflation in the piglet. J Pediatr Surg 29: 1276–1280

11. Gutt CN, Berguer R, Stiegmann GV (1993) Laparoscopic surgery inthe rat: description of a new technique. Zentralbl Chir 118: 631–634

12. Hashimoto S, Hashikura Y, Munakata Y, Kawasaki S, Makuuchi M,Hayashi K, Yanagisawa K, Numata M (1993) Changes in the cardio-vascular and respiratory systems during laparoscopic cholecystectomy.J Laparoendosc Surg 3: 535–539

13. Ho HS, Gunther RA, Wolfe BM (1992) Intraperitoneal carbon dioxideinsufflation and cardiopulmonary functions. Laparoscopic cholecys-tectomy in pigs. Arch Surg 127: 928–932; discussion 932–923

14. Ishizaki Y, Bandai Y, Shimomura K, Abe H, Ohtomo Y, Idezuki Y(1993) Changes in splanchnic blood flow and cardiovascular effectsfollowing peritoneal insufflation of carbon dioxide. Surg Endosc 7:420–423

15. Joris JL, Noirot DP, Legrand MJ, Jacquet NJ, Lamy ML (1993) He-modynamic changes during laparoscopic cholecystectomy. AnesthAnalg 76: 1067–1071

16. Kayton M, Morales A, Chen M, Treat M, Nowygrod R (1994) Lapa-roscopic surgery in the rat: a model for teaching laparoscopic suturingtechniques. Surg Endosc 8: 547

17. Odeberg S, Ljungqvist O, Svenberg T, Gannedahl P, Backdahl M, vonRosen A, Sollevi A (1994) Haemodynamic effects of pneumoperito-neum and the influence of posture during anaesthesia for laparoscopicsurgery. Acta Anaesthesiol Scand 38: 276–283

18. Safran DB, Orlando RR (1994) Physiologic effects of pneumoperito-neum. Am J Surg 167: 281–286

19. Trokel M, Allendorf J, Treat M, Whelan R, Nowygrod R, Bessler M(1994) Inflammatory response is better preserved after laparoscopy vslaparotomy. Surg Endosc 8: S30

20. Trokel MJ, Bessler M, Treat MR, Whelan RL, Nowygrod R (1994)Preservation of immune response after laparoscopy. Surg Endosc 8:1385–1387; discussion 1387–1388

21. Watson R, Redmond H, McCarthy J, Burke P, Bouchier-Hayes D(1995) Exposure of the peritoneal cavity to air regulates the earlyinflammatory responses to surgery in a murine model. Br J Surg 82:1060–1065

918

A tactile sensor for laparoscopic cholecystectomy

S. Matsumoto, R. Ooshima, K. Kobayashi, N. Kawabe, T. Shiraishi, Y. Mizuno, H. Suzuki, S. Umemoto

Department of Surgery, Second Teaching Hospital, Fujita Health University, 3-6-10 Otohbashi, Nakagawa-ku, Nagoya, Aichi 454, Japan

Abstract. During laparoscopic surgery, surgeons observethe three-dimensional abdominal cavity on a two-dimensional TV monitor, which is a limitation. Anotherlimitation is that surgeons are unable to estimate the soft-ness of organs or tissues during laparoscopic surgery as theyare only allowed to use instruments which touch objects anddirect palpation is not permitted during the procedure. Thetactile sensor which we used displays the object softnessimmediately as a digital score, which can then be superim-posed on a TV monitor as a graph. With the tactile sensor,we were able to ascertain the presence of a gallstone in thegallbladder or cholecystic duct during laparoscopic chole-cystectomy and also able to discriminate between a stoneand an air bubble during intraoperative cholangiography.We were convinced that the tactile sensor would be usefulin laparoscopic surgery, which does not permit surgeons topalpate objects with human fingers.

Key words: Tactile sensor — Laparoscopic cholecystecto-my — Cholelithiasis — Diagnosis — Ultrasound

Laparoscopic surgery started with laparoscopic cholecystec-tomy, and assessments have since been made of the possi-bility of its application to various ailments typically treatedby surgery. Both laparoscopy and thoracoscopy allow easyobservation and identification of objects as long as they arelocated on surface areas. However, there is the disadvantagethat they cannot identify lesions which do not producechanges on the surface. It is obvious that intraluminal le-sions cannot be observed from outside. If an intraluminalobject is large and sufficiently hard, it may be possible touse a dissector or other instrument to touch and identify itfrom the outside. It is a defect of this method of surgery that,unlike with human hands, one cannot determine with lapa-roscopic graspers or other instruments whether objects aresoft or hard. We were frustrated when we could not be sureof the location of the lesion when performing laparoscopicor thoracoscopic local resection for early gastric, coloniccancer, or pulmonary metastases.

We recently had opportunities to use an ultrasonic tac-tile sensor probe, which has been developed by ProfessorOmata and Olympus Co. [1, 2], as an instrument to detect agallstone in the gallbladder as well as an auxiliary means toconfirm the absence of a stone in the cholecystic duct. Wefelt that this instrument was also supportive for intraopera-tive cholangiography, given our experiences of 20 laparo-scopic cholecystectomy cases.

Case report

A 60-year-old male patient underwent a laparoscopic cholecystectomy.The tactile sensor was used to scan the surface of the gallbladder from thefundus to the cholecystic duct to see whether it could detect a gallstonefrom outside. Three scans were made, and the bottom line shows thetypical curve, which shifts upward in the later scan. This is evidence ofhard material in the gallbladder under the line which the probe scanned(Fig. 1).

Materials and methods

Principles of ultrasonic tactile sensor

The sensor tip contains a piezoelectric transducer (PZT) which consists ofceramics such as lead zirconate itanate and a vibration pickup. When the tipcomes into contact with body tissue, the resonance condition changesaccording to the hardness of the tissue, and the resonance frequency shiftsin a negative direction. It has been proven scientifically that the shiftincreases according to the softness of the tissue. The shift in resonancefrequency can be displayed on a computer monitor as a graph in which thehorizontal axis is time. The quantity of value shift at each point could bemeasured by scanning a probe.

System structure

We have constructed a laparoscopic probe which is 295.5 mm long whosesensor tip is 8 mm in diameter and whose shaft diameter is 10 mm (Figs.2, 3.). The tactile sensor tip is installed in the front of the probe and facesforward. The circuit generating the vibration is contained in the probehandpiece. Power supply to the probe is contained in and calculation of theresonance frequency is made in the measuring section, and the resonancefrequency which is detected according to the hardness of the organic tissueis digitally processed in the frequency counter and sent to a personalcomputer through a universal counter (GP-1B). In the personal computer,records of time changes in digitalized resonance frequency information andCorrespondence to:S. Matsumoto

Surg Endosc (1997) 11: 939–941

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

data are converted into a graph. The graph of resonance frequency data issuperimposed on a laparoscopic monitor image through a scanner con-verter.

Preliminary test and results

In order to check whether the probe was capable of detecting a gallstone,a gallstone was buried in the preliminary test in silicone gum as illustratedin Fig. 4 and the tactile sensor probe was moved as shown in Fig. 5. Theblack material in the photograph shows the gallstone. Figure 5 shows thatthe curved line shifts upward when the probe is passed over the gallstone,and we can thus tell that there is a hard object beneath the sensor.

Discussion

In this study, we confirmed that with the tactile sensor wecould detect hard material in the gallbladder provided thatwe could place the probe on the surface of the gallbladderfrom the fundus to the cystic duct and could scan correctly

Fig. 1. Graph superimposed on the screen during scanning of gallbladder.Thebottom lineindicates the gallbladder softness during scanning from thefundus to the neck. Thecurveshows an upward shift at the right side, andthis indicates a gallstone.

Fig. 2. System structure. The ultrasonic tactile sensor probe is connectedto a personal computer through a universal counter. The result is super-imposed on the monitor through a scan converter.

Fig. 3. Ultrasonic tactile sensor probe.

Fig. 4. A gallstone was embedded in the silicone gum. The black roundmaterial shows a gallstone.

Fig. 5. Model of the ultrasonic tactile sensor probe scan. When the surfaceof the silicone gum is scanned after a gallstone has been embedded in it, thecurved line,which shifts toward a minus direction according to the softnessof silicone gum, shifts upward only when the probe scans over the em-bedded gallstone.

940

over the gallstone. Through experience we learned that astone could be missed if the sensor was not placed over astone when there were only a few small stones in a largegallbladder in which stones could move about during scan-ning. But as the lumen of the cystic duct is narrow and the8-mm probe is wide enough to cover the width of the wholecystic duct lumen, it seems unlikely that we would miss astone impacted in the cystic duct. The tactile sensor thuswas useful when making a decision as to at which point thecystic duct should be cut when there are many stones pass-ing from the gallbladder to the cystic duct. We also shouldbe able to examine stones in the common bile duct and inthe hepatoduodenal ligament with the tactile sensor, whichcan scan the common bile duct in that area.

We often have problems with air bubbles because wecannot distinguish them from stones during intraoperativecholangiography. It is difficult to get good images of dis-sected free cystic duct with the ultrasound diagnostic de-vice. In such cases, it will be unnecessary to repeat cholan-

giography if the tactile sensor has not shown a positivestone reaction. Clinically, this is very useful.

Based on our experiences with a case of cholelithiasis,we think the tactile sensor would be useful for finding alesion which was not exposed on the organ surface—forexample, intraluminal lesions or metastatic lesions in thelung or the liver. The tactile sensor is particularly advanta-geous when locating pulmonary lesions because ultrasounddiagnostic devices are not helpful when detecting lesions.

References

1. Omata S, Terunuma Y (1992) New tactile sensor like the human handand its application. Sensors Actuaters A 35: 9–15

2. Omata S, Yoshida S, Constantinou C, Kayata K, Yamaguchi O, Shi-raiwa Y (1994) New medical sensor for detecting compliance of livingtissue and its applications. Technical Digest of the 12th Sensor Sym-posium, pp 245–248

941

Predictive factors for synchronous common bile duct stones in patientswith cholelithiasis

A. Alponat,1 C. K. Kum, 1 A. Rajnakova,1 B. C. Koh,2 P. M. Y. Goh1

1 Department of Surgery, National University Hospital, 5 Lower Kent Ridge Road, Singapore2 Biostatistic Consultancy Unit, Clinical Research Centre, Singapore

Received: 29 November 1996/Accepted: 9 April 1997

AbstractBackground:To determine the predictive factors of syn-chronous common bile duct (CBD) stones, data from 878consecutive patients who underwent cholecystectomy in auniversity clinic from June 1991 to June 1996 were retro-spectively analyzed.Methods: Based on clinical, biochemical, and ultrasono-graphic criteria, 194 patients were selected for ERCP, 180preoperative and 14 postoperative.Results:Cannulation of CBD was successful in 192 (99%)patients. Stones were identified in 62 (32%) patients andsphincterotomy was performed in 56 (90%). Duct clearancewas achieved in 43 (77%) cases. There was a high predic-tive value for the presence of CBD stones in patients withcholangitis, present jaundice, and dilated CBD with evi-dence of stones on ultrasound (75%, 72%, and 67% respec-tively). A dilated CBD without stone on ultrasound andelevated liver enzymes had less than 40% positive predic-tive value. History of previous jaundice, pancreatitis, pre-viously raised liver enzymes, and present pancreatitis waspredictive in less than 20% of the cases. Univariate analysesrevealed that clinical findings of cholangitis and obstructivejaundice, elevated liver enzymes (previous and present), andultrasonographic findings of stones in a dilated CBD weresignificant positive predictors. Subanalysis of each elevatedliver enzyme revealed that alanine transaminase, aspartatetransaminase, alkaline phosphatase, and gamma glutamyltranspeptidase were significant predictors. Both elevatedconjugated and total bilirubins were also significant predic-tors for CBD stones.Conclusion: Multivariate logistic regression analysis onthese significant predictors showed that cholangitis (oddsratio [OR]: 10.5), dilated CBD with evidence of stones onultrasound (OR: 7.4), elevated aspartate transaminase (OR:2.9), and conjugated bilirubin (OR: 5.3) were jointly sig-nificant. The likelihood of having stones in the duct without

any of these predictors was 7%, but 99% when all the pre-dictors were positive.

Key words: Predictive factors — Synchronous commonbile duct stones — Cholelithiasis

The introduction of endoscopic retrograde cholangiopancre-aticography (ERCP) and sphincterotomy in the early 1970schanged the practice of biliary surgery [4, 13]. Endoscopicsphincterotomy (ES) was first introduced for removal ofrecurrent or retained stones after cholecystectomy [5]. How-ever, its rapid success has expanded the indications. Endo-scopic sphincterotomy with stone extraction became a well-established procedure in the management of choledocholi-thiasis in high-risk elderly patients in whom a noninflamedgallbladder might be left in situ [18]. Patients with cholan-gitis not responding to antibiotics and patients with severegallstone pancreatitis not responding to conservative treat-ment have also become the candidates for ERCP andsphincterotomy during the acute illness [19].

In the last decade, laparoscopic cholecystectomy (LC)has gained widespread acceptance and become the new‘‘gold standard’’ for the treatment of symptomatic gallblad-der disease. Laparoscopic management of associated cho-ledocholithiasis via a transcystic or transcholedochal routehas also been advocated [1]. However, it remains a demand-ing approach that cannot be applied by the majority of sur-geons. Technical difficulties and necessity of advancedskills to manage synchronous CBD stones laparoscopicallyhave fueled the use of preoperative ERCP and sphincterot-omy prior to LC—an approach which fits in well with theminimally invasive concept [14].

ES and duct clearance are successful in 90–95% of pa-tients with an acceptable morbidity and mortality rate [15].Most of the complications are self-limited pancreatitis.However, CBD stones are detected in only 25–50% of theCorrespondence to:P. M. Y. Goh

Surg Endosc (1997) 11: 928–932

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

patients who undergo ERCP. Therefore, knowledge of re-liable factors predictive of synchronous CBD stones is es-sential to avoid unnecessary ERCP and reduce avoidablemorbidity and costs.

The aim of this study was to determine such factors andtheir reliability in predicting CBD stones.

Patients and methods

Between July 1991 and July 1996, 878 consecutive patients underwentcholecystectomy (783 laparoscopic and 95 open), of which 194 were se-lected for either preoperative or postoperative ERCP.

Data collected prior to ERCP included age, sex, history of right hypo-chondrial pain, indication for procedure—previous or present elevated se-rum liver enzymes, clinical findings of cholangitis, jaundice, pancreatitis,dilated CBD over 6 mm with or without stones on ultrasound, serum levelof each liver enzyme (aspartate transaminase [AST], alanine transaminase[ALT], alkaline phosphatase [ALP], gamma glutamyl transpeptidase[GGT], lactate dehydrogenase [LDH])—bilirubins, and ultrasonographicfindings.

Age was categorized as young (<65) and old (65 and above). Bio-chemical analysis of blood was evaluated as abnormal when liver enzymesand bilirubin levels were elevated greater than two times normal. Elevatedliver enzyme tests were considered present when any three of AST, ALT,ALP, GGT, and LDH were elevated.

ERCP was performed with an Olympus JFIT side-viewing duodeno-scope under fluoroscopic control. If sphincterotomy was indicated, the bileduct was cannulated with a papillotome (Wilson-Cook Med., Inc., NC,USA); a combination of coagulation and cutting diathermy was used.Stones were extracted with the Dormia basket or balloon catheter. Chole-cystectomy was done between 24 to 72 h after ERCP.

Statistical analysis

All statistical analyses were performed with the SPSS for Windows version6.0. In the preliminary analysis, univariate statistical methods were used todetermine which factors were significantly related to the presence of CBDstones. The chi-square test with Yates’ correction or Fisher’s exact test wasused for categorical variables;p < 0.05 was considered significant.

All significant factors from the univariate analysis were subsequentlyincluded in the multivariate logistic regression procedure. The FORWARDautomatic variable selection procedure was used in determining which ofthese variables were predictors of presence of CBD stones. Presence ofCBD stones on ERCP was the dependent variable. Four of 194 patientswith missing data were excluded from the analysis. The missing data wasultrasonographic evaluation of CBD in two and ERCP findings due tofailure in cannulation in two.

Results

There were 73 males (38%) and 121 females (62%). Theaverage age was 52 (range 13–86) years. Cannulation ofCBD was successful in 192 cases (99%). Cannulation fail-ure occurred in two patients (1%) in whom the ampulla waslocated in a large diverticulum; they were excluded from thestudy.

Some 180 patients (93%) had preoperative ERCP and14 (7%) had postoperative ERCP. The indication for post-operative ERCP was obstructive jaundice in six patients—retained stones were found in all of them. Elevated liverenzymes and serum amylase levels were the indications inanother six patients—ERCP findings were normal in all. Inthe remaining two patients, incomplete clearance of CBD inopen exploration required postoperative ERCP. Stone ex-traction was successful in these cases.

Stones were found in 62 cases out of 192 (32%). Sixpatients with multiple large stones (>15 mm) in the CBDwere referred to surgery for open CBD exploration withoutendoscopic sphincterotomy attempt during our early expe-rience with interventional ERCP. Endoscopic sphincteroto-my was performed in 56 patients (90%). Both wire-guidedcannulation and pre-cut sphincterotomy techniques wereemployed to improve success rate. Stone retrieval was suc-cessful in 43 (77%) cases. Failure of stone extraction oc-curred in eight patients (14%) because of the large size ofstones (mean: 18.3 mm, ranging from 15 to 23 mm). Theseeight patients underwent open exploration of CBD. Twopatients who had single small stones (<5 mm) had onlysphincterotomy and subsequently underwent LC. Three pa-tients who had septic cholangitis underwent sphincteroto-my, one of which also required stenting. The mean durationfor diagnostic ERCP was 25 (SD: 16 min) and 42 (SD: 21.5min) when sphincterotomy and stone extraction were per-formed.

Indications for ERCP and positive predictive values(PPV) for these predictors are shown in Table 1.

Complications included pancreatitis in seven cases(4%), all of which resolved spontaneously. Pancreatitis oc-curred following sphincterotomy in two and diagnosticERCP in five. One patient (0.5%) had a retained stone bas-ket which required surgery and duodenotomy. Two patients(1%) had minor bleeding that stopped spontaneously with-out the need for blood transfusion.

Univariate analysis of each set of pre-op ERCP data wascarried out to assess its significance in predicting the pres-ence of CBD stones (Table 2). Cholangitis, jaundice, pastand present elevated liver enzymes, and ultrasonographicfinding of CBD stones were predictors on univariate analy-sis even after using the Bonferroni-Holm method to adjustfor multiple comparisons. Analysis of individual enzymesrevealed that elevated AST, ALT, ALP, GGT, and raisedbilirubins (conjugated and total) were significant predictors.

Multiple logistic regression analysis on these 11 vari-ables showed that only cholangitis, dilated CBD over 6 mmwith stone(s) on ultrasound, elevated AST, and conjugatedbilirubin were jointly significant (Table 3). The probabilityof having stones in CBD is thus: log odds ratio (OR)4−2.62 + (2.26 × cholangitis) + (2 × dilated CBD with stone

Table 1. Indications for endoscopic retrograde cholangiopancreaticogra-phy and positive predictive values

Indication

Number ofpatients (%),n 4 194

Positivepredictivevalue (%)

Previous LFTa elevation 31 (16) 16Previous pancreatitis 7 (4) 14Previous jaundice 8 (4) 13Present LFTa elevation 139 (72) 37Present pancreatitis 31 (16) 19Present jaundice 52 (27) 67Cholangitisb 8 (4) 75Dilated CBDc without stone on ultrasound 53 (27) 36Dilated CBDc with stone on ultrasound 36 (19) 72

a Liver function testb Jaundice with fever, right hypochondrial painc Common bile duct

929

on ultrasound) + (1.08 × elevated AST) + (1.66 × elevatedconjugated bilirubin), in which, cholangitis4 1 if presentor 0 if not; dilated CBD over 6 mm with stone on ultrasound4 1 if yes or 0 if not; elevated AST4 1 if yes or 0 if not;and elevated conjugated bilirubin4 1 if present and 0 ifnot.

Discussion

CBD stones are present in 10–15% of patients undergoingcholecystectomy and the incidence rises to 30% in thoseover 80 years of age [6]. In the prelaparoscopic cholecys-tectomy era, the traditional management for CBD stoneswas to perform a choledochotomy and stone extraction.However, open CBD exploration is associated with signifi-cant morbidity and retained stone rate [5, 18]. Preoperative

ERCP and sphincterotomy were introduced to avoid thecomplications associated with open CBD exploration [11].To date, the endoscopic approach has never been convinc-ingly proven to be superior to open exploration [19, 22].

After widespread acceptance of LC, the management ofCBD stones has regained renewed interest. Several strate-gies have emerged to manage synchronous CBD stones. Thefirst strategy is to ignore them. But it has been shown that55% of untreated CBD stones became symptomatic, andhalf had complications [12].

The second option is routine preoperative cholangiog-raphy and sphincterotomy if stones are present in a dilatedduct. However, this approach is not cost-effective and rou-tine ERCPs are not without danger. ERCP has a morbidityrate of 3–6.4% and a mortality rate of 0.05–0.1%, increasingto a morbidity rate of 5–10% and a mortality rate of 1–3.1%when sphincterotomy is performed [6]. In addition, it is

Table 2. Analysis of preoperative parameters for statistical significance by using chi-square or Fisher’s exact test*

Predictors

CBDa stoneSignificance

Absent(n 4 130)b

Present(n 4 62) x2 p

Age (years)>65<65

41 (32)89 (68)

29 (47)33 (53) 3.574 0.06

Sex (F percent) 83 (63) 38 (61) 0.002 0.93Past history

Right upper quadrant pain 130 (100) 61 (98) 0.54*Right upper quadrant pain radiating back 60 (46) 23 (37) 1.058 0.19

IndicationCholangitis 2 (2) 6 (10) 0.02*Resolved pancreatitis 6 (5) 1 (2) 0.43*Previous jaundice 7 (5) 1 (2) 0.44*Previously elevated liver enzymes 26 (20) 5 (8) 0.04*Present elevation of liver enzymes 88 (67) 52 (84) 0.02*Present pancreatitis 25 (19) 6 (10) 0.14*Present jaundice 17 (13) 35 (56) 37.826 0.01CBD >6 mm without stone on ultrasound 34 (26) 19 (31) 0.331 0.56CBD >6 mm with stone on ultrasound 10 (8) 26 (42) 31.189 0.01

Laboratory investAspartate transaminase 78 (60) 55 (89) 0.01*Alanine transaminase 80 (62) 54 (87) 0.03*Alkaline phosphatase 70 (54) 52 (84) 0.01*Gamma glutamyl transpeptidase 102 (78) 54 (87) 0.05*Lactate dehydrogenase 68 (54) 42 (68) 3.481 0.06Conjugated bilirubin 28 (22) 42 (71) 41.678 0.01Total bilirubin 22 (17) 38 (61) 36.425 0.01

Ultrasonography findingsGB stone41 11 (9) 6 (11)

>1 111 (87) 52 (90) 0.78*GB st. size <1 cm 102 (84) 42 (81)

>1 cm 19 (16) 10 (19) 0.66*

a CBD, common bile ductb Values in parantheses are percentages

Table 3. Predicted odds ratio of having stones in the common bile duct based on the multivariate logistic regression model

Predictor Odds ratio (OR) 95% confidence interval p value

Cholangitis 10.5 1.55 71.79 0.02Dilated common bile duct with stone on ultrasound 7.4 2.85 18.99 0.01Elevated aspartate transaminase 2.9 1.25 6.88 0.01Elevated conjugated bilirubin 5.3 2.35 11.83 0.01

930

expensive [7]. Therefore, routine use of preoperative ERCPcannot be recommended in all patients undergoing chole-cystectomy.

The third strategy is to perform intraoperative cholan-giogram and retrieve CBD stones laparoscopically. Intraop-erative cholangiography is successful in 88% of patientswith a positive predictive value of 63–92% [16]. This pro-cedure also delineates the biliary anatomy and may preventbile duct injury. Furthermore, complete treatment of biliarylithiasis in one operation may avoid the joint risk of twoprocedures (ERCP and LC) [17]. Although routine use ofintraoperative cholangiography adds about 20 min to theLC, fluoroscopic cholangiogram can be completed in anaverage time of 6.9 min [2]. However, it has a false-positiverate ranging from 2.1 to 67%, which leads to unnecessaryCBD exploration or conversion to open surgery [16].

The fourth strategy is preoperative ERCP in selectedcases. Preoperative ERCP reduces the need for intraopera-tive cholangiography, avoids the need for most postlaparo-scopic studies, and provides important information for thebile duct anatomy prior to LC. However, the significantdifference in developing recurrent biliary symptoms follow-ing sphincterotomy (21%) compared to the biliary surgicalgroup (6%) and high cost of ERCP should be kept in mind[8, 24].

Experience with selective preoperative ERCP andsphincterotomy prior to LC has been accumulating [3, 8–10,20, 26]. Clinical, biochemical, and ultrasonographic evi-dence of stones has been utilized as criteria in patient se-lection for ERCP. Some studies reported that only the se-verity of patient’s initial illness (acute cholangitis, persistentobstructive jaundice, and acute gallstone pancreatitis) hadhigher than 85% positive predictive value [21, 23, 25]. El-evated liver enzymes correlated well with duct calculi witha positive predictive value of 30–60% in other studies [9,11]. Ultrasonographic finding of a dilated duct was predic-tive in 20–70% of the cases, whereas evidence of stones ina dilated duct on ultrasound was almost uniformly accurate[3, 20, 26].

Multivariate analysis in this study has identified chol-angitis, dilated CBD over 6 mm with stone(s) on ultrasound,

elevated AST, and conjugated bilirubin as the main predic-tive parameters.

A combination of predictors increases the odds of hav-ing stones in the CBD. By applying the derived formula,likelihood of having stones in the duct can be estimated;thereby, the number of the unnecessary ERCPs can be re-duced. According to the formula, probability of havingstones in the duct without any of these predictors is 7%whereas it increases to 99% if all four predictors are present(Table 4).

This study has identified important predictors of syn-chronous CBD stones in patients who require cholecystec-tomy. Selective use of ERCP in these patients will ensure ahigh yield rate and thus improve cost-effectiveness. Thesmall number of patients who are ‘‘false negatives’’, i.e.,have associated CBD stones but also do not have any ab-normal parameters, usually have small stones that are likelyto pass spontaneously. If not, postoperative ERCP can dealwith most of them.

References

1. Arregui ME, Davis CJ, Arkush AM, Nagan RF (1992) Laparoscopiccholecystectomy combined with endoscopic sphincterotomy and stoneextraction or laparoscopic choledochoscopy and electrohydrauliclithotripsy for management of cholelithiasis with choledocholithiasis.Surg Endosc 6: 10–15

2. Carroll BJ, Phillips EH, Rosenthal R, Gleischman J, Bray JF (1996)One hundred consecutive laparoscopic cholangiography. Results andconclusions. Surg Endosc 13(3): 319–323

3. Cisek PL, Greaney GC (1994) The role of endoscopic retrograde chol-angiopancreaticography with laparoscopic cholecystectomy in themanagement of choledocholithiasis. Am Surg 60: 772–776

4. Classen M, Demling L (1974) Endoskopische sphinkterotomie derpapilla Vateri und steinextraktion aus dem ductus choledochus. DischMed Worchensahr 99: 496–497

5. Danilewitz MD (1989) Early postoperative endoscopic sphincteroto-my for retained common bile duct stones. Gastrointest Endosc 35:298–299

6. Fletcher DR (1994) Changes in the practice of biliary surgery andERCP during the introduction of laparoscopic cholecystectomy toAustralia: their possible significance. Aust N Z J Surg 64: 75–80

7. Frazee RC, Roberts J, Symmonds R, Hendricks JC, Snyder S, Smith R,Custer MD, Stoltenberg P, Avots A (1993) Combined laparoscopic

Table 4. Predicted probability of having stones in CBD based on the pattern of predictors present in any individual patient

Model CholangitisDilated CBD withstone on ultrasounda

Elevated aspartatetransaminase

Elevated conjugatedbilirubin

Predictedprobability (%)

1 No No No No 72 Yes No No No 433 No Yes No No 354 No No Yes No 185 No No No Yes 286 Yes Yes No No 857 Yes No Yes No 698 Yes No No Yes 809 No Yes Yes No 61

10 No Yes No Yes 7411 No No Yes Yes 5312 Yes Yes Yes No 9413 Yes Yes No Yes 9714 Yes No Yes Yes 9215 No Yes Yes Yes 8916 Yes Yes Yes Yes 99

a CBD, common bile duct

931

and endoscopic management of cholelithiasis and choledocholithiasis.Am J Surg 166: 702–705

8. Graham SM, Flowers JL, Scott TR, Biley RW, Scovill WA, ZuckerKA, Imbembo AL (1993) Laparoscopic cholecystectomy and commonbile duct stones. Ann Surg 216: 61–67

9. Hainsworth PJ, Rhodes M, Gompertz RHK, Armstrong CP, LennardTWJ (1994) Imaging of common bile duct in patients undergoinglaparoscopic cholecystectomy. Gut 35: 991–995

10. Hawasli A, Llyoyd L, Pozios V, Veneri R (1993) The role of endo-scopic retrograde cholangio-pancreaticogram in laparoscopic chole-cystectomy. Am Surg 59: 285–289

11. Heinerman MP, Boeckl O, Pimpl W (1988) Selective ERCP and pre-operative stone removal bile duct surgery. Ann Surg 209: 267–272

12. Johnson AG, Hosking SW (1987) Appraisal of the management of bileduct stones. Br J Surg 74: 555–560

13. Kawai K, Akasaka Y, Murakami M (1974) Endoscopic sphincteroto-my of the ampulla of Vater. Gastrointest Endosc 20: 148–151

14. Kum CK, Goh PMY (1996) Preoperative ERCP in the management ofcommon bile duct stones before laparoscopic cholecystectomy. Eur JSurg 162: 205–210

15. Leese T, Neoptolemos JP, Carr-Locke DL (1985) Successes, failures,early complications and their management following endoscopicsphincterotomy: results in 394 consecutive patients from a single cen-tre. Br J Surg 72: 215–219

16. Madhavan KK, Macintyre IMC, Wilson RG, Saunders JH, Nixon SJ,Hamer-Hodges DW (1985) Role of intraoperative cholangiography inlaparoscopic cholecystectomy. Br J Surg 82: 249–252

17. Millat D, Deleuze A, Atger J, Briandet H, Fingerhut A, Marrel E,de-Seugin C, Soulier P (1996) Treatment of common bile duct lithiasisunder laparoscopy. A prospective multicenter study in 189 patients.Gastroenterol Clin Biol 20 (4): 339–345

18. Neoptolemos JP, Carr-Locke DL, Fossard DP (1987) Prospective ran-

domised study of preoperative endoscopic sphincterotomy versus sur-gery alone for common bile duct stones. Br Med J 294: 470–474

19. Neoptolemos JP, London N, Slater ND, Carr-Locke DL, Fossard DP,Moosa AR (1986) A prospective study of ERCP and endoscopicsphincterotomy in the diagnosis and treatment of gallstone acute pan-creatitis. Arch Surg 121: 697–702

20. Rieger R, Wayand W (1995) Yield of prospective, noninvasive evalu-ation of the common bile duct combined with selective ERCP/sphincterotomy in 1390 consecutive laparoscopic cholecystectomy pa-tients. Gastrointest Endosc 42: 6–12

21. Rijna H, Borgstein PJ, Meuwissen SGM, De Brauw LM, WildenborgNP, Cuesta MA (1995) Selective preoperative endoscopic retrogradecholangiopancreaticography in laparoscopic biliary surgery. Br J Surg82: 1130–1133

22. Stiegmann GV, Goff JS, Mansour A, Pearlman N, Reveille RM, Nor-ton L (1992) Precholecystectomy endoscopic cholangiography andstone removal is not superior to cholecystectomy, cholangiography,and common duct exploration. Am J Surg 163: 227–230

23. Surick B, Washington M, Ghazi A (1993) Endoscopic retrograde chol-angiography in conjunction with laparoscopic cholecystectomy. SurgEndosc 7: 388–392

24. Targarona EM, Ayuso RM, Bordes JM, Ros E, Pros I, Martinez J,Teres J, Trias M (1996) Randomized trial of endoscopic sphincterot-omy with gallbladder left in situ versus open surgery for common bileduct in high risk patients. Lancet 347 (9006): 926–929

25. Vanneman W, Kingsbury R, Duberman E, Lee M (1992) When isERCP indicated before laparoscopic cholecystectomy. GastrointestEndosc 38: 265

26. Vitale GC, Larson GM, Wieman TJ, Cheadle WG, Miller FB (1993)The use of ERCP in the management of common bile duct stones inpatients undergoing laparoscopic cholecystectomy. Surg Endosc 7:9–11

932

Impact of laparoscopic cholecystectomy on indications for surgicaltreatment of gallstones

E. W. Steinle,1 R. L. VanderMolen,1 A. Silbergleit,1,2 M. M. Cohen2,3

1 Department of Surgery, St. Joseph Mercy Hospital, Pontiac, MI, USA2 Wayne State University, Detroit, MI, USA3 Rose Medical Center, Denver, CO, USA

Received: 25 September 1996/Accepted: 14 March 1997

AbstractBackground:The objective of this study was to compare thehistology of gallbladders removed prior to the introductionof laparoscopic cholecystectomy with that found after theintroduction of the laparoscopic technique to determine ifthere has been a change in the indications for surgical treat-ment of gallbladder disease.Methods:A retrospective review of all patients undergoingcholecystectomy during 1989, 1992, and 1993 was com-pleted at two large community teaching hospitals in twodifferent geographic regions of the United States. Patientswho underwent cholecystectomy as the primary procedurewere studied. A total of 1,815 cases met the criteria foranalysis. Histological diagnoses were categorized as acutecholecystitis with or without cholelithiasis, or chronic cho-lecystitis with cholelithiasis.Results:The number of cholecystectomies performed in-creased by 58% from 1989 to 1993 (p < 0.05). The numberof cholecystectomies for acute cholecystitis did not change.Conclusions:With the advent of laparoscopic cholecystec-tomy, the number of cholecystectomies significantly in-creased and the proportion of cholecystectomies performedfor chronic disease also increased. There has been a signifi-cant change in the surgical management of gallbladder dis-ease with increased willingness to recommend elective cho-lecystectomy. Further study is needed to determine if thereis real benefit from earlier elective cholecystectomy.

Key words: Cholecystectomy — Cholecystitis — Laparos-copy — Surgery rate surgical indications

Laparoscopic cholecystectomy accounts for 80–83% of thecholecystectomies performed nation wide [1], and since itsintroduction in 1987 [5] its use has been accompanied by amajor increase in the cholecystectomy rate [8, 11, 15]. Theprocedure has been shown to have a low morbidity andmortality rate [1, 2, 11, 15, 16]. Laparoscopic cholecystec-tomy has also reduced the total cost per case of cholecys-tectomy by decreasing the hospital stay [9]. The NationalInstitutes of Health (NIH) consensus conference recom-mended laparoscopic cholecystectomy as the treatment ofchoice for symptomatic gallstones [10]. The NIH also con-cluded that ‘‘the indications for laparoscopic cholecystec-tomy, in general, are similar to those for open cholecystec-tomy. Indeed, the availability of laparoscopic cholecystec-tomy should not expand the indications for gallbladderremoval’’ [10]. Despite this warning, published reports haveshown large increases (29–60%) [8, 11, 15] in the totalnumber of cholecystectomies since the introduction of thelaparoscopic technique compared to the cholecystectomyrate in the 1980s [3]. These data have led some to speculatethat the indications for cholecystectomy have changed,leading not only to an increase in the total number of pro-cedures but also in the total cost of treatment of gallbladderdisease [4, 8]. This study was designed to determine if theintroduction of laparoscopic cholecystectomy has led to achange in the indications for cholecystectomy.

Patients and methods

During the years 1989, 1992, and 1993, a total of 1,815 patients underwentcholecystectomy as the primary procedure at our two hospitals. Both hos-pitals are large community teaching hospitals, one (St. J) is in Pontiac,Michigan, the other (RMC) is in Denver, Colorado. Patients having inci-dental cholecystectomy, or cholecystectomy for chronic acalculous chole-cystitis or trauma were excluded. Cases of carcinoma of the gallbladderwere also excluded. The average age of the patients was 51 and 52 yearsold, respectively, for 1989 and 1992. The percentage of patients who werefemale was 70% in 1989 and 71% in 1992.

Permanent sections from each case were reviewed by attending pa-thologists. Standard criteria for acute and chronic cholecystitis were used.

Correspondence to:M. M. Cohen, Department of Surgery, Grace Hospital,6071 W. Outer Drive, Detroit, MI 48235, USA

Surg Endosc (1997) 11: 933–935

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Gallbladders were designated as acute cholecystitis with or without cho-lelithiasis or chronic cholecystitis with cholelithiasis. Results were ana-lyzed using the chi-squared test.

Results

Figure 1 shows the total number of cholecystectomies per-formed at the two institutions during the years 1985 through1993. The number of patients was relatively stable over the5 years prior to the introduction of the laparoscopic tech-nique; 1989 was the last year in which no laparoscopiccholecystectomies were performed. By 1991, all generalsurgeons with privileges at the two hospitals had obtainedthe required credentials to perform laparoscopic cholecys-tectomy. The average number of cholecystectomies per-formed from 1985 to 1989 was 429 per year. In 1993 therewere 679, a 58% increase (p < 0.05). Throughout the period1985 to 1993, the hospitals remained the same size, thepopulations for which they provided service were stable,and the general surgical staffs were not significantly ex-panded. This stability of the population being served wasconfirmed by demonstrating that the total number of appen-dectomies (considered a good indicator of the need for sur-gical care) did not change from 1989 to 1992. In 1989 a totalof 362 appendectomies were performed at the two hospitalswhereas in 1991 a total of 350 were performed. It is alsointeresting to note the remarkable penetration of the lapa-roscopic technique at both hospitals within about 2 years(Fig. 2). No laparoscopic procedures were performed in1989, yet by 1993, 86% of the cholecystectomies were per-formed laparoscopically. The ratio of acute to chronic dis-ease of the gallbladders during these years also changedfrom 0.17 in 1989 to 0.1 in 1992 and 0.12 in 1993. As canbe seen in Fig. 3, while the total number of patients withacute disease remained constant, the entire increase in cho-lecystectomies was due to a large and significant increase inthe number of cases with chronic disease. Chronic chole-cystitis accounted for only 338 cases in 1989 compared to546 cases in 1993. This represents an increase of 62%. In1992, 22% of the open cholecystectomies were performedfor acute cholecystitis and by 1993 this had increased to33%.

Discussion

In the early 1990s laparoscopy produced a remarkablechange in the technique of cholecystectomy. This has beenaccompanied by improved patient recovery time, decreasein cost per procedure, and an acceptable morbidity rate.Recently, several authors have noted that along with thesechanges, there has also been a substantial increase in thetotal number of patients undergoing cholecystectomy [8, 11,15]. We have seen a similar increase in the cholecystectomyrate at both of our institutions. However, the absolute num-ber of cases of acute cholecystitis has remained unchanged.The observed increase was entirely among patients withchronic cholecystitis and cholelithiasis. There is a largepopulation of patients with cholelithiasis who are eitherasymptomatic or who never opt for surgical treatment [13].It seems clear that the recent increase in the cholecystecto-my rate comes primarily from the group of patients withmild or asymptomatic disease. There were probably initiallysome patients with moderate to severe symptoms who weremore willing to undergo a laparoscopic than an open opera-tion. But the higher cholecystectomy rate has been main-

Fig. 3. The number of gallbladders showing acute or chronic disease dur-ing the years 1989, 1992, and 1993.

Fig. 1. The number of cholecystectomies performed as the primary pro-cedure during the years 1985 to 1993.

Fig. 2. The number of open and laparoscopic (Lap) cholecystectomiesperformed each year since the introduction of laparoscopic cholecystecto-my at each hospital during 1990.

934

tained even after that pool of patients has been treated.Escarce et al. [6] have produced data from Medicare pa-tients in Pennsylvania over the period 1986 to 1993 whichstrongly suggests that a 22% increase in the rate of chole-cystectomy in that state was accompanied by a lowering ofthe threshold for performing surgery. Lowering the thresh-old for cholecystectomy is not necessarily inappropriate ifthe patients benefit overall. But in mildly symptomatic orasymptomatic patients, for benefit to outweigh risk the op-eration must be extremely safe, as serious complicationsoccurring in the course of the natural history are rare in suchindividuals [14].

Apparently physicians, both primary care and specialist,are now more likely to recommend laparoscopic cholecys-tectomy than previously. This is probably due to the rapidrecovery and the avoidance of a large abdominal incision.This trend has already eliminated the anticipated cost sav-ings of laparoscopic cholecystectomy for the payors [4]. Bycontrast, Ho et al. [7] in a study of cholecystectomy from1988 to 1994 reported that their cholecystectomy rate hadstabilized at a level only 11% higher than in the years pre-ceding the introduction of laparoscopic cholecystectomy.This finding could be explained by the fact that theirs is atertiary referral center and may not reflect what is happen-ing in the country as a whole. As stated in the NIH consen-sus conference, the challenge to the clinician is ascertainingwhich symptoms are and which are not due to gallstones[10]. A risk of operating on patients with stones whosesymptoms are not caused by the stones is that the symptomswill frequently recur. As has recently been observed [12], ifwe broaden the indications for cholecystectomy to includepatients with minimal or even no symptoms, the postchole-cystectomy syndromes will become of much greater clinicalrelevance. More importantly, the small but real risk of bileduct injury and its sequelae must be weighed very carefullyagainst the potential benefit for a patient with minimal orequivocal symptoms. It is perhaps too easy to recommend aminimally invasive operation which can be done as an am-bulatory procedure. A cost/benefit analysis is needed to de-termine if the US population has been well served by the

introduction of laparoscopic cholecystectomy and by thesignificant increase in the cholecystectomy rate that this hasproduced.

References

1. Baird D, Wilson J, Mason E, Duncan TD, Evans JS, Luke JP, RubenDM, Lucas GW (1992) An early review of 800 laparoscopic chole-cystectomies at a university-affiliated community teaching hospital.Am Surg 58: 206–210

2. Cohen MM (1992) Initial experience with laparoscopic cholecystec-tomy in a teaching hospital. Can J Surg 35: 59–63

3. Diehl A (1987) Trends in cholecystectomy rates in the United States.Lancet 2: 683

4. Diehl A (1993) Laparoscopic cholecystectomy: too much of a goodthing? J Am Med Assoc 270: 1469–1470

5. Dubois F, Icard P, Berthelot G, Levard H (1990) Coelioscopic chole-cystectomy. Ann Surg 211: 60–62

6. Escarce JJ, Chen W, Schwartz JS (1995) Falling cholecystectomythresholds since the introduction of laparoscopic cholecystectomy. JAm Med Assoc 273: 1581–1585

7. Ho HS, Mathiesen KA, Wolfe BM (1996) The impact of laparoscopiccholecystectomy on the treatment of symptomatic cholelithiasis. SurgEndosc 10: 746–750

8. Legorreta A, Silber J, Constantino G, Kobylinski R, Zata S (1993)Increased cholecystectomy rate after the introduction of laparoscopiccholecystectomy. J Am Med Assoc 270: 1429–1432

9. McIntyre RC, Zoeter MA, Weil KC, Cohen MM (1992) A comparisonof outcome and cost of open vs laparoscopic cholecystectomy. J Lapa-roendosc Surg 2: 143–148

10. NIH consensus development panel on gallstones and laparoscopic cho-lecystectomy (1993) Gallstones and laparoscopic cholecystectomy. JAm Med Assoc 269: 1018–1024

11. Orlando R, Russell J, Lynch J, Mattie A (1993) Laparoscopic chole-cystectomy: a statewide experience. Arch Surg 128: 494–499

12. Phillips SF (1996) Diarrhea after cholecystectomy: if so, why? Gas-troenterology 111: 816–818

13. Ransohoff D, Gracie W (1990) Management of patients with symp-tomatic gallstones: a quantitative analysis. Am J Med 88: 154–160

14. Ransohoff DF, Gracie WA (1993) Treatment of gallstones. Ann InternMed 119: 606–619

15. Steiner CA, Bass EB, Talamini MA, Pitt HA, Steinberg EP (1994)Surgical rates and operative mortality for open and laparoscopic cho-lecystectomy in Maryland. N Engl J Med 330: 403–408

16. The Southern Surgeons Club (1991) A prospective analysis of 1518laparoscopic cholecystectomies. N Engl J Med 324: 1073–1078

935

Comparative retrospective study of surgical treatment ofspontaneous pneumothorax

Thoracotomy vs thoracoscopy

R. Jimenez-Merchan, F. Garcıa-Dıaz, C. Arenas-Linares, J. C. Giron-Arjona, M. Congregado-Loscertales,J. Loscertales

Department of General and Thoracic Surgery, Hospital Universitario de Sevilla, Avda. Dr. Fedriani s/n, 41071 Seville, Spain

Received: 23 September 1996/Accepted: 28 February 1997

AbstractBackground:Since May 1992 we have used video-assistedthoracic surgery (VATS) for the treatment of the spontane-ous pneumothorax. In this study we review the results ob-tained with this technique and we compare them with thoseobtained by conventional surgery (1976–1992).Methods:110 patients (98 males and 12 females, age 15–83years) were operated by VATS, and 627 patients (567 malesand 60 females, age 14–89 years) by conventional surgery.The surgical technique and the complications are compared.No one died.Results:Although the number of complications was greaterin the first group than in the second one (29.5% vs 15.1%),this is attributed to the lack of experience among surgeonsin the first cases; which complications nearly disappeared inthe last 60 patients. Less pain, better recovery, and shorterhospital stay resulted.Conclusions:For us VATS is the ideal technique withwhich to treat spontaneous pneumothorax.

Key words: Spontaneous pneumothorax — VATS — Opensurgery

Spontaneous pneumothorax is one of the afflictions mostcommonly treated by thoracic surgeons. The aim of treat-ment is to completely reexpand the lung and avoid recur-rence. Until a shorttime ago, there were two options: place-ment of endopleural drainage or surgical treatment by tho-racotomy. With the former, a failure and recurrence rate thatsurpasses the 30% has been reported [2, 9, 15]. The surgicaltreatment tends to thoroughly solve this problem; recurrencedoes not surpass 1% [11, 17]. Nevertheless, despite being

only a medium-risk intervention with low morbidity andgenerally no mortality, many patients complain of consid-erable pain in the postoperative period that tends to take along time to disappear [6]. With both options, the loss ofworking days is important.

Since the introduction of videothoracoscopy and video-assisted thoracic surgery (VATS), the problems inherent tothis type of surgery have been notably reduced. Postthora-cotomy pain is minimized by using only three workingports, and this results in a greater ability of the patient toperform respiratory physiotherapy. Benefits are fewer pleu-ropulmonary problems and shorter postoperative hospitalstay. All this contributes to a more rapid recovery and anearlier return to work [14, 18]. However, long-term resultsof the VATS approach are not well known.

From the technical point of view, blebs and bullae areeasily identifiable with this approach; sometimes vision isbetter than in open surgery. But blebs and bullae do notgenerate problems when removed with the help of the newendostapler devices [3], either, or in the realization of thepleural abrasion [5] or pleurectomy [4].

The aim of this study is to revise the results obtained inour series of patients operated by VATS and to comparethem to a previous series of patients operated by thoracot-omy.

Materials and methods

Until May 1992 in our department 627 patients (567 men and 60 women)had been operated on for spontaneous pneumothorax using conventionalopen surgery. The average age of these patients was 31 years (range, 14 to89 years). In 324 cases the pneumothorax was located in the right side, in262 in the left side; 11 were bilateral synchronous and 30 bilateral meta-chronous. Though we have used several approaches (Table 1), the mostcommonly used one has been Noirclerc’s lateral thoracotomy [13]. Thisthoracotomy allows entry into the thorax with only section of the inter-costal muscle, since the opening of the serratus is made without sectioningit, only opening its fibers in their own course.Correspondence to:J. Loscertales

Surg Endosc (1997) 11: 919–922

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Since May 1992, VATS has been used in the treatment of spontaneouspneumothorax in our department: 110 patients (98 men and 12 women)with an average age of 29.5 years (range, 15 to 83 years) have beenoperated. On 55 occasions the pneumothorax was located on the right side(50%) and in 49 on the left hemithorax (44.6%); in four cases it wasbilateral metachronous (3.6%) and in two bilateral synchronous (1.8%).There are no significant differences regarding the age of the patients andthe location of the pneumothorax between groups.

We use the following technique for videothoracoscopy: patient on lat-eral decubitus opposite to the affected side, under general anesthesia, withdouble lumen tracheal tube. During the operation the lung collapse ismaintained through the occlusion of one of the lumens of the double lumentube. On certain occasions the anesthetist is asked to insufflate the lung inorder to observe small blebs/bullae in a better way, and at the end to checkif there are any air leaks. Three ports are opened with 12-mm trocars: thefirst one, in seventh–eighth intercostal space, middle axillary line, for thecamera (0° straight vision), using two high-resolution monitors; thereaftertwo other ports are opened at the level of the third–fourth space, anterioraxillary line, and fifth–sixth space under the tip of the scapula, throughwhich the working instruments are introduced. Once the adherences of thelung to the wall, if any, are freed, the bullae or blebs are located, and theyare extirpated with the endostapler device Endo-Gia. (Between one andeight loaders, an average of three, have been used.) The change of positionof the camera in the anterior and/or posterior port must be always accom-plished, since bullae/blebs can be found in the posterior segment, and thesebullae cannot be properly visualised with the camera in its initial position.After that, a plastic sponge with its size adapted to the ports is introducedto accomplish the abrasion of the parietal pleura, and thereafter we instil-late 150–200 ml of 0.5% iodine alcohol solution to bath the whole visceralpleura. The possible air leaks are proven with this liquid, after asking theanesthetist to insufflate the lung, if there is any possible doubt about theline of suture. The action of the sponge in the pleurodesis is reinforced dueto the irritation produced by the iodine. This is an original procedure ofours [8]. Upon ending the intervention, through the inferior port we intro-duce a 28 F drainage tube connected to the aspiration system with a waterseal, and another small-caliber drainage-type Pleurocath, through the an-terosuperior port, checking its location and the total lung reexpansion withthe video camera. Normally, the fine drainage is withdrawn in 24–48 h,while the tube in the inferior location is withdrawn in 48–72 h if there isno air leak or important pleural effusion. The patient leaves the hospital the1st day after the second tube removal, sending the patient to the Rehabili-tation Department in order to continue with the assisted respiratory phys-iotherapy over 20–30 days.

The technique we have used in conventional surgery is similar to theone used here with VATS: removal of the lesion with staplers, also fol-lowed by parietal pleural abrasion with plastic sponge and bath of thevisceral pleural with iodine solution [6].

After leaving the hospital, the patients are followed up in the policlinic.The first visit is at 1 month; visits then occur every 3 months during the 1styear and every 6 months until 5 years. The patients are annually followedup thereafter.

Results

The pathologic findings both in open surgery and in thepatients operated by videothoracoscopy are presented inTables 2 and 3, respectively. There are no significant dif-ferences.

The complications that appeared in the postoperativeperiod of the patients operated by videothoracoscopy weremore frequent in our first 30 cases (learning curve), sincethe technique was initiated in our department for the treat-ment of this disease. One must also indicate the direct re-

lationship between the complications and the lack of activecollaboration on the part of the patient in the physiotherapy;in the last 60 patients we have had practically no compli-cations. The global number of complications has been 26(23.6%), distributed as follows: eight air leaks of more than5 days of duration (8.2%); eight transitory apical air cham-bers (8.2%), which disappeared with physiotherapy in 7days maximum, and among these ones only two needed anew pleural drainage Pleurocath; on four occasions thechamber was not eliminated either with physiotherapy orwith the placement of the new drainage; six patients pre-sented a few quantity pleural effusion, which has evolvedtoward its resolution after a drainage by thoracocentesis. Intwo cases (1.8%) the pneumothorax had a recurrence in15–30 days after the intervention. The four patients with theunresolved air chamber and the two ones with recurrenceswere operated through a lateral thoracotomy. In the formerones, bullae were not in evidence, having enough with anew pleural abrasion and a forced expansion of the lung tosolve the problem. In the two patients with recurrence, onecould be observed to have a bulla not removed (this was oursecond case) and the other one to have the suture staplersopen, perhaps due to the improper use of a white loader(vascular) with 3-mm staples, instead of a blue one (paren-chymatosous) with 3.5-mm staples. The exeresis of thebulla in the first patient and the resuture with the blue loaderin the second one solved the problem.

In the follow-up of the patients, between 1 month and 4years, no more complications, or sequelae, or recurrencesappeared.

In the patients operated by thoracotomy the followingcomplications appeared: 46 air leaks of more than 5 days ofduration (7.6%); 32 apical air chambers (5.1%), and 10 ofthem required new placement of endopleural drainage; 12pleural effusions (1.9%), which were solved with drainage;five wound surgical infections (0.8%). No recurrences havebeen observed in these patients.

The duration of the operation in the patients operated byvideothoracoscopy has gone from 100 min in the first casesto 30–40 min nowadays, which is similar to open surgery.

Postoperative hospital stay has decreased from 8.5 daysmean in the open surgery to 4.5 days mean in those operatedby thoracoscopy.

The analgesia used in the patients that underwent tothoracotomy was 2 g intravenous magnesic metamizol four

Table 1. Type of thoracotomy in open surgery

Axillary thoracotomy 14Lateral thoracotomy 603Posterolateral thoracotomy 2Middle sternotomy 8

Table 2.Spontaneous pneumothorax; findings at operation by thoracotomy

Section of adhesions 174 (27.7%)Exeresis of Bullae-blebs 580 (92.5%)Electrocoagulation of blebs 31 (4.9%)Exeresis of giant bullae 27 (4.3%)Exeresis of apical scars 105 (16.7%)Others 47 (7.5%)

Table 3. Spontaneous pneumothorax; findings at operation by VATS

Section of adhesions 37 (33.6%)Exeresis of bullae-blebs 101 (91.8%)Electrocoagulation of blebs 9 (8.2%)Exeresis of giant bullae 4 (3.6%)Exeresis of apical scars 15 (13.6%)

920

times a day during the first 3 days, followed by one 575-mgpill every 6 h after that. This was carried out during thewhole hospital stay; analgesia was maintained during anextended postoperative period (about 2 months) in 50% ofthe patients. However, the patients operated by VATS re-quired maximum analgesic doses (magnesic metamizol, or,more recently, 30 mg intravenous ketorolac trometamol,four times a day) only the 1st day, and later one 10-mg pillevery 6 h during the following 3–4 days, generally until thewithdrawal of the posteroinferior drainage, continuing withanalgesia thereafter only in 10% of the patients after leavingthe hospital.

The reintegration of the patients to their working lives isimportant and it is closely related to the postoperative painand to the respiratory rehabilitation. The patients operatedby videothoracoscopy are immediately sent to the Rehabili-tation Department after they leave the hospital, and thiskeeps them from 20 to 30 days in respiratory physiotherapyprograms. After that, most of the patients return to activitiesthat they were involved in before the operation. On thecontrary, more than the 60% of the patients operated byopen surgery delay the return to their working activities 2 ormore months, which implies an important loss of workingdays.

Though less important than the parameters previouslyindicated, we should not forget the cosmetic results. Whilewe can operate by videothoracoscopy with only three inci-sions of about 15 mm, in open surgery we need a thoracot-omy of at least 10–15 cm, and this implies worse cosmeticresults.

Discussion

Since May 1992 we have performed surgical treatment ofspontaneous pneumothorax using videothoracoscopy [8].Until then, we were strong supporters of surgical treatmentby limited lateral thoracotomy since we consider that theresection of the bullae and the pleural abrasion is the onlyeffective treatment against this affliction. We do not agreewith treatment based on instillation of irritant agents in thepleural space, such as tetracycline or talc, because resultsare poor, although on certain occasions we are obliged touse them in patients of advanced age with concomitant im-portant diseases that contraindicate the surgical treatment[7].

On the other hand, conventional thoracotomy causesconsiderable postoperative pain, and this implies less col-laboration by the patient during the postoperative period,which increases the postoperative complications and hospi-tal stay. With the introduction of videothoracoscopy, theseproblems are enormously minimized, reducing pain andstay in the hospital [1, 12, 16]. When we compare the resultsin our experience obtained by VATS (n 4 110) with thosefor thoracotomy (n 4 627), we see that although it is dif-ficult to specify the intensity of the pain, the patients in thefirst group complain much less and carry out the respiratoryphysiotherapy with more intensity and less trouble; the an-algesic doses, equal during the 1st day, are reduced there-after in the VATS group. Equally, in the follow-up, theypresent less pain and they do not present dysesthesia, whichcustomarily occurs after the thoracotomy.

With regard to pleurodesis, we do not create any scari-fication—neither with a laser (because we do not have any)nor with electrocoagulation because considerable pain isreported in such cases [10]. Neither do we accomplish api-cal pleurectomy since it does not provide more benefits thanthe abrasion with sponge and the iodine instillation, and itsometimes produces serious bleeding problems.

Upon analyzing the complications presented betweenboth groups, we observe that, of the 627 patients operatedby thoracotomy, we had a complication rate of 15.1%,smaller than in the thoracoscopy group (23.6%). Neverthe-less, one must take into account that VATS is a new tech-nique that we began to use precisely in this type of pathol-ogy, the complications having almost practically disap-peared in the last 60 patients. It is quite possible that theintensity with which we accomplished the pleurodesis at thebeginning was not enough to achieve a right pleural sym-physis. On the other hand, we also began to place a Pleu-rocath in the inferior port, but we abandoned it because werealized it was not enough to evacuate the effusion gener-ated after the abrasion.

With regard to the duration of the intervention, eventhough at the beginning it lasted 100 min mean, currently itis 30–40 min mean, similar to the time used with thoracot-omy.

No one died in either group and the hospital stay meanhas decreased from 8.5 days (range, 5–22) for those oper-ated by thoracotomy to 4.5 days (range, 2–14) for thosetreated by VATS.

In conclusion, we consider that VATS must currently bethe elective method of choice for treatment of spontaneouspneumothorax, even in the first episode. Minimal thoracicincisions reduce the postoperative pain; this results in betterrecovery of the patient, decrease in the postoperative hos-pital stay, earlier return to working life, and consequentsocioeconomic savings. Finally cosmetic results are alsogood.

References

1. Bertrand PC, Regnard JF, Spaggiari L, Levi JF, Magdeleinat P, Guib-ert L, Levasseur P (1996) Immediate and long-term results after sur-gical treatment of primary spontaneous pneumothorax by VATS. AnnThorac Surg 61: 1641–1645

2. Elfeldt RJ, Schroeder D, Meinicke O (1991) Spontanpneumothoraxuberlegungen zur aetiologie und therapie. Chirurg 62: 540–546

3. Hazelrigg SR, Landreneau RJ, Mack JM, Acuff T, Seifert PE, Aver JE,et al. (1993) Thoracoscopic stapled resection for spontaneous pneu-mothorax. J Thorac Cardiovasc Surg 105: 389–393

4. Inderbitzi R, Leiser A, Furrer MU (1994) Three years’ experience invideo-assisted thoracic surgery for spontaneous pneumothorax. J Tho-rac Cardiovasc Surg 107: 1410–1415

5. Liu HP, Lin PJ, Hsieh MJ, Chang JP, Chang CH (1995) Thoracoscopicsurgery as a routine procedure for spontaneous pneumothorax. Resultsfrom 82 patients. Chest 107: 559–562

6. Loscertales J, Ayarra FJ, Garcı´a Dıaz FJ, Arenas Linares C, RicoAlvarez A (1988) Neumoto´rax esponta´neo. Ed Elba S.A.

7. Loscertales J, Garcı´a Dıaz F, Jime´nez Mercha´n R, Ayarra FJ, ArenasLinares C, Giro´n Arjona JC (1994) Tratamiento del neumoto´rax es-pontaneo en pacientes mayores de 70 an˜os. Arch Bronconeumol 30:344–347

8. Loscertales J, Jime´nez Mercha´n R, Ayarra Jarne FJ, Garcı´a Dıaz FJ,Arenas Linares C, Giro´n Arjona JC (1995) Nuestra experiencia en el

921

tratamiento del neumoto´rax esponta´neo por videotoracoscopia. Cir Es-panola 57: 526–529

9. Mercier C, Page A, Verdant A, Cossette R, Dontigny L, Pelletier C(1976) Outpatient management of intercostal tube drainage in sponta-neous pneumothorax. Ann Thorac Surg 22: 163–165

10. Mitchell RL (1990) The lateral limited thoracotomy incision standardfor pulmonary operation. J Thorac Cardiovasc Surg 590–596

11. Murray K, Matheny R, Howanitz P, Myerowitz D (1993) A limitedaxillary thoracotomy as primary treatment for recurrent spontaneouspneumothorax. Chest 103: 137–142

12. Nathanson LK, Shimi SM, Wood RAB, Cuschieri A (1991) Video-thoracoscopic ligation of bulla and pleurectomy for spontaneous pneu-mothorax. Ann Thorac Surg 62: 316–319

13. Noierclerc M, Dor V, Chauvin G (1973) Extensive lateral thoracotomywithout muscle section. Ann Chir Thorac Cardiovasc 12: 181–186

14. Sampietro R, Biraghi T, de Angelis G, d’Urbano C, Fuertes F (1993)Videotoracocirugı´a: un nuevo me´todo para el tratamiento de la pato-logıa pulmonar. Cir Espan˜ola 53: 287–291

15. Van Test WF, Roukema JA, Verpalen MC, Lobach HJ, Palmen FM(1991) De spontane pneumothorax; operen of niet? Ned Tijdschr Ge-neeskd 135: 1174–1178

16. Waller DA, Forty J, Morritt GN (1994) Video-assisted thoracoscopicsurgery versus thoracotomy for spontaneous pneumothorax. Ann Tho-rac Surg 58: 372–377

17. Weeden D, Smith G (1983) Surgical experience in the management ofspontaneous pneumothorax, 1972–82. Thorax 38: 737–743

18. Yim APC, Ho JK, Chung SS, Ng DYC (1994) Video-assisted thora-coscopic surgery for primary spontaneous pneumothorax. Aust NZ JSurg 64: 667–670

922

A new approach in the management of incarcerated hernia

Emergency laparoscopic hernia repair

K. Liao, J. Ramirez, S. Carryl, G. W. Shaftan

Department of Surgical Services, The Brookdale University Hospital and Medical Center, Linden Boulevard and Brookdale Plaza,Brooklyn, NY 11212, USA

Received: 18 November 1996/Accepted: 26 December 1996

Abstract. The authors describe the case of a 74-year-oldmale presenting with an incarcerated epigastric hernia. Analgorithm for successful management of such a case is pro-posed.

Key words: Emergency laparoscopy — Incarcerated hernia

Although laparoscopy has been widely used for electivehernia repair, little is known about its role in the emergencymanagement of incarcerated hernias. A case of incarceratedepigastric hernia which was successfully managed by lapa-roscopy is reported here.

Case report

A 74-year-old black male with a past medical hisotry of hypertension,diabetes, heart disease, and two episodes of stroke with resultant righthemiplegia was brought to the Emergency Department of the BrookdaleUniversity Hospital and Medical Center after his primary care physicianfound a painful mass in the abdominal wall about 14 h previously. Prior tothis visit, the patient had abdominal discomfort and recurrent vomitingduring the past month. His wife recalled that the patient had a golf-ball-sizemass at the midabdomen about 2 weeks ago but it disappeared spontane-ously 5 h after she ‘‘massaged’’ the mass. The patient had three episodesof vomiting after he arrived at the hospital; his last bowel movement wasthe night before coming to the Emergency Department. Physical exami-nation revealed a moderately obese man lying supine who was not suffi-ciently oriented to answer any questions. The vital signs were: blood pres-sure 136/80 mmHg, pulse 89 per min, respiration 20, and temperature36.3°C. The abdomen was slightly distended with poor abdominal muscletone, especially on the right side. A 5 × 6 × 6 cmsupraumbilical mass waspalpated that was firm, tender, and irreducible. The surface of the mass wassmooth and it was normothermic. There was moderate guarding around themass and in the right lower quadrant but there was no obvious rebound.Bowel sounds were audible; rectal examination was unremarkable andstool guaiac was negative. Hemoglobin was 12.8 g/dl, hematocrit 39.8%,white blood cell 7,200/mm3, and urinalysis was normal. Abdominal X-ray

showed a few slightly dilated small-bowel loops with no air-fluid levelsand air was noted in the colon. A diagnosis of incarcerated epigastric herniawas made.

Emergency operation was planned; however, during the transport ofpatient to the operating room, the hernia reduced spontaneously. A 2.5-cmfascial defect was found in the midline, 2 cm above the umbilicus. Therestill was some tenderness and guarding in the supraumbilical area and inthe right lower quadrant. Since ischemic bowel could not be ruled outbased on the clinical findings, laparoscopy was performed. After nasogas-tric and bladder intubation, a 1-cm skin incision was made along the rightanterior axillary line at the level of midabdomen through which a 10-mmtrocar was placed into the abdominal cavity by an open approach. Theabdomen was insufflated with carbon dioxide to maintain 15 mmHg pneu-moperitoneum and then the camera was inserted. Another two 5-mm tro-cars were placed under direct vision at the right upper and right lowerquadrants. A 2.5 cm linear defect was seen in the linea alba just above theumbilicus and a 8 cmsegment of small bowel was found to be erythem-atous and edematous with patchy ecchymoses on the serosa which involvedthe entire circumference. No free fluid was found inside the abdomen. Withthe help of the graspers, a piece of 4 × 8 cm2 Gore-Tex DualMesh (W. L.Gore & Associates, Inc., Flagstaff, AZ) was inserted inside the abdomenand was laid flat over the defect. Two 2-0 nylon straight needle sutureswere placed percutaneously at about 1 cm above and below the defect toanchor the mesh at 6 and 12 o’clock: The needle passed through theabdominal wall and the mesh and with the help of the grasper inside; theneedle then was pushed back through the mesh and the abdominal wall,exiting close to the entrance. The knots were tied just underneath the skin.A Tacker (Origin Medsystems, Inc., Menlo Park, CA) was inserted via a5-mm trocar and multiple titanium tacks were fired to fix the mesh aroundthe defect. Good covering of the mesh over the fascial defect was achievedafter 12 tacks were placed. The previous hernia defect area appeared tightwhen the pressure was applied from outside. The compromised small-bowel segment was reevaluated before the camera was removed, and thecolor of the bowel has improved dramatically compared to its appearanceat the beginning of the operation. Good peristalsis also was noted in thatsegment of bowel. After removal of all the trocars, the fascia and the skinopenings were closed. Postoperatively the patient recovered without com-plications and he was discharged from the surgical service on the 5thpostoperative day.

Discussion

Traditionally, the decision to perform closed taxis or emer-gency operative repair for an incarcerated hernia is based onthe presence or absence of the clinical signs of impending orCorrespondence to:K. Liao

Surg Endosc (1997) 11: 944–945

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

actual strangulation of the hernial contents. These includepain in the region of the hernial swelling, tenderness topalpation, and discoloration of the tissues over the swelling[4]. The clinical ability to objectively identify strangulationof bowel is far from ideal, especially in the aged patientswith multiple medical problems and compromised mentalstatus. Too-aggressive emergency surgical intervention forincarcerated hernias, especially when laparotomy is re-quired, carries a much higher mortality and morbidity thando those operated electively [3, 5]. On the other hand, too-vigorous efforts to reduce an incarcerated hernia may resultin ‘‘reductio en masse’’ or reduction of strangulated, non-viable bowel [4]. Most surgeons have found it very difficultto choose between emergency operation and observation ifconfronted with the clinical situation of an unreliable, dis-oriented patient. Laparoscopy should be considered in in-carcerated hernias under certain circumstances. Althoughincarcerated epigastric hernia is rare, similar principles ofmanagement also should apply with other abdominal her-nias. With laparoscopic herniorrhaphy one can thoroughlyevaluate the viability of reduced hernial contents withoutexposing them to the external environment, and, at the same

time, repair the hernia; there is much less trauma for thepatient compared with open repair, especially when celi-otomy is required to evaluate bowel viability [1, 2]. In theera of laparoscopic surgery and managed care, the applica-tion of laparoscopy to the management of incarcerated her-nias certainly offers an extra choice for the surgeon. Analgorithm for management is proposed (Fig. 1).

References

1. Barkun JS, Wexler MJ, Hinchey EJ, Thibeault D, Meakins JL (1995)Laparoscopic versus open inguinal herniorrhaphy. Surgery 118: 703–710

2. Feldman MG, Russell JC, Lynch JT, Mattie A (1994) Comparison ofmortality rates for open and closed cholecystectomy in the elderly:Connecticut statewide survey. J Laparoendosc Surg 4(3): 165–172

3. Hjaltason E (1981) Incarcerated hernia. Acta Chir Scand 147: 263–2674. Nyhus MN, Bombeck CT, Klein MS (1991) Hernias. In: Sabiston DC

(ed) Textbook of surgery. 14th ed. WB Saunders, Philadelphia, pp1134–1148

5. Stephens BJ, Rice WT, Koucky CJ, Gruenberg JC (1992) Optimaltiming of elective indirect inguinal hernia repair in healthy children:clinical considerations for improved outcomes. World J Surg 16: 952–957

Fig. 1. Clinical approach to theincarcerated hernia.H + P,evaluated by history and physicalexamination; +, success; −, failure.

945

Review article

Laparoscopic anatomy of the region of the esophageal hiatus

G. G. R. Kuster,1 F. A. Innocenti2

1 Division of General Surgery, MS213, Scripps Clinic and Research Foundation, 10666 North Torrey Pines Road, La Jolla, CA 92037, USA2 Department of Surgery, University of Concepcion School of Medicine, Concepcion, Chile

Received: 24 October 1996/Accepted: 26 December 1996

AbstractBackground:The wide patient acceptance of hiatal and gas-troesophageal surgery performed by laparoscopy has sud-denly generated a large volume of procedures frequentlydone by surgeons with limited experience in this area. Thishas resulted in an excessive number of complications.Knowledge of the normal and pathologic laparoscopicanatomy is essential for safe dissection around the esopha-geal hiatus.Methods:This description is based on the experience gainedduring 850 open and 150 laparoscopic surgeries in andaround the hiatus and on the review of the literature.Results:Laparoscopic approach, dissection, and accessibil-ity of the hiatus and surrounding organs are different thanthose experienced through cadaveric dissection and opensurgeries.Conclusions:Clear understanding of the normal and patho-logic anatomy and its variations facilitates laparoscopic dis-section of the hiatus and neighboring structures and shouldhelp the surgeon avoid complications.

Key words: Laparoscopic anatomy — Esophageal hiatus— Diaphragm — Laparoscopic fundoplication — Gastro-esophageal junction

Complications such as excessive bleeding [12, 16, 20, 21];esophageal and gastric perforations [9, 12, 20, 21, 41, 49,54]; pneumothorax [12, 16, 20, 21, 24, 54]; pneumodiasti-num [10]; paraesophageal herniation [10, 21, 54]; dysphagia[12, 16, 21, 38]; and delayed gastric emptying and diarrheapossibly due to vagal injury [12, 37, 48] have occurred moreoften after laparoscopic procedures in the hiatal area thanafter open surgery. This is due, in part, to the surgeon’s lackof familiarity with the laparoscopic anatomy of the region.

Many procedures have been performed by surgeons withgreat expertise in advanced laparoscopic procedures butlimited experience with open surgery of the hiatal area [13].A clear understanding of the normal and pathologic laparo-scopic anatomy of the hiatus and surrounding organs shouldminimize the risks of injuries and poor functional results,making the laparoscopic approach comparable to or betterthan the open procedure [34, 37, 39, 46, 50, 51, 54]. Expe-rience gained with laparoscopic cholecystectomy has dem-onstrated that knowledge of the anatomy and its variations,as well as the distortions resulting from disease, is essentialto anticipate and prevent technical complications [1, 11, 29,42]. The region of the hiatus involves several organs, whichin some pathologic conditions may vary significantly intheir relationship, shape and size, tissue consistency, andfragility.

This study is based on the anatomic findings of 850open and 150 laparoscopic surgical dissections performed inthe area of the hiatus and on a review of the literature. Thedetailed anatomic descriptions of the surgical dissectionsrecorded by the senior author were reviewed for location,size, consistency, variations, pathologic distortions, andother elements of interest encountered in the structuresaround the hiatus. These operations included hiatal herniarepairs; antireflux procedures; distal esophagectomies;esophageal myotomies; local removal of tumors, diverticuli,or cysts; total and proximal gastrectomies; splenectomies;vagotomies; and miscellaneous procedures involving thediaphragm and perihiatal areas.

The anatomic description will be presented in the orderthat the surgeon encounters the organs and planes of thetissue when dissecting the area for various surgical proce-dures. Special emphasis is placed on the frontal view of theanatomic structures as seen and felt with the laparoscopicinstruments.

Patient position and trocar insertion forhiatal approach

The patient is placed in a semilithotomy position with orwithout a lumbar cushion [18] and support under the gluteal

Presented at the 4th International Congress of the European Association forEndoscopic Surgery, Trondheim, Norway, June 1996

Correspondence to:G. G. R. Kuster

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 883–893

area to prevent the patient from sliding when placed in steepFowler’s position. The laparoscope can be introducedthrough any area of the upper abdominal wall. However, thefact that the diaphragmatic hiatus is quite posterior andcephalad within the abdominal boundaries should be con-sidered. The location of a mid-distance between the xiphoidand the umbilicus is used for most operations. A 0° anglescope placed at the level of the umbilicus or below will offera tangential view of the hiatus and will not provide a goodretroesophageal visualization. This can be offset by the useof a 30° or 45° angle scope. The exact position of the scopeshould be tailored to the size of the patient and the actualprocedure. The size of the lateral segment of the left lobe ofthe liver will also influence the position of the scope. Thisis particularly important for the location of the port whichwill accommodate the liver retractor. It is a good practice tointroduce the first trocar in a site which is not critical, likethe left subcostal margin, and to insert the rest of the trocarsunder direct vision after a complete abdominal exploration,tailoring the site and angle of insertion according to thefunction of each port. For procedures that will involve onlythe front part of the hiatus, umbilical access is adequate[30].

A general view of the upper abdomen will disclose thefront portion of the dome of the diaphragm, the lateral seg-ment of the left lobe of the liver, and part of the stomach. Inlarge paraesophageal hiatal hernias, half or more of thestomach is in the chest cavity and the general examinationof the upper abdomen may not reveal any portion of thestomach without lifting the liver and/or reducing the stom-ach by exerting traction on the gastrocolic omentum [27,28]. In the majority of adult patients, the spleen is not vi-sualized during this general examination of the upper abdo-men. Part of the transverse colon might also be seen, de-pending on the thickness of the greater omentum, whichmay preclude the frontal view of the entire transverse colon,particularly in obese people. The falciform ligament is alsowell visualized, and an examination of the right side of theupper abdomen will disclose the diaphragm, the rest of theliver and, most of the time, the fundus of the gallbladder.

The rest of the ports are placed in the subcostal areas.Depending on the body build of the patient, most accesscannulas will be located well below the coronal plane of thehiatus, except for those entering at the right or left side ofthe xiphoid, which may reach the hiatus directly posteriorlyin individuals with a wide habitus.

Left lobe of the liver

For exposure of the hiatus, the lateral segment of the leftlobe of the liver needs to be retracted upward with a non-traumatic instrument (Fig. 1) or folded to the right afterdividing the left triangular ligament. Thetriangular liga-mentdoes not insert exactly along the posterior border ofthe left hepatic lobe but instead on its superior surface,about 1 cm from the posterior border [35]. The insertion onthe liver reaches the posterior border as it approaches theinferior vena cava. This ligament is thin and contains minorvessels, particularly at the left end, but it widens at the rightto accommodate the left hepatic veins and the inferior venacava. The left triangular ligament measures an average of

1.5 cm wide in its left half and narrows to the right. Thelength of the left triangular ligament varies widely, follow-ing the transverse length of the left lobe of the liver. Thereare cases of long ligaments extending well beyond the leftend of the liver to the lateral abdominal wall, as well ascases of short or rudimentary ligaments. If the triangularligament is transected, attention should be given to avoidinjury to the upper end of the spleen (particularly in thepresence of a very long left hepatic lobe reaching over andbeyond the upper pole of the spleen), as well as the liver, thefundus of the stomach and esophagus, the left suprahepaticveins, and branches of the left and right diaphragmatic ar-teries and veins.

Variations in thesize and shapeof the left lobe of theliver are frequent. The liver as seen by frontal radiocolloidscintigraphy shows a smaller craniocaudal dimension in su-pine position than standing position [22]. This is particu-larly true in the left lobe of normal individuals (27% de-crease is seen). Cirrhosis allows less pliability, and thereforethe decrease in size is only 18%. The left lobe of the livermay present anomalous lobulations, accessory lobes with apedicle, ectopic hepatic tissue without pedicle, and hyper-plasia, hypoplasia, or absence of the lateral segment. Ac-cessory hepatic lobes presenting as large and pedunculatedmasses are rare [25]. Jaques et al. [23] described ten casesof hypoplasia or aplasia of one or both segments of the leftlobe among 517 abdominal computed tomographic scans.The left lobe is frequently elongated and embraces thespleen [4].

The left lobe may be congentially atrophic or totallyabsent. Congenital atrophy has also been described in asso-ciation with hypertrophy of the caudate lobe, with gastricvolvulus [3], and with hepatodiaphragmatic colon interpo-sition (Chilaiditi’s syndrome) [23]. Acquired atrophy of theleft lobe of the liver can be focal or diffuse. Volume dimi-nution is a frequent finding in cirrhosis.

The consistency, texture, and colorof the liver vary in

Fig. 1. Exposure of the hiatal area by retracting the lateral segment of theleft lobe of the liver (a); (b) triangular ligament; (c) stomach and esophagusbeing retracted to the left; (d) gastrohepatic omentum with its thin area(‘‘window’’) between a branch of the left inferior phrenic artery (above)and the hepatic division of the vagus nerve and an aberrant hepatic artery(below).

884

normal as well as pathologic conditions. The increased ri-gidity noted in cirrhosis in relation to the degree of fibrosismay interfere with the mobility and bendability of the organ.The presence of soft and fragile liver should be of majorconcern to the laparoscopist because of the possibility ofsurgical trauma resulting in intra- and postoperative hem-orrhage, bile leaks, or tissue necrosis. Some elderly patientsand patients with fatty infiltration of the liver, acute hepa-titis, hypercortisonism, or lymphoma may present with anextremely fragile liver. The surgeon and assistants should beextremely gentle with liver retractors and the blind intro-duction of instruments that could injure the liver. The Glea-son capsule may detach easily from the liver parenchymawhen traction is applied to adhesions or hepatic ligaments.

Gastrohepatic omentum

The lesser omentum, also known as the gastrohepatic omen-tum, ligament, or membrane, extends from the porta hepatisto the lesser curvature of the stomach, the antero-right wallof the esophagus, and the diaphragm. The gastrohepaticomentum is best visualized by exertingtraction on the stom-ach toward the left side and caudally (Fig. 1). It containsbranches of the left gastric artery and vein, the hepaticbranch of the anterior vagus nerve, the posterior and anteriorLatarjet nerves, and some lymphatic vessels and nodes. Italso frequently carries branches of the right gastric arteryand vein, an aberrant hepatic artery, and branches of the leftinferior phrenic artery and vein.

The caudal portion of the lesser omentum correspondsto the hepatoduodenal ligament, extending from the liver tothe first 2.5 cm of the duodenum. The free edge of thisligament forms the anterior boundaries of the hiatus of Win-slow and contains the hepatic artery, the portal vein, theextrahepatic biliary tract, lymphatic vessels, and nodes. Thecephalad end of the lesser omentum extends to the dia-phragm, where its anterior layer reflects upward behind theposterior border of the left lobe of the liver to join thetriangular ligament [35].

The thicknessof the gastrohepatic omentum changesfrom the diaphragm to the hiatus of Winslow. The cephaladportion (0.5 to 2 cm long by 2 to 3 cm wide) is fibrous,opaque, and tense. It joins the phrenoesophageal ligamentand frequently contains branches of the left inferior phrenicand left gastric vessels. Caudally to this thick segment is anarea 3 to 6 cm long by 3 to 4 cm wide which forms atransparent ‘‘window,’’ with only the anterior and posteriorperitoneal layers and minimal fat surrounding a loose net-work of very thin vessels. This area, called the ‘‘Kusterwindow’’ by the trainees, permits the laparoscopic view ofthe caudate lobe of the liver and the right crus of the dia-phragm. The ‘‘window’’ portion of the lesser omentum is avery important landmark in laparoscopic dissection of thearea of the hiatus since it is a constant finding and can beeasily torn with or without minimal bleeding, gaining im-mediate access to the right side and the posterior portions ofthe hiatus and its contents (Fig. 2). In large hiatal hernias,particularly paraesophageal, the lesser omental windowmay not be readily apparent until reducing the stomach to itsintraabdominal position.

Caudal to the ‘‘window,’’ the lesser omentum becomes

thicker with fat and contains the hepatic vagal branch cross-ing from the lesser curvature to the liver and the aberranthepatic artery, which is present in 11 to 23% of the cases[17, 20].

Caudate lobe of the liver

The caudate lobe can be seen through the transparent gas-trohepatic omentum and can be reached by opening thelesser omentum at the ‘‘window’’ (Fig. 2). This lobe issituated between the inferior vena cava to the left and theright crus of the diaphragm to the right. Occasionally, thecaudate lobe protrudes to the right between the porta hepatisand the inferior vena cava, constituting the so-called caudateor papillary process ]19, 52].

Particular attention should be given to avoid injury tothe caudate lobe during dissection of the right crus of thediaphragm and during repair of hiatal hernias at the time ofplacement of sutures between the left and right crus of thediaphragm behind the esophagus.

Esophageal hiatus

The esophageal hiatus is an inverted teardrop-shaped open-ing located about 1 cm to the left of midline, at the level ofthe tenth thoracic vertebra. Its anterior and lateral bordersare formed by the muscular and fibrous edge of thedia-phragmatic crura,and the posterior angle is formed by themedial arcuate ligament.The crura start in the anterolateralsurface of the first lumbar vertebrae, the intervertebraljoints, and the anterior longitudinal ligament. The right crusinserts on the anterior surface of L1 to 3 or 1 to 4 and the leftcrus on L1 to 2 or 1 to 3, as well as on the intervertebraldisks and the arcuate ligament. The crura extend cephaladand anteriorly to form the hiatus, inserting anteriorly intothe transverse ligament of the central tendon of the dia-

Fig. 2. The thin ‘‘window’’ portion of the gastrohepatic omentum hasbeen opened to show the (a) caudate lobe of the liver and (b) right crus ofthe diaphragm. The dissection of the hiatus is initiated by blunt dissectionof the free border of the right crus with a Peanut sponge mounted on a rod(c).

885

phragm. There are many variations in the formation of thehiatus [32]. In more than 90% of cases, the right crus formsthe right and left margins of the hiatus, or the left cruscontributes to the right margin. The right crus is immedi-ately behind the gastrohepatic omentum and is located cau-dally in relation to the left crus (Fig. 3). The hiatus closelysurrounds the esophagus. There may be space between theesophagus and the hiatus posteriorly, filled with areolar orfatty tissue, but normally no more than 2 cm. There shouldnot be any sac herniation of peritoneum between theesophagus and the hiatus in any place, unless there is ahiatal hernia.

Theorientationof the hiatus in relation to the esophagusis neither perpendicular nor symmetrical. The ellipticalopening of the hiatus is slanted to the right in its posteriorend in such a way that the esophagus passes in front of theleft crus. In addition, the hiatal opening is oriented in asagittal view from anterior-cephalad to posterior-caudal,with the esophagus passing into the abdomen first in itsanterior wall and then its lateral walls obliquely and finallyits posterior wall at a subdiaphragmatic position 2 to 2.5 cmfurther caudally. In large paraesophageal hernias the meresurgical displacement of the esophagus from its posteriorlocation to the anterior border of the hiatus brings 2 to 3 cmof esophagus to an intraabdominal position.

The laparoscopic dissection of the hiatus should be donebluntly, since there are no important vessels between thehiatal border and the esophagus (Fig. 2). This is greatlyfacilitated by the use of a gauze sponge mounted at the tipof a rod (‘‘Peanut sponge,’’ Ideas for Medicine, Inc., Clear-water, FL 34622) impregnated with a 1:10,000 epinephrinesolution [26]. The dissection is started byopening the lesseromental ‘‘window’’ and cutting the peritoneum along themedial border of the right crus.The right crus is thenpushed to the right and the hiatal contents are pushed to theleft, separating the two until the arcuate ligament is reached,so the dissection of the left crus may be started from below.This dissection is performed retroperitoneally.To reenter

the peritoneum in the left side of the hiatus, a rent should bemade in the peritoneum between the left crus and the rightwall of the esophagus.This can be done from behind theesophagus or from the left side by retracting the cardia andesophagus caudally and to the right. If the dissection isadvanced cephalad to the edge of the left crus, the medias-tinum is entered, andfurther dissection will perforate theleft pleura.

Theanterior borderof the hiatus is dissected bydetach-ing and lifting the phrenoesophageal ligamentfrom the an-terior wall of the esophagus or by cutting through thephrenoesophageal ligament transversely at the edge of thediaphragm until the transverse muscle bundles of the hiatusare identified in front of the longitudinal fibers of theesophagus and the anterior vagus nerve.

Dissection of the hiatus is extended into the mediasti-num, where one encounters the pericardium in the front, theright and left mediastinal pleuras laterally, the aorta poste-riorly, and the esophagus with its vessels, lymph nodes, andvagus nerves in the center.

In open surgery, the dissection of the hiatus is usuallystarted anteriorly. In laparoscopic surgery, the dissectionshould be started posteriorly;otherwise drops of bloodmight drain down from the front, obscuring the tissues: redmuscle fibers, yellow fat, grayish fascia and ligaments, andwhite vagus nerves.

Both right and left crus can be, and have been, mistakenfor the esophagus and can be dissected, sometimes enteringthe pleural cavity. This error is avoided by paying attentionto the direction of the muscle fibers and by moving thecardial end of the stomach right and left; the esophagusfollows the stomach while the crura are fixed posteriorly.

The right and left crus can be transected at midlevel ifthe hiatus needs to be enlarged (i.e., to accommodate thestomach or colon to replace a resected esophagus). In thiscase the pleura should be retracted to avoid entering thechest cavity and branches of the left inferior phrenic arterymay need ligature.

The crura on the sides and behind the esophagus havelittle or no tendinous fibers. The hiatal border is strongeranteriorly because of the endothoracic and endoabdominalfascia and the central tendon of the diaphragm. The crurabecome tendinous as they reach their vertebral origin andtheir fusion with the arcuate ligament. Most of the length ofthe right and left crus is formed by a 1–1.5-cm-thick bundleof very soft parallel muscle fibers which tear easily ifgrasped or sutured and tied firmly.

Phrenoesophageal ligament or membrane

The phrenoesophageal ligament is formed by the cephaladextension of the endoabdominal fascia and the caudal ex-tension of the mediastinic fascia fusing at the hiatus. Thefibers which come from the endoabdominal fascia gothrough the hiatus and insert in the adventitia and the con-nective tissue located between the muscle fibers of theesophagus 1 to 2 cm above the hiatus, completely encirclingthe esophagus. A second sheet of this same fascia descendsto insert in the adventitia of the anterior half of the perimeterof the abdominal esophagus, 1 to 2 cm caudal to the hiatalborder (Fig. 4). The fibers that come from the endothoracic

Fig. 3. View of the hiatus after complete dissection: (a) anterior and pos-terior vagus nerves retracted together with the esophagus; (b) right and leftcrura; (c) upper pole of the spleen.

886

fascia are thinner and less constant than the abdominal fas-cia, and they join the endoabdominal fibers. During fetallife, these two fascias are very strong, with fibrous andmuscular components [5]. They regress during the matura-tion of the fetus although muscle fibers may persist. Withthe active swallowing and respiratory movement after birth,this union is modified and becomes more lax; connectiveand fatty tissue fill the space between the two layers. Themembrane is less evident in the adult except in its anteriorportion, where a very distinct, strong, white fascia is clearlyseen. Patients with large hiatal hernias may show a veryelongated, thin, and practically nonexistent phrenoesopha-geal ligament [17].

Vessels between esophageal and inferior phrenic arter-ies and veins pass over the ligament and need control beforedetaching the phrenoesophageal ligament from the esopha-gus.

Esophagus

The length of the abdominal segment of the esophaguswhich can be seen by laparoscopy varies from 0.5 to 2.5 cm.It is located between T11 and L11, but this position varieswith height and body build of the individual. Before dis-section, the location of the esophagus is not clearly apparentby laparoscopy.

The abdominal esophagus is in contact with the left lobeof the liver anteriorly, with the caudate lobe and part of theright crus on the right side, with part of the left crus andfundus of the stomach on the left, and with the left crus ofthe diaphragm, the arcuate ligament, and the aorta posteri-orly. The upper pole of the spleen may reach the left wall ofthe esophagus in cases of splenomegaly.

The abdominal esophagus is essentially aretroperitone-al organ, since it is covered with peritoneum anteriorly andpartially covered at its lateral walls, whereas the posterior

wall is not covered by peritoneum (Fig. 4) and is separatedfrom the retroesophageal elements described above by are-olar or fatty tissue.

In sliding hiatal hernias,the entire abdominal esopha-gus may be found in the mediastinum, or it may float be-tween the abdomen and the supradiaphragmatic area ac-cording to the prevailing pressures between the chest andabdominal cavities. Inpure paraesophageal herniasthegastroesophageal junction remains in a subdiaphragmaticposition, with herniation of the stomach and/or other tissuesalongside the esophagus, between the esophagus and thehiatus. Retroesophageal herniations of retroperitoneal fatand posterior gastric wall do not create a peritoneal sac inthe mediastinum, whereas paraesophageal hiatal hernias lo-cated anteriorly will form a sac by elongation of the phreno-esophageal ligament and peritoneum [27, 28].

The distal mediastinal esophagus can be mobilizeddown to the abdomen for 3 to 15 cm (Fig. 3). This mayrequire detachment from periesophageal areolar mediastinaltissue, right and left pleura, aorta, and pericardium, and mayrequire division of arteries coming directly from the aorta.The vagus nerves limit the amount of mobilization. Maxi-mal mobilization is obtained by division of the anterior andposterior vagus nerve in the lower mediastinum.

Gastroesophageal junction

The gastroesophageal junction is normally located to the leftof the midline at the level of T10, at the level of the an-teroposterior projection of the seventh costal cartilage.

The longitudinal muscle layers of the esophagus runstraight through the gastroesophageal junction into thestomach where they spread like a fan (Fig. 5). The internaltransverse muscular sheet of the esophagus is formed bymuch shorter bundles of incomplete circles around the sub-mucosa. These two layers arevery distinct and can be easily

Fig. 4. The phrenoesophageal ligament (a) covers the anterior muscularborder of the hiatus and the proximal 1 to 2 cm of the abdominal esopha-gus. The distal anterior and lateral walls of the esophagus are covered byperitoneum, making the esophagus a retroperitoneal organ (b).

Fig. 5. The muscle layers of the distal esophagus and gastroesophagealjunction: (a) external longitudinal muscle bundles of the esophagus andstomach; (b) internal transverse layer of the esophagus; (c) crisscrossing ofmuscle fibers starting at the cardia.

887

dissected apart from each other and from the submucosaofthe esophagus.

The fibers of the inner muscle layer change their direc-tion at the gastroesophageal junction to create a pattern oflong and short bundles with different orientation. The longbundles form the so-called gastric sling embracing the an-terior and posterior gastric walls to the left of the cardial endof the stomach. They create the cardiac notch or angle ofHis between the esophagus and the fundus of the stomach.The short bundles form the so-called semicircular claspsand maintain the transverse orientation of the semicircularinner esophageal layer, embracing the lesser curvature. Bothsling and clasp fibers are more abundant at the gastro-esophageal junction, forming a thicker layer than in theesophagus [47].

The crisscrossingof longitudinal, oblique, and trans-verse fibers is seen at the entrance to the stomach when alongitudinal esophageal myotomy is performed. In addition,the separation of muscle layers and thedissection awayfrom the gastric mucosa is notoriously more difficultthan inthe esophagus. These features greatly facilitate the laparo-scopic myotomy which is frequently performed for achala-sia of the esophagus. However, the fibrosis and inflamma-tion of the esophageal wall observed in peptic stricturespreclude an easy distinction between esophagus and stom-ach, since the longitudinal and transverse layers may befused and adherent to the mucosa. The same problem is seenin posttraumatic strictures of the lower esophagus and afterperforations. Endoscopic biopsies of the esophageal mucosamay result in the same problem of fixation to the muscularlayers. Preoperative mucosal biopsies over a leiomyoma orlipoma may lead to penetration of the esophageal lumenduring laparoscopic enucleation of the tumors.

A circular or annular muscular formation constituting asphincter does not exist in the human gastroesophagealjunction [47]. The manometric high-pressure zone noted atthe level of the ‘‘lower esophageal sphincter’’ appears to berelated to the sling and clasp fibers at the gastroesophagealjunction. Myotomy of these fibers abolishes the high-pressure zone and the gastroesophageal competence [33,34].

Stomach

The stomach’s anterior surface is at least partly covered bythe liver and transverse colon. In most cases, the majorcurvature extends for 5 to 25 cm below the xiphoid. Thestomach is very mobile, depending on the position of thepatient and the amount and weight of the gastric contents.The gastroesophageal junction is fairlyfixed in normal sub-jects by the retroperitoneal fixation of the esophagus and thegastropancreatic ligament. Some mobility limitations areobserved in the upper part of the greater curvature due to thegastrophrenic and gastrosplenic ligaments. The proximalpart of the lesser curvature and upper posterior wall arefixed due to the gastropancreatic ligament and left gastricvessels [15]. The posterior wall of the stomach is adjacent tothe diaphragm, spleen, left kidney, left adrenal gland, pan-creas, and transverse mesocolon.

The pathologic changes of the stomach due to cancerand peptic ulcer disease are vast and beyond the scope of

this report. The change ofpositionof the stomach in hiatalhernias varies from sliding of the cardia into the lower me-diastinum to complete rolling of the stomach into a largehernial sac in paraesophageal hiatal hernias. In some casesthe entire stomach is herniated, assuming an ‘‘upside-down’’ position with significant elongation of the phreno-esophageal, gastrophrenic, and gastropancreatic ligaments.The posterior wall of the upper stomach may extend behindthe hernial sac and remain in the mediastinum after apparentreduction of the stomach down to the abdomen and totalevacuation of the sac [28].

Complete mobilization of the fundus of the stomachrequires transection of the gastrosplenic omentum contain-ing the short gastric vessels, the gastrophrenic ligament, andthe portion of gastropancreatic ligament extending from theleft of the left gastric artery toward the spleen. The anteriorwall of the fundus is the most distensible and mobile area ofthe stomach, allowing its passage to the level of the caudatelobe, in front or behind the esophagus, without the need todivide any gastric ligament in the great majority of indi-viduals, thus allowing the Rossetti modification of Nissenfundoplication [40].

Gastric wall thickness

The wall of the gastroesophageal junction is thicker at theside corresponding to the major curvature of the stomachthan toward the lesser curvature (average 4.5 mm vs 3.8mm). It is also slightly thicker in its posterior than its an-terior walls (average 3.5 mm vs 3.3 mm). The esophagusbecomes thinner as it leaves the gastroesophageal junctionin a cephalad direction [6, 7, 47].

During surgery,iatrogenic perforation of the distalesophagus and gastroesophageal junctionmay happen dur-ing blind and traumatic dissection. Special considerationshould be given to the fact that theanterior gastric wall isthe thinnest area of the stomach,particularly the proximalcentimeter below the cardia. Traction on the upper part ofthe anterior wall of the stomach should be avoided when thestomach is mobilized. A generous grasp of the stomach andpart of the lesser curvature with nontraumatic forceps ispreferable.

Arcuate ligament

The arcuate ligament is an area of fusion of the right and leftcrus of the diaphragm behind the esophagus (Fig. 6). Whenthis fusion contains a tendinous portion of the crura, theligament is a fibrous cord located 1–3 cm behind theesophagus and in front of the aorta, at the level of the firstlumbar vertebra. The arcuate ligament is located just ceph-alad to the origin of the celiac axis. When this fusion isformed only by muscle fibers, the arcuate ligament may bedifficult to identify [31]. The arcuate ligament may be ab-sent in other cases, leading to a common hiatal opening forthe esophagus and aorta (Fig. 7). This is seen frequently inlarge paraesophageal hernias [28].

Aortic hiatus

The aorta, the thoracic duct, and sometimes the azygos veincross the diaphragm through the aortic hiatus at the level of

888

Fig. 6. Arcuate ligament (a); (b) aorta.

Fig. 7. Large hiatal opening common tothe esophagus (a) and aorta (b).

Fig. 8. The elements behind theesophageal hiatus: (a) aorta and aortichiatus, (b) cisterna chyli, (c) thoracic duct,(d) esophagus, (e) inferior vena cava.

Fig. 9. Arteries to the hiatus andesophagus: (a) left inferior phrenic arteryoriginating in the aorta; (b) esophagealbranch of the left gastric artery. The leftinferior phrenic artery may originate in theceliac trunk.

Fig. 10. Veins of the hiatal region: (a) leftinferior phrenic vein; (b) left adrenal vein;(c) inferior vena cava.

Fig. 11. Vagus nerves.

889

the 12th thoracic vertebra (Fig. 8). The aortic hiatus is lo-cated posterior to the esophageal hiatus. The median arcuateligament separates these two openings, and the crura formsits lateral borders. The aorta has been mistakenly dissectedfor the esophagus in large hiatal hernias with absent arcuateligament.

Arteries of the hiatal region

The proximal portion of the stomach and the abdominalesophagus are supplied by the esophageal branches of theleft gastric artery.These branches join the esophageal ar-teries comingdirectly from the aortaabove the diaphragm.Frequently, the distal esophagus also receives branches pro-vided by theleft inferior phrenic artery.The borders of thehiatus are supplied by a branch of this artery [17]. The leftinferior phrenic artery may originate from the aorta or fromthe celiac trunk (Fig. 9).

Veins of the hiatal region

The left inferior phrenic veinis seen coursing parallel to theleft triangular ligament, passing in front of the hiatus anddraining into the inferior vena cava (Fig. 10). Theleft gas-tric vein can be identified 2 or 3 cm distal to the hiatus; itreceives tributary branches of the esophagus and may anas-tomose with the left inferior phrenic vein. In portal hyper-tension, portosystemic collateral channels are formed in thelower esophagus and stomach. A network of tortuous andhypertensive veins are seen which can bleed easily and aredifficult to control.

Vagus nerves

The vagal trunks cross the hiatus next to the esophagus (Fig.11). They are distinctly identified as anterior and posteriorbranches in 80–85% of the cases [2, 44]. The anterior vagusnerve is located more to the left than the posterior, but bothnerves, seen through a sagittal anterior/posterior view, arelocated on the right half of the esophagus in the majority ofinstances. The posterior branch may be situated closer to theaorta than to the esophagus. In 12–20% of the cases, thevagus nerves divide into multiple smaller branches wellabove the hiatus [14].

The anterior vagus nervedivides into two branches:anterior gastric and hepatic. The anterior gastric branch pro-vides the anterior Latarjet nerve, which runs 0.5 to 1 cm tothe right of the lesser curvature. The hepatic branch of theanterior vagus nerve usually separates from the anterior gas-tric branch at the level of the abdominal esophagus, advanc-ing toward the liver through the lower border of the avas-cular window of the gastrohepatic omentum. Frequently, itis formed by multiple parallel bundles.

The posterior vagus nervealso divides into twobranches: the posterior gastric and the celiac. The posteriorgastric branch provides the posterior Latarjet nerve, withbranches to a more cephalad portion of the stomach than theanterior Latarjet nerve. The celiac branch is always a single,thick nerve extending through the gastropancreatic ligamentto join the celiac plexus.

Unless some type of vagotomy is intended as part of thesurgical procedure,the vagus nerves should be protectedagainst injury during dissection of the hiatus, the esophagus,and surrounding tissues. The use ofblunt dissectionwith aPeanut sponge (as described above) will rarely result indamage to the main vagus trunks while dissecting the bor-ders of the crura away from the hiatal contents or whendissecting the esophagus in the lower mediastinum.

The anterior vagus nerve, and particularly the hepaticand Latarjet branches, may be mistakenly divided whenperforming esophageal myotomy (Heller procedure) forachalasia [49]. It is preferable to perform the myotomy tothe left of the anterior vagus nerve [36], since the branchesextending to the fundus of the stomach can be divided with-out clinical consequences. The posterior vagus nerve shouldbe retracted with the esophagus (Figs. 3 and 7). Its celiacbranch sometimes limits the caudal mobilization of the pos-terior wall of the cardia. The hepatic branch is frequentlysacrificed when opening the gastrohepatic omentum.

Celiac plexus

The celiac ganglia and plexus are located below the originof the celiac artery. The exact location is variable. They aresituated between T12–L1 intervertebral disk and the middleof L2. The celiac plexus is formed by one to five gangliawith a diameter of 0.5 to 4.5 cm [53] and reaches the areaof the insertion of the diaphragmatic crura.

Damage to the celiac plexus during surgery in this area(particularly during dissection of the crura of the diaphragmin front of the aorta and during approximation of the crurain hiatal hernia repairs) does not appear to produce clini-cally significant effects. Injection of neurolytic solutionsinto this plexus has been used for pain control in carcinomaof the pancreas. The laparoscopic injection is performed byopening the ‘‘window’’ of the gastrohepatic ligament andretracting the lesser curvature of the stomach to the left. Theneurolytic solution is applied to the front and to the rightand left lateral periaortic tissues between the crura and theupper border of the pancreas.

Lymphatics

The cisterna chyli,when it is present, lies in front of thebody of the first and second lumbar vertebra between theright crus of the diaphragm and the aorta (Fig. 8). Fromthere, thethoracic ductadvances in a cephalad direction.Perforation of this duct may lead to chylous ascites. On theother hand, its ligation does not result in a significant prob-lem.

Gastrosplenic ligament

The membrane that joins the spleen with the greater curva-ture of the stomach is called the gastrosplenic ligament oromentum. There is a space between the anterior and poste-rior leaflets of the cephalad portion of this ligament, leavinga small segment of the posterior wall of the stomach un-covered by peritoneum. This area is situated immediately infront of the left crus of the diaphragm and may contain

890

vessels between the diaphragm, the pancreas, and the pos-terior wall of the stomach. To the left in this area is the leftadrenal gland. The gastrosplenic omentum contains theshort gastric vesselsin its upper portion and the left gas-troepiploic vessels as well as some branches from thesplenic artery in its most caudal area. There are lymphaticvessels and nodes all along the gastrosplenic omentum.

Gastrophrenic ligament

Cephalad to the gastrosplenic omentum is the gastrophrenicligament extending to the esophagus. This contains none orvery few vessels toward the esophagus, allowing a bluntdissection to approach the left and posterior wall of thedistal esophagus. The gastrophrenic ligament, however,contains many gastric and phrenic vessels extending for 2 to3 cm to the left of the angle of His until joining the shortgastric vessels to the spleen. This ligament joins the gastro-pancreatic ligament and upper border of the pancreas.

Gastropancreatic ligament

Seldom described under this name since it is formed by awide area containing theleft gastric vessels,the gastropan-creatic ligament separates the portion of the lesser sac lo-cated behind the gastrocolic omentum from the portion be-hind the posterior wall of the stomach. The only communi-cation between these two areas is an orifice caudal to thegastropancreatic ligament known as theomental foramen.The cephalad portion of the gastropancreatic ligament is lostin the retroperitoneum below the esophageal hiatus and be-hind the cardia. The right wall is covered by the peritoneumextending from the proximal part of the lesser curvature tothe right of the aorta and head of the pancreas. The leftboundaries continue with the posterior leaflet of the gastro-phrenic ligament.

The gastropancreatic omentum contains the left gastricvessels; the posterior gastric artery, which is a branch of theproximal segment of the splenic artery; lymphatic vesselsand nodes; the celiac division of the posterior vagus nerve;and small retroperitoneal, diaphragmatic, and pancreaticvessels.

The gastropancreatic and gastrophrenic ligaments maybe very elongated and may extend into the mediastinum inlarge hiatal hernias. The presence of thick, fatty, and vas-cular tissue ‘‘pulled’’ into the posterior mediastinum to-gether with the cardial end of the stomach may confuse thesurgeon when trying to identify and dissect the posteriorwall of the esophagus. The key to avoiding difficulties is tofirst dissect the cruraand then retract all tissue interposedbetween the junction of the right and left crus and theesophagus into an intraabdominal position.

Diaphragm

The central tendon of the diaphragm ends posteriorly at thelevel of the anterior portion of the crura. The tendinousportion gives way to muscle fibers on the sides of the crura.There are several diaphragmatic vessels in the vicinity ofthe esophageal hiatus. Theleft inferior phrenic arterycomes

from the celiac trunk or directly from the aorta (Fig. 9).Rarely it arises from the left gastric, left renal, or accessoryleft hepatic artery. It crosses between the left crus and theesophagus and distributes its branches to the left hemidia-phragm and the left adrenal gland. The left inferior phrenicartery may give branches to the esophagus and anastomosewith branches of the left gastric artery and thoracic esoph-ageal arteries. Attention should be given to these vesselswhen dissecting the esophagus away from the hiatus. Someof them may require clipping before transection.

The left inferior phrenic veinmay drain into the inferiorvena cava or the left adrenal vein (Fig. 10). In cases inwhich part or all of the drainage is into the vena cava, thisvein passes transversely in front of the hiatus, usually belowthe triangular ligament. The left inferior phrenic vein mayalso anastomose with esophageal branches of the left gastricvein and with veins of the greater curvature of the stomachthrough the gastrophrenic and gastropancreatic ligaments.

Congenital diaphragmatic defects should be recognizedduring laparoscopic procedures since they can be easily re-paired [30] and may lead to complications like pneumotho-rax if left uncorrected [8].

Transverse colon

In one particular congenital syndrome, the colon may belocated between the liver and the diaphragm. This is knownas Chilaiditi’s syndrome, and may be found associated withhypoplasia or absence of the left hepatic lobe as describedabove. The colon is frequentlyfound in the sacof largeparaesophageal hernias and in Morgagni hernias [27, 28,30].

Spleen

Although not in direct relation with the hiatus, except incases of splenomegaly, the spleen is an important organ toconsider in any laparoscopic dissection of the hiatus or itssurroundings because of the frequent risk of injury, tears ofits capsule, or bleeding from the many fragile vessels con-tained in its ligaments. The surgeon and assistants mustconstantly be aware of the position of the spleen and thedangers of exerting traction in the gastrosplenic, spleno-colic, and splenophrenic ligaments [45]. These ligamentsvary in size and site of insertion in the spleen.

The upper pole of the spleen is seen after dissecting thegastrophrenic ligament or during dissection of the left crus(Fig. 3). Trauma to the upper pole of the spleen should beavoided when passing instruments behind the esophagusfrom right to left. To avoid this, the instrument is advancedin front and just to the left of the left crus, the esophagus isthen retracted to the right, and the instrument is furtheradvanced under direct visualization.

The upper pole of the spleen may be in direct contactwith the lateral end of the left lobe of the liver and the lefttriangular ligament which sometimes surround it [4].

Pancreas

The pancreas is not in direct contact with the esophagealhiatus. Its upper border is caudal to the celiac trunk and the

891

caudal insertions of the crura are well behind the body of thepancreas. However, the pancreas may be traumatized duringdissection in the hiatal area, producing postoperative eleva-tion of serum amylase. The pancreas does not slide up intothe mediastinum in hiatal hernias.

Conclusion

Familiarity with the normal and pathologic laparoscopicanatomy is essential for safe surgical interventions in thearea of the esophageal hiatus. Training in the animal labo-ratory is technically valuable but may lead to excessiveconfidence due to the anatomic simplicity of the animalmodel.

Some laparoscopic complications, like pneumothoraxand esophageal and gastric perforations, rarely seen duringopen procedures, are usually the result of the difficultiesencountered in finding proper anatomic landmarks or theconsequences of struggles to provide adequate exposure fordissection.

This review of the laparoscopic anatomy of the area ofthe esophageal hiatus, based on published material and onpersonal experience, emphasizes the anatomic and patho-logic variations that the surgeon may encounter, the lapa-roscopic approach for dissection, and the technical maneu-vers that may help the surgeon avoid complications.

Acknowledgments.This work was supported in part by the Mr. and Mrs. A.Foroni and Family funds and by the Mr. and Mrs. M. C. Weiner funds.

References

1. Adams DB (1993) The importance of extrahepatic biliary anatomy inpreventing complications at laparoscopic cholecystectomy. Surg ClinNorth Am 73: 861–871

2. Aharinejad S, Firbas W (1989) The innervation of the human esopha-gus. Acta Anat 136:715–720

3. Ahmed AF, Bediako AK, Rai D (1988) Agenesis of the left hepaticlobe with gastric volvulus. NY State J Med 88: 327–328

4. Arenson AM, McKee JD (1986) Left upper quadrant pseudolesionsecondary to normal variants in liver and spleen. J Clin Ultrasound 14:558–561

5. Ashour M, Butchart EG, Breckenridge IM (1984) Phreno-oesophagealmuscle anomaly or rudimentary oesophageal sphincter? J R Coll SurgEdinb 29: 387–388

6. Bolondi L, Casanova P, Santi V, Caletti G, Barbara L, Labo G (1986)The sonographic appearance of the normal gastric wall: anin vitrostudy. Ultrasound Med Biol 12: 991–998

7. Caletti G, Bolondi L, Labo G (1984) Anatomical aspects in ultrasonicendoscopy for the stomach. Scand J Gastroenterol Suppl 94: 34–42

8. Childers JM, Caplinger P (1995) Spontaneous pneumothorax duringoperative laparoscopy secondary to congenital diaphragmatic defects.A case report. J Reprod Med 40: 151–153

9. Collet D, Cadiere GB (1995) Conversions and complications of lapa-roscopic treatment of gastroesophageal reflux disease. Formation forthe development of laparoscopic surgery for gastroesophageal refluxdisease group. Am J Surg 169: 622–626

10. Coster DD, Bower WH, Wilson VT, Butler DA, Locker SC, BrebrickRT (1995) Laparoscopic Nissen fundoplication—A curative, safe, andcost-effective procedure for complicated gastroesophageal reflux dis-ease. Surg Laparosc Endosc 5: 111–117

11. Crist DW, Gadacz TR (1993) Laparoscopic anatomy of the biliary tree.Surg Clin North Am 73: 785–798

12. Cuschieri A, Hunter J, Wolfe B, Swanstrom LL, Hutson W (1993)Multicenter prospective evaluation of laparoscopic antireflux surgery:preliminary report. Surg Endosc 7: 505–510

13. Dallemagne B, Weerts JM, Jehaes C, Markiewicz S (1996) Causes offailure of laparoscopic antireflux operations. Surg Endosc 10: 305–310

14. Dia A, Ouedraogo T, Zida M, Sow ML (1994) Anatomic pitfalls ofvagotomies. Ann Chir 48: 452–461

15. Erenbourg L, Reggiani P (1985) A new ligament: the ‘‘gastro-pancreatic.’’ Anat Clin 7: 143–144

16. Fontaumard E, Espalieu P, Boulez J (1995) Laparoscopic Nissen-Rossetti fundoplication. Surg Endosc 9: 869–873

17. Gray SW, Rowe JS, Skandalakis JE (1979) Surgical anatomy of thegastroesophageal junction. Am Surg 45: 575–587

18. Hay JM, Magdeleinat P, Regnard JF (1993) Positioning of the patientfor oesophageal hiatus surgery. Ann Chir 47: 80–83

19. Heloury Y, Leborgne J, Rogez JM, Robert R, Barbin JY, Hureau J(1988) The caudate lobe of the liver. Surg Radiol Anat 10: 83–91

20. Hinder RA, Filipi CJ, Wetscher G, Neary P, DeMeester TR, PerdikisG (1994) Laparoscopic Nissen fundoplication is an effective treatmentfor gastroesophageal reflux disease. Ann Surg 220: 472–483

21. Hunter JG, Trus TL, Branum GD, Waring JP, Wood WC (1996) Aphysiologic approach to laparoscopic fundoplication for gastroesopha-geal reflux disease. Ann Surg 223: 673–687

22. Jackson ML, Zuckier L, Goldfarb CR, Ongseng F (1986) Effect ofpatient positioning on liver size estimation. J Nucl Med 27: 1632–1634

23. Jacques PF, Warshauer DM, Keefe B, Mauro MA, McCall JM (1974)Variations in liver-colon anatomic relationship: relevance to interven-tional radiology. J Vasc Interv Radiol 5: 637–641

24. Joris JL, Chiche JD, Lamy ML (1995) Pneumothorax during laparo-scopic fundoplication: diagnosis and treatment with positive end-expiratory pressure. Anesth Analg 81: 993–1000

25. Kuster G, Biel F (1967) Accuracy of laparoscopic diagnosis. Am JMed 42: 388–393

26. Kuster GG, Fischer B (1993) Pharmacologic hemostasis in laparosco-py: topical epinephrine facilitates cholecystectomy. Am Surg 59: 281–284

27. Kuster GGR, Gilroy S (1993) Laparoscopic repair of paraesophagealhiatal hernia. Surg Endosc 7: 362–363

28. Kuster GGR, Gilroy S (1993) Laparoscopic technique for repair ofparaesophageal hiatal hernia. J Laparoendosc Surg 3: 331–338

29. Kuster GGR, Gilroy SBC (1995) Intraoperative trans-gallbladder chol-angiography intended to delineate bile duct anatomy. J LaparoendoscSurg 5: 377–384

30. Kuster GGR, Kline LE, Garzo G (1992) Diaphragmatic hernia throughthe foramen of Morgagni: laparoscopic repair case report. J Laparo-endosc Surg 2: 93–100

31. Lindner HH, Kemprud E (1971) Clinicoanatomical study of the arcu-ate ligament of the diaphragm. Arch Surg 103: 600–605

32. Listerud MB, Harkins HN (1958) Anatomy of the esophageal hiatus.Anatomic studies on two hundred four fresh cadavers. Arch Surg76:835

33. Mattioli S, Pilotti V, Felice V, Di Simone MP, D’Ovidio F, Gozetti G(1993) Intraoperative study on the relationship between the loweresophageal sphincter pressure and the muscular components of thegastroesophageal junction in achalasic patients. Ann Surg 318: 635–639

34. McKernan JB, Laws HL (1994) Laparoscopic Nissen fundoplicationfor the treatment of gastroesophageal reflux disease. Am Surg 60:87–93

35. Min PQ, Yang ZG, Lei QF, Gao XH, Long WS, Jiang SM, Zhou DM(1993) Peritoneal reflectons of left perihepatic region: radiologic-anatomic study. Radiology 182: 553–557

36. Oddsdo´ttir M (1996) Laparoscopic management of achalasia. SurgClin North Am 76: 451–458

37. Patti MG, Arcerito M, Pellegrini CA, Mulvihill SJ, Tong J, Way LW(1995) Minimally invasive surgery for gastroesophageal reflux dis-ease. Am J Surg 170: 614–718

38. Peters JH, Heimbucher J, Kauer WKH, Incarbone R, Bremner CG,DeMeester TR (1995) Clinical and physiologic comparison of laparo-scopic and open Nissen fundoplication. J Am Coll Surg 180: 385–393

39. Richards KF, Fisher KS, Flores JH, Christensen BJ (1996) Laparo-scopic Nissen fundoplication: cost, morbidity and outcome comparedwith open surgery. Surg Laparosc Endosc 6: 140–143

40. Rossetti M, Hell K (1977) Fundoplication for treatment of gastro-esophageal reflux in hiatal hernia. World J Surg 1: 439

41. Schauer PR, Meyers WC, Eubanks S, Norem RF, Franklin M, PappasTN (1996) Mechanisms of gastric and esophageal perforations duringlaparoscopic Nissen fundoplication. Ann Surg 223: 43–52

892

42. Scott-Conner CEH, Hall TJ (1995) Anatomic pitfalls of laparoscopicsurgery in the right upper quadrant. Contemp Surg 47: 307–314

43. Siewert JR, Jennewein HM, Waldeck F (1973) Experimental studieson the function of the lower esophageal sphincter following intratho-racic displacement, myotomy and circular myectomy. Bruns Beitr KlinChir 22: 818–828

44. Skandalakis LJ, Donahue PE, Skandalakis JE (1993) The vagus nerveand its vagaries. Surg Clin North Am 73:769–784

45. Skandalakis PN, Colborn GL, Skandalakis LJ, Richardson DD,Mitchell WE Jr (1993) The surgical anatomy of the spleen. Surg ClinNorth Am 73:747–768

46. Stein HJ, Feussner H, Siewert JR (1996) Failure of antireflux surgery:causes and management strategies. Am J Surg 171: 36–40

47. Stein HJ, Liebermann-Meffert D, De Meester TR, Siewert JR (1995)Three-dimensional pressure image and muscular structure of the hu-man lower esophageal sphincter. Surgery 117: 692–698

48. Swanstrom L, Wayne R (1994) Spectrum of gastrointestinal symptomsafter laparoscopic fundoplication. Am J Surg 167: 538–541

49. Swanstrom LL, Pennings J (1995) Laparoscopic esophagomyotomyfor achalasia. Surg Endosc 9: 286–292

50. Swanstrom LL, Pennings JL (1995) Safe laparoscopic dissection of thegastroesophageal junction. Am J Surg 169: 507–511

51. Trus TL, Laycock WS, Branum G, Waring JP, Mauren S, Hunter JG(1996) Intermediate follow-up of laparoscopic antireflux surgery. AmJ Surg 171: 32–35

52. Van Minh T, Galizia G, Lieto E (1992) Anatomy of the caudate lobeof the liver. New aspects and surgical applications. Ann Chir 46:309–318

53. Ward EM, Rorie DK, Nauss LA, Bahn RC (1979) The celiac gangliain man: normal anatomic variations. Anesth Analg 58: 461–465

54. Watson DI, Jamieson GG, Devitt PG, Matthew G, Britten-Jones RE,Game PA, Williams RS (1995) Changing strategies in the performanceof laparoscopic Nissen fundoplication as a result of experience with230 operations. Surg Endosc 9: 961–966

893

The author replies

We have read with care the letter to the editor of ProfessorBax concerning the intraperitoneal use of cetrimide and theunfortunate death of this 2-year-old patient. We do feel,with the author, that this complication may be a cetrimide-related toxicity; however, we would like to clarify severalissues related to the clinical use of cetrimide:

The diagnosis in this 2-year-old boy was not docu-mented by a CT scan, which gives the typical appearance ofa cyst with multiple daughter cysts, which is pathognomon-ic of hydatid disease [2]. Also, there was no mention of theindirect hemagglutination test or any other hydatid studies,which usually are positive. Hydatid cysts at this age are rare,and other possibilities, including hepatoblastoma may beconsidered.

We have reported in our experience [5, 6] that only cystssurfacing to the peritoneal cavity should be treated laparo-scopically. The cyst is punctured in the fibrous capsulerather than in the liver tissue, cetrimide or hydatid fluid indirect contact with bloodstream [3], or highly vascular areaslike liver or epidural space [7] could result in anaphylacticshock. Especially in this 2-year old boy, whose estimatedblood volume is 800–1000 ml. The cetrimide was left in theperitoneal cavity for 1 h; 200 ml of this fluid was not re-trieved and could have been absorbed. Concerning the liverpuncture, it was not clear whether cetrimide was injectedinadvertently into the liver. The findings suggest an obviouscase of cetrimide overdose, although not documented bymethemoglobinemia [1]. In our described technique the ce-trimide instillment does not exceed 1,000 ml in adults; thehydatid fluid aspirated from the cyst is also replaced byequal amounts of cetrimide. In a vast experience with morethan 1,000 cases in our institution [4] of using cetrimidesince 1972, no cetrimide overdose has been reported and it

is a safe practice to use 1% cetrimide, which is not toxic inthis concentration [8]. We conclude that this is an obviouscase of cetrimide overdose in a 2-year-old child, but we stillconsider the use of cetrimide 1% a safe practice provided itis not injected in highly vascular areas or in the bloodstreamand is used in moderate amounts for a short period of timein the peritoneal cavity. Needless to say, it should be used inhydatid asepsis after documentation of hydatid disease.

References

1. Baraka A, Yamut F, Walid N (1980) Cetrimide induced methaemoglo-binaemia after surgical excision of hydatid cyst. Lancet 11: 88–89

2. Beggs I (1985) The radiology of hydatid disease. AJR Am J Roentgenol145: 639–48

3. Gode GR, Jayalaksmi TS, Kalla GN (1975) Accidental intravenousinjection of cetrimide. Anesthesia 30: 508–510

4. Jidejian YD (1979) Hydatid disease. Dar El-Mashreq, Beirut, Lebanon5. Khoury G, Geagea T, Hajj A, Jabbour-Khoury S, Baraka A, Nabbout G

(1994) Laparoscopic treatment of hydatid disease of the liver. SurgEndosc 8: 1103–1104

6. Khoury G, Jabbour-Khoury S, Bikhazi K (1996) Results of laparoscopictreatment of hydatid cyst of the liver. Surg Endosc 10: 57–59

7. Klouche GR, Charlotte N, Kaaki M, Beraud JJ (1994) Coma hemolysisafter cetrimide washout of epidural hydatid cyst. Intensive Care Med20: 613

8. Martindale (1996) The extra pharmacopocia. 22nd ed. The Pharmaceu-tical Press, London, p 1122

G. KhouryS. Jabbour-KhouryK. Bikhazi

Department of SurgeryAmerican University of BeirutMedical CenterP.O. Box 113-6044Beirut, LebanonCorrespondence to:G. Khoury

Surg Endosc (1997) 11: 969

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Letters to the editor

Laparoscopic bariatric surgery

We read with interest the article ‘‘Laparoscopic gastric by-pass: another option in bariatric surgery’’ by Lo¨nroth etal. [Surg Endosc(1996) 10: 636–638]. The gastric bypasstechnique which they describe (loop gastrojejunostomy,with afferent-efferent jejunojejunostomy for diversion ofbile and pancreatic juice) is acceptable. However, thismethod has been largely replaced over the past 15 years bythe Roux-en-Y gastric bypass. The Roux-limb facilitatestechnically the movement of the jejunum to a high position.However, with the advent of, enthusiasm for, and progres-sive application of video-assisted laparoscopic surgery, cau-tion must be advised with respect to bariatric surgery. Theauthors mention that laparoscopic surgery ‘‘should ofcourse be based on experiences already reached in the areaof open surgery,’’ but this statement demands further em-phasis.

In 1993, Robert J. Greenstein of Mount Sinai MedicalSchool and Joseph F. Amaral of Brown University spear-headed a move to formulate basic guidelines regarding theapplication of laparoscopic techniques to obesity surgery inorder to avoid disastrous consequences. This resulted in therecommendation by the American Society for Bariatric Sur-gery (ASBS) that ‘‘Laparoscopic obesity operations shouldbe undertaken only by surgeons who are experiencedbothin video-laparoscopic techniqueand in the complexities ofopen bariatric operations and the field of morbid obesity.’’This recommendation has been circulated by the ASBS tothe chiefs of surgery in the USA and Canada.

Bariatric surgery has become a specialized field in gen-eral surgery. Expertise is necessary in dealing with the mas-sively obese patient population, in their selection and co-morbidities, in the specific measurements required to avoidfailure, and in the postoperative treatment of sequelae andnutrition. In the event that a laparoscopic bariatric operationhas to be converted, experience in the open abdominal sur-gery is required.

We see a parallel here to the action of SAGES, whenlaparoscopic cholecystectomy catapulted into the surgicalrealm 6 years ago. SAGES then issued a statement with thelogical recommendation that laparoscopic cholecystectomybe performed only by surgeons experienced in open chole-cystectomy. This recommendation helped to avoid majorcomplications which would have devalued early video-laparoscopic surgery; it also deterred a number of medicalgastroenterologists experienced in laparoscopy who wereanxious to undertake laparoscopic cholecystectomies.

SAGES was instrumental in setting criteria for creden-

tialling laparoscopic surgery in hospitals and provided manycomprehensive courses by experts. In like fashion, sessionson laparoscopic bariatric surgery have been an integral partof the ASBS meetings for the past 4 years.

Lonroth, Dalenba¨ck, Haglind, and Lundell fromGothenburg are experienced bariatric abdominal surgeonsand laparoscopists [7]. However, there are learning curveswith laparoscopic gastric banding [1, 5] vertical bandedgastroplasty [2, 3], Roux-en-Y gastric bypass [9], and ileo-gastrostomy [4]. These procedures have been done by de-voted individuals and groups experienced in both the openand laparoscopic approaches, thus avoiding catastrophesand setbacks for bariatric surgery. The laparoscopic proce-dures, where exposure and the ability to measure may beless optimal in the massively obese, must be shown to be atleast as effective and safe as the open procedures. A numberof groups are using the laparoscopic adjustable gastric bandin a controlled study, but further evaluations are necessary[6, 8].

References

1. Broadbent R (1993) Lessons learned from laparoscopic gastric banding.Obes Surg 3: 415–420

2. Catona A, Gossenberg M, Mussini G, La Manna L, De Bastiani T,Armeni E (1995) Videolaparoscopic vertical banded gastroplasty. ObesSurg 5: 323

3. Chua TY, Mendiola RM (1995) Laparoscopic vertical banded gastro-plasty: the Milwaukee experience. Obes Surg 5: 77–80

4. Cleator IGM, Litwin DEM, Phang PT, Brosseuk DT, Rae AJ (1994)Laparoscopic ileogastrostomy. Obes Surg 4: 40–43

5. Favretti F, Cadiere GB, Segato G, Bruyns G, De Marchi F, Himpens J,Foleto M, Lise M (1995) Laparoscopic adjustable silicone gastric band-ing: technique and results. Obes Surg 5: 364–371

6. Greenstein RJ, Rabner JG, Green S, Hodge-Penn G, Kaiser S, HalpenNA (1996) The laparoscopic adjustable gastric band (Lap-band): ob-servations from the preliminary Mount Sinai experience of a nation-wide multi-center, FDA-moderated study. Obes Surg 6: 124 (abstract)

7. Lonroth H, Dalenba¨ch J, Haglind E, Lundell L (1995) Laparoscopicvertical banded gastroplasty: 1-year follow-up. Obes Surg 5: 249 (ab-stract)

8. O’Brien P, McMurrick P (1995) Posterior gastric wall prolapse afterLap-band placement. Obes Surg 5: 247 (abstract)

9. Wittgrove AC, Clark GW (1996) Laparoscopic gastric bypass, Roux-en-Y: experience of 27 cases, with 3–18 months follow-up. Obes Surg6: 54–57, 329

M. Deitel

Department of Nutritional Sciences and SurgeryUniversity of Toronto2238 Dundas Street West, Suite 201Toronto, Ontario M6R 3A9, Canada

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 965

Erroneous CO2 pressure readings in a morbidly obese patient

Since the initial reports in 1992 [2] laparoscopic splenec-tomy has become an accepted technique for patients withidiopathic thrombocytopenic purpura (ITP) [4]. Simulta-neously, increasing experience with laparoscopic tech-niques has broadened patient selection criteria to includeobses patients [6], who had previously been excluded. Ex-perience has also dictated alterations in patient positions andtrocar sites to facilitate exposure and manipulation of thespleen, resulting in decreasing operative times [9]. Refine-ments in technology, such as the harmonic scalpel, havealso facilitated dissection and hemostasis [7]. With eachrefinement, however, new potential problems can arise andmust be identified. In the case reported here the combinationof lateral position and morbid obesity presented the poten-tial for unrecognized overinsufflation of CO2.

A 37-year-old man with ITP inadequately responsive tomedical therapy was referred for splenectomy. The patientwas morbidly obese, weighing 318 lb on a 5’10’’ frame.Due to anticipated exposure problems with either open orsupine laparoscopic approaches, the lateral or ‘‘hangspleen’’ laparoscopic approach was selected [1]. The patientsize limited the usual table rotation position [3] and he wasin almost a true lateral position. The Hassan open techniquewas utilized for placement of a supraumbilical trocar withthe blunt finger technique of Filipi [5] used to access theperitoneal cavity due to the difficult exposure.

Following trocar placement and confirmation of posi-tion by endoscopic visualization of omental tissue the lapa-roscope was withdrawn and CO2 insufflation was initiated.Insufflation was at 6 l per min via a Stryker high-flowinsufflator with a maximum pressure setting at 15 cm ofwater. It was noted that although several liters had beeninstilled, the abdominal pressure reading remained zero. Al-though our initial thought was a loose tubing connection,the scope was immediately reinserted to check for position.This afforded visualization of fatty tissue flapping over theend of the trocar like a ‘‘flutter valve’’ mechanism. Whenthe scope was advanced beyond this fatty tissue a largepneumoperitoneum was visualized and the pressure readingjumped to 18 cm of water. This was immediately released tothe proper pressure reading of 15. The remainder of thesplenectomy was completed uneventfully using a four-trocar technique, the 30° laparoscope, and the harmonicscalpel. Blood loss was negligible and a 72-g spleen wasmorsalated and removed via an 18-mm trocar site (theweight is artificially decreased due to blood aspiration dur-ing the morsalation process). The total procedure time was118 min. The patient did well postoperatively and was dis-charged on the 2nd postoperative day.

While for the experienced laparoscopic surgeon opera-tions in morbidly obese patients can be simplified relative tothe open technique due to better visualization and depth ofaccess, other problems can arise. In this case, I believe thepatient’s morbid obesity combined with the lateral positionand midline trocar site resulted in intermittent occlusion ofthe trocar tip to the pressure measurement while allowingcontinued insufflation. If this situation is not recognized,CO2 overinsufflation with dire consequences can result [8].The patient’s obesity further obscured the problem by lim-iting visual external inspection of the abdomen. Our scopeis often removed during initial insufflation to allow betterflow and to facilitate setup of other equipment. In a mor-bidly obese patient continuous visualization during insuf-flation appears safer.

Special precautions should be taken during insufflationof CO2 via a midline trocar with a morbidly obese patient inthe lateral position. Low pressure readings should not beassumed to be accurate as a flutter valve effect of tissue overthe trocar can occur. Visual endoscopic inspection duringinsufflation is indicated and pressure tubing should bechanged to a more lateral trocar when this site is available.

References1. Delaitre B (1995) Laparoscopic splenectomy: the ‘‘hanged spleen’’

technique. Surg Endosc 9: 528–5292. Delaitre B, Maignien B (1992) Laparoscopic splenectomy—technical

aspects. Surg Endosc 6: 305–3083. Dexter SPL, Martin IG, Alao D, Norfolk DR, McMahon MJ (1996)

Laparoscopic splenectomy: the suspended pedicle technique. Surg En-dosc 10: 393–396

4. Grossbard ML (1996) Is laparoscopic splenectomy appropriate for themanagement of hematologic and oncologic diseases? Surg Endosc 10:387–388

5. Laparoscopic Surgery Update. May 1996: Vol. 4, No. 5: 49–606. Robles AE, Owens CC, Bianchi C, Pitombo C, Garberoglio C, Ta-

buenca A (1996) Outpatient laparoscopic cholecystectomy is safe forthe obsese and morbidly obese patient. SAGES Meeting; Poster Ab-stract; 205

7. Rothenberg SS (1996) Laparoscopic splenectomy using the harmonicscalpel. J Laparoendosc Surg 6: S61–S62

8. Smith I, Benzie RJ, Gordon NLM, Kelman GR, Swapp GH (1971)Cardiovascular effects of peritoneal insufflation of carbon dioxide forlaparoscopy. Br Med J 3: 410–411

9. Trias M, Targarona EM, Balague´ C (1996) Laparoscopic splenectomy:an evolving technique. Surg Endosc 10: 389–392

K. D. Stalter

449 Main StreetOneonta, NY 13820USA

Surg Endosc (1997) 11: 970

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

Original articles

Short-term outcome of laparoscopic paraesophageal hernia repair

A case series of 58 consecutive patients

T. R. Huntington

Department of Surgery, St. Luke’s Regional Medical Center Department of Surgery, Boise Veteran’s Affairs Medical Center, 222 N. 2nd Street, Suite107, Boise, ID 83702, USA

Received: 24 July 1996/Accepted: 20 November 1996

AbstractBackground:The purpose of this study is to determine themorbidity, mortality, and short-term outcomes associatedwith laparoscopic paraesophageal hernia repair (LPHR).Methods:A series of 58 consecutive LPHRs performed bythe author were reviewed with an average 1-year follow-up.Morbidity and mortality rates were compared with historicalseries of open repairs. Anatomy and technical consider-ations pertinent to LPHR were reviewed.Results:There were no procedure-related or perioperativedeaths in this series of patients undergoing LPHR. Fourmajor complications occurred (7%), two of which requiredreoperation, all in urgently repaired patients. One patientrequired conversion to laparotomy (1.7%). Based on symp-toms, there were no reherniations. No patients had long-term dysphagia worse than preoperatively. Preoperativesymptoms of chest pain, esophageal obstruction, hemor-rhage, and reflux were resolved in all patients.Conclusions:LPHR is safe, effective, and compares favor-ably to historical series of open paraesophageal hernia re-pair.

Key words: Paraesophageal hernia — Laparoscopy — Lap-aroscopic — Hiatal hernia — Gastroesophageal reflux —Fundoplication

Numerous case reports of laparoscopic paraesophageal her-nia repairs (LPHRs) appearing in the literature [1–3, 7–9,11, 14] have confirmed that the procedure is technicallyfeasible and safe. This report describes the author’s series of58 consecutive LPHRs. The overall morbidity and mortalityin this series were compared with large historical series ofparaesophageal hernia repair (PHR) by laparotomy [4, 5, 10,18] or thoracotomy [13]. The large case series reported hereprovides data regarding technique, morbidity, mortality, and

short-term outcome. This is the largest series of laparoscop-ic repairs reported to date.

Materials and methods

Patient population

Between August 1992 and October 1995, 218 of the author’s patients fromSt. Luke’s Regional Medical Center and the Boise Veteran’s Affairs Medi-cal Center had laparoscopic repair of hiatal hernias. Twenty-six of thesewere operated on for symptoms or radiographic findings of a paraesopha-geal hernia (chest pain, dysphagia, esophageal obstruction, hemorrhage,and aspiration). Of the remaining 187 who had a chief complaint of gas-troesophageal reflux, 32 of these were identified as having a significantparaesophageal hernia (PH) either preoperatively or at the time of surgery.The referral base was biased in favor of paraesophageal hernias and largehiatal hernias due to somewhat selective referral of this type of hernia.

A PH was defined as a peritoneal lined defect through the esophagealhiatus typically containing at least one-quarter to one-third of the stomachwithin the hernia sac. The most superior portion of the fundus was locatedsuperior to the gastroesophageal junction.

All patients had preoperative esophagogastroduodenoscopy (EGD) per-formed by various physicians. Only 38% were specifically noted as havinga PH on EGD. The remainder were described as having a large or hugehiatal hernia, sometimes associated with a description of torsion, obstruc-tion, or complete intrathoracic stomach. One patient, not included in the 58cases reported here, had an incorrect preoperative diagnosis of PH based onsymptoms and EGD; he underwent fundoplication for reflux symptoms.

Of those patients repaired primarily for symptoms of PH, 82% hadpreoperative upper gastrointestinal series (UGI) performed at various in-stitutions. Seventy percent were interpreted by their radiologist as havinga PH. The remainder were described as having hiatal hernias that werehuge, large, completely or largely intrathoracic, or in some as being par-tially torsed. One patient with a PH demonstrated on EGD and confirmedat surgery had an UGI interpreted as normal.

Esophageal manometry was done in elective cases when technicallyfeasible (43 cases). Four of these had dysmotility demonstrated.

All 58 patients underwent laparoscopic repair, although one requiredconversion to laparotomy. During the 3 years covered by this study, nopatients seen by the author for gastroesophageal reflux or paraesophagealhernia were denied laparoscopic repair in favor of repair by laparotomy.

Ten patients in this series underwent urgent repair.

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 894–898

Five of the 58 patients had had previous failed fundoplication or para-esophageal repair, four by laparotomy and one by laparoscopy.

The average follow-up was 12 months (383 days) and ranged from 6weeks to 36 months. An attempt was made to telephone or examine allsurvivors between January and March 1996, during which time 45 patientswere successfully interviewed or examined by the authors. Data on theremaining 13 patients, including five deceased, were taken from the mostcurrently available medical records. Patients were evaluated for symptomsof recurrent herniation, gastroesophageal reflux, dysphagia, hemorrhage,and complications related to surgery.

Technique

The surgical approach utilized a standard five-port placement consisting ofone 5-mm subxyphoid port and four 11-mm ports placed in a near straightline crossing the upper abdomen approximately 10 cm below the xyphoid.The patient was supine with the surgeon standing on the right utilizing thetwo midline ports. The assistant, or in later cases, the robotic arm (AESOP,Computer Motion, Goleta, CA), was on the patient’s left. Gentle tractionwas applied to the hernia contents, reducing it as much as possible. Themost superior aspect of the fundus, which was retroperitoneal, was fre-quently not reducible. The peritoneum was divided at the hiatus anterior tothe esophagus and stomach except in reducible hernias, in which case thesac was divided adjacent to the cardioesophageal junction. Short gastricvessels were found stretched and frequently elongated over the left side ofthe hiatus and did not require division. The gastrohepatic ligament wasdivided over the caudate lobe of the liver exposing the lesser sac.

It is important to recognize that in a large PH the lesser sac extends intothe mediastinum. The mediastinal extension of the lesser sac, or lesser sachernia, is bounded anteriorly by the gastrohepatic ligament, the herniatedstomach, and the herniated gastrosplenic ligament. In extreme forms thesestructures form a transverse plane dividing the hiatus hernia into two com-partments, the greater and lesser sac. Figure 1 illustrates a cross section ofthe esophageal hiatus demonstrating the two components of the hernia.

The lesser sac hernia into the mediastinum was the key to easy reduc-tion of the paraesophageal hernia. After entering the lesser sac through thegastrohepatic ligament, the most posterior layer of lesser sac peritoneumwas identified and entered where it was draped over the left and right crurato the right of and posterior to the esophagus. Further dissection wasundertaken posterior and external to the lesser sac (Fig. 2). This allowsreduction of the lesser sac and coincidentally mobilizes the posterioresophagus and herniated stomach.

The greater sac hernia was approached through the same plane, ergo,from posteriorly and to the right, again keeping external to the sac. This

plane of dissection was utilized to remove the entire sac and the slidingportion of the fundus. The entire sac and stomach were thus reduced.

The esophagus was mobilized as far as necessary for adequate length,sometimes past the point at which the esophagus passes posterior to theposterior aspect of the aortic arch. Of note, all patients had adequate esoph-ageal length, although some required extensive mobilization of the esopha-gus. The hernia sac was excised, or at a minimum was incised down to thecardioesophageal junction.

Crural repair was performed by placing sutures between the right andleft crura posterior to the esophagus and, if needed, anterior to the esopha-gus. In the event of excessive tension on the crura, a relaxing incision wasmade in the right crus. The relaxing incision defect was covered with meshaffixed by hernia staples.

Following this, a Nissen or Toupet fundoplication completed the pro-cedure in all but two cases. Short gastric vessels were usually stretched andelongated by the herniated fundus. These were divided only in the rareinstance that there was tension on the fundoplication.

A Toupet plication was chosen in favor of a Nissen plication in thosepatients with esophageal dysmotility demonstrated on manometry (peakmean body amplitude <40 mm Hg or less than 50% primary peristalsis). Inlater cases a Toupet was used in all patients who did not have preoperativemanometry.

Although no problems were encountered with esophageal damage fromnasogastric tubes or dilators, increasing awareness of this possible com-plication strongly militated against intraoperative instrumentation of theesophagus. If there was difficulty identifying the esophagus, a Malonydilator was cautiously introduced by the anesthesiologist while the surgeonstraightened the esophagus and esophagogastric-gastric junction as muchas possible. Tightness of the warp was judged by unimpeded passage of aBabcock clamp between the wrap and the esophagus.

If there was any question about esophageal or gastric integrity at anypoint in the procedure, 250 ml of methyl-blue-stained saline was intro-duced by NG tube in an attempt to identify any leakage.

Results

All patients

Fifty-eight patients with a mean age of 67.3 years, range of24 to 90 years, underwent LPHR over a 40-month period(Table 1). Fifty-seven had successful completion of the pro-

Fig. 1. Transverse view of esophageal hiatus showing the diaphragmaticcrura, herniated gastroesophageal junction, and the herniated lesser omen-tal sac posterior to the gastrohepatic ligament and stomach.C, caudate lobeof liver; GH, gastrohepatic ligament;E, esophagus;F, fundus;Sp,spleen.

Fig. 2. Sagittal view through the mediastinum showing the paraesophagealhernia with lesser sac hernia posterior to fundus (F ). The initial dissectionof the hernia is through the plane posterior and external to the lesser(posterior) hernia sac.St,stomach;E, esophagus.

895

cedure laparoscopically; one patient required conversion toa laparotomy. All but two patients had either Toupet (eight)or Nissen (48) fundoplication. One patient had reductionand excision of the sac with crural repair only. Anotherpatient, who had end-stage Alzheimer’s dementia, had re-duction of her torsion and gastrostomy only. She died 6weeks postoperatively.

Average total length of stay for all patients was 2.8 days.There were four major complications, two of which re-

quired laparoscopic reoperation for esophageal obstruction.One patient had a pulmonary embolism (survived) and onepatient required a feeding gastrostomy for esophageal dys-motility and central dysphagia. All four major complica-tions were in urgently repaired patients.

There were no procedure-related deaths, although therewere four early deaths from unrelated causes: two of Alz-heimer’s dementia, one of ovarian carcinoma, and one oflung carcinoma at1⁄2 months, 6 months, 2 months, and 6months from surgery, respectively. One additional patientdied of a myocardial infarct 13 months post-LPHR.

At a mean of 12 months there were no known recur-rences or symptoms of recurrence.

Gastroesophageal reflux

Forty-six patients had complained of gastroesophageal re-flux symptoms preoperatively. All of these patients under-went fundoplication and reported resolution of reflux symp-toms postoperatively.

Total intrathoracic stomach PH

A total of 12 patients had herniation of the entire stomachinto the chest; six were repaired urgently. These six urgentlyrepaired patients with intrathoracic stomach had three of thefour major complications encountered in this series.

Elective LPHR

All 48 patients undergoing elective repair had either Nissenor Toupet fundoplication. Forty-two (87%) of these hadsymptoms of gastroesophageal reflux (GER) preopera-tively. There were no major complications in the patientsundergoing elective repair. Mean operative time was 115min and average total length of stay (LOS) was 1.9 days.

Minor complications consisted of urinary retention, a singleepisode of congestive heart failure requiring an additional 2days of hospitalization, and one patient requiring esophage-al dilation for dysphagia at 2 months postoperative. Milddysphagia lasting 4 weeks or, rarely, up to 12 weeks, wascommon. One patient with normal esophageal motility hadrepair of an epiphrenic diverticulum 6 months postopera-tively. This was either missed or not present on the initialEGD. No splenic injuries were encountered.

Previous hiatal hernia repair

As noted above, five patients had previous failed hiatalhernia repairs (one laparoscopic and four by laparotomy)performed by other surgeons. The patient with a previousfailed laparoscopic repair had an aborted attempt at repairwithout laparotomy. All normal tissue planes were obliter-ated, requiring identification of the esophagus by manualpalpation. Those patients with a previous failed repair bylaparotomy had extensive adhesions in the upper abdomenand mediastinum, making the cause of failure difficult todetermine. One of these patients had an essentially un-touched left crus, suggesting that the left crus was not prop-erly identified at the first procedure.

Urgently repaired LPHR

Of the 10 patients who underwent urgent LPHR, two wereoperated on for pain suggestive of incarceration or strangu-lation; one of these was strangulated but not infarcted. Eighthad esophageal obstruction; four of these had aspirationpneumonia.

Only one of the 10 underwent preoperative manometricstudies. This patient had an uneventful postoperative course.

Two patients with torsed total intrathoracic stomach PHdeveloped postoperative esophageal obstruction with aspi-ration pneumonia. One of these, who did not undergo fun-doplication, had infarcted the remnants of the excised herniasac. This was corrected by dividing the remnant of the sacdown to the serosa/muscularis at the cardioesophageal junc-tion at the angle of His. The other patient had a Nissenfundoplication and developed edema in the area of the wrapand cardioesophageal junction. This was converted to aToupet repair with prompt relief of obstruction. Both reop-erations were done laparoscopically; they were dischargedat 5 and 6 days following the second procedure.

The additional two complications in this group of ur-gently repaired PHs were a pulmonary embolism at 3 weekspostoperative (survived) and one patient with dysphagia re-quiring a postoperative percutaneous endoscopic gastros-tomy (PEG). This patient with Alzheimer’s and a total in-trathoracic stomach hernia was unable to swallow preopera-tively. Postoperatively she was able to swallow much betterbut still required supplemental tube feedings. Her dysphagiawas from dysmotility demonstrated on esophagogram andwas compounded by a failure to eat secondary to Alzhei-mer’s dementia. No evidence of mechanical obstruction waspresent; she died of complications of Alzheimer’s dementia4 months postoperatively. Average LOS for these urgentlyrepaired patients, including preoperative stay, was 7.4 days.

Table 1.Results of current series of 58 LPHR

ElectiveLPHR

UrgentLPHR Total

% oftotal

Perioperative mortality 0 0 0 0Morbidity 0 4 4 7%Length of staya 1.9 7.4 2.8Recurrence 0 0 0 0Previous failed repair 3 2 5 9%Total intrathoracic stomach 6 6 12 21%Preoperative GER 42 4 46 79%Postoperative GER 0 0 0 0Total patients 48 10 58 100%

a Includes preoperative stay

896

Discussion

Elective LPHR had a very low morbidity with no majorcomplications. Although follow-up is fairly limited (12months), there have been no known recurrences. GER hadbeen either cured or greatly improved in all cases. Electivehospital stay is commensurate with other laparoscopic pro-cedures.

Large series of PHR infrequently included morbidityrates or length of stay. Morbidity, when given, sometimesuses a different perspective than we presently use. Althoughcarefully reported, Ellis [4] did not consider six splenecto-mies as morbidity and Menguy [10] did not consider 2weeks of pseudo-obstruction or a gastric fistula as majorcomplications. Comparison of the author’s series to histori-cal series must be kept in perspective. Open PHR morbidityrates averaged 14% [4, 5, 10, 13, 18] and mortality averaged3% [4, 5, 10, 12, 13, 17, 18]. Length of stay, when noted,averaged 10.1 days [4, 5] for open repairs. LPHR in theauthor’s series had a morbidity rate of 7%, no mortality, andan average LOS of 2.8 days (Table 2).

Three of the four morbid events in the author’s serieswere esophageal obstruction or dysphagia, all occurring inthe group of urgently repaired patients. These three wereprobably to some extent preventable. Two occurred intorsed PH with total intrathoracic stomach undergoing Nis-sen fundoplication who did not have preoperative motilitystudies. From this experience, it is clear that if a fundopli-cation is used in urgent repairs without preoperative ma-nometry, it should be a Toupet plication. The third patient,who did not have a fundoplication, was a failure to fullydivide the remnants of the hernia sac. The 1–2-cm wide rimof sac infarcted and obstructed the cardioesophageal junc-tion. Division of this rim down to the muscularis/serosa ofthe cardioesophageal junction is mandatory. Patients with-out preoperative esophageal manometry should be consid-ered for Toupet fundoplication.

The true definition of PH as been discussed by nearlyevery paper written on PH. Laparoscopy may offer someadvantage over laparotomy in establishing the diagnosis ofPH. The mere presence of a peritoneal defect adjacent to theesophagus does not presently qualify as a true PH, althoughit is hard to call this a pure sliding hernia. This type of defectis probably much more noticeable through a laparoscope for

two reasons. First, exposing the hiatus by laparotomy usu-ally involves retracting the stomach inferiorly, reducingsmall paraesophageal peritoneal defects prior to visualizingthem. Second, positive-pressure inflation during laparosco-py should make potential defects more prominent. In thisstudy only patients with herniation of significant portions ofintraperitoneal stomach in which the fundus was superior tothe cardioesophageal junction were considered PH [14].

Preoperative UGI and EGD often described typical find-ings of PH but failed to actually label the diagnosis as a PH.These studies were done by numerous physicians and sug-gest that the definition and ramifications of PH are notuniversally understood.

Whether or not to plicate PHs remains controversial.This series and some others [9, 13, 16, 17] have a highincidence of preoperative GER. Treacy [17], Pearson [13],and Styger [16] advised fundoplication in all PHR. Ana-tomically this makes sense. The diameter of the distalesophagus with associated wrap is significantly larger thanthe repaired hiatus and should help prevent subsequent re-herniation. Also of concern is the rearrangement of anatomyresulting from PHR, which may allow reflux to developwhen it was not formerly present. Other large series [4, 9,10, 15, 16] conclude that there is a significant (10% overall)incidence of GER developing in patients who do not haveantireflux procedures and presumably did not have preop-erative GER (Table 3). Postoperative GER appears in sev-eral studies [4, 9, 10, 15, 16] as one of the sources ofpostoperative morbidity.

With the exception of full wraps in acute torsions, theaddition of fundoplication did not appear to add to morbid-ity in LPHR. Technically, fundoplication was simple andstraightforward, the majority of dissection having been al-ready completed during reduction of the hernia. The gastricfundus was uniformly floppy, likely because of elongationof the gastrosplenic ligament. Short gastric vessels wereusually not divided. Fundoplication carries a low risk witha significant potential benefit in LPHR.

Although only eight patients had preoperative esopha-geal obstruction, experience with them strongly suggeststhat patients with complete esophageal obstruction shouldbe repaired promptly in spite of any underlying pneumonia.Pulmonary status and SaO2 improved dramatically within

Table 2. Results of previous paraesophageal hernia repair by laparotomy and thoracotomy

Totalpatients Emergent Morbidity Mortality

Lengthof stay Recurrence

Ellis [4] 51 —a 13b 1 9.5 4Haas [5] 21 10 4 4 11.4 0Hill [6] 22 —a —a —a —a 0Menguy [10] 30c 12 2 0 —a 0Ozdemir [12] 31 12 —a 2 —a 3Pearson [13] 53 1 5 1 —a 0Treacy [17] 53 18 —a 2 —a —a

Williamson [18] 119 7 14 2 —a 13Averaged 46 20% 14% 3% 10.1 6%This study 58 17% 7% 0 2.8 0

a —Not statedb Does not include six splenectomiesc All total intrathoracic stomachd Percentage of those stated

897

hours of repair in these patients. Delaying surgery to correctmalnutrition almost inevitably resulted in the patient devel-oping aspiration pneumonia.

The use of a laparoscope to visualize the hiatus andmediastinum allows visualization and appreciation of ana-tomical details that are difficult to appreciate at laparotomy.Dissection of the hernia sac and subsequent reduction of thestomach is straightforward and relatively easy if theanatomy is understood. The dissection into the mediastinumshould be started by incising the posterior peritoneum of thelesser sac. The lesser sac is dissected free and reduced fromexternal to the sac. The greater sac hernia is likewise ap-proached externally. The esophagus is easily identified andmobilized by this technique.

The low morbidity and low mortality of LPHR, as notedin this study, are significant in that old and debilitated pa-tients who had been refused open repair were included inthis series. The low morbidity of LPHR should allow repairin patients formerly thought to be too risky for PHR bylaparotomy or thoracotomy. Although many surgeons ad-vocate repair of all PH, it is apparent that there are manyphysicians taking the view advocated by Treacy [17] ofwatchful waiting in elderly or debilitated patients. The lowmorbidity of LPHR has convinced many of our local phy-sicians that early repair should be undertaken. Indeed, it isa change in this threshold that has contributed to the largenumbers of LPHR in this series.

As more and more laparoscopic antireflux proceduresare done by increasing numbers of surgeons, we need to be

aware of and familiar with techniques for LPHR. Frequentconversion to laparotomy is denying these patients a proce-dure of lower morbidity. One-half of the procedures in thisseries were encountered in patients with a primary diagnosisof gastroesophageal reflux. Surgeons attempting laparo-scopic fundoplication should have adequate experience andexpertise to consistently complete these procedures laparo-scopically. Although long-term follow-up is needed, thelaparoscopic approach should be strongly considered in pa-tients undergoing PH repair.

References

1. Cloyd DW (1994) Laparoscopic repair of incarcerated paraesophagealhernias. Surg Endosc 8: 893–897

2. Congreve DP (1992) Laparoscopic paraesophageal hernia repair. JLaparoendosc Surg 2: 45–48

3. Cuschieri A (1993) Laparoscopic antireflux surgery and repair of hia-tal hernia. World J Surg 17: 40–45

4. Ellis FH, Crozier RE, Shea JA (1986) Paraesophageal hiatus hernia.Arch Surg 121: 416–420

5. Haas O, Rat P, Christophe M (1990) Surgical results of intrathoracicgastric volvulus complicating hiatal hernia. Br J Surg 77: 1379–1381

6. Hill LD, Tobias JA (1968) Paraesophageal hernia. Arch Surg 96:735–744

7. Koger KE, Stone JM (1993) Laparoscopic reduction of acute gastricvolvulus. Am Surg 59: 325–328

8. Kuster GG, Gilroy S (1993) Laparoscopic technique for repair ofparaesophageal hiatal hernias. J Laparoendosc Surg 3: 331–338

9. McKernon JB, Champion JK (1995) Laparoscopic antireflux surgery.Am Surg 61: 530–536

10. Menguy R (1988) Surgical management of large paraesophageal her-nia with complete intrathoracic stomach. World J Surg 12: 415–422

11. Oddsdottir M, Franco AL, Laycock WS, Waring JP, Hunter JG (1995)Laparoscopic repair of paraesophageal hernia. Surg Endosc 9: 164–168

12. Ozdemir IA, Burke WA, Ikins PM (1973) Paraesophageal hernia. Alife-threatening disease. Ann Thorac Surg 16: 547–554

13. Pearson FG, Cooper JD, Ilves R (1983) Massive hiatal hernia withincarceration: a report of 53 cases. Ann Thorac Surg 1: 45–51

14. Pitcher DE, Curet MJ, Martin DT (1995) Successful laparoscopic re-pair of paraesophageal hernia. Arch Surg 130: 590–596

15. Streitz JM, Ellis FH (1990) Iatrogenic paraesophageal hiatus hernia.Ann Thorac Surg 50: 446–449

16. Styger S, Ackermann C, Schuppisser JP, Tondelli P (1995) Refluxdisease following gastropexy for para-esophageal hiatal hernia. Sch-weiz Med Wochenschr 125: 1213–1215

17. Treacy PJ, Jamieson GG (1987) An approach to the management ofpara-esophageal hiatus hernia. Aust N Z J Surg 57: 813–817

18. Williamson WA, Ellis FH, Streitz JM, Shahian DS (1993) Paraesopha-geal hiatal hernia: is an antireflux procedure necessary? Ann ThoracSurg 56: 447–521

Table 3.Patients with gastroesophageal reflux (GER) after paraesophagealhernia repair without antireflux procedures

Totalhiatalhernias

Paraesoph-ageal

Anatomic repairwithout anti-refluxprocedure

Post-operativeGERb

Ellis [4] 222 51 49 4McKernon [9] 283 16 6 1Menguy [10] —a 30 21 1Streitz [15] —a 13 3 1Styger [16] —a 28 19 6Williamson [18] —a 119 100 6Total 257 198 19 (10%)This study 218 58 2 0

a —Not statedb Of those not having had an antireflux procedure who subsequently re-quired therapy for GER

898

Extracorporeal knotting simplified with a new instrument

E. Croce, S. Olmi

1st Department of General and Thoracic Surgery, Center for Minimally Invasive Surgery, Fatebenefratelli and Oftalmico Hospital, C.so Porta Nuova23, Milan, Italy

Received: 11 June 1995/Accepted: 26 January 1996

Abstract. We report a quick, reliable, inexpensive methodthat uses a new, reusable instrument which can be used inlaparoscopic and thoracoscopic surgery to execute any kindof extracorporeal suture.

Key words: Extracorporeal laparoscopic suturing —Roeder slip-knot — Surgical knot

In minimally invasive surgery the most common techniqueused for suturing (for example, in Nissen Rossetti fundo-plication, of a blood vessel, a cystic duct, or the base of theappendix) is the standard extracorporeal knot technique.

An instrument created by us allows speed and ease ofexecution and the possibility of doing different types ofknots with any type of suture material that is at least 90 cmlong.

Is it possible not only to do some autostopping slip-knots in an easier way, equivalent to a Roeder knot,but,more importantly, all the simple and double surgical knotswhich one traditionally used can be executed with ease?

After having made the knot, the instrument works as aknot pusher.

Technique

The instrument for tying laparoscopic knots (produced by Karl StorzGmbh) is made of steel; it is about 55 cm long, with a proximal part 3 mmin diameter and 6 cm long that ends in a flattened portion which containsan aperture. The distal part, 5 mm in diameter and 49 cm long, has arounded end that allows its use as a small probe for palpation (Fig. 1).

Let us now examine the technique of simplified external knotting.Having introduced the needle into the cavity, while the other end of thesuture stays external to the trocar, it is passed through the two edges inorder to stitch and is then extracted via the same trocar in which it had beenintroduced. The surgical knot or slip knot is made quickly and autono-mously. There is no need for an assistant’s finger to remain on the trocarwhere the two ends come out, to keep them separated and to prevent the

escape of CO2—such as occurs with the Roeder slip knot, which engagesboth the surgeon and an assistant.

Thus the needle is passed through the hole and, while the right hand isholding both the instrument and the part of the thread ending with theneedle (Fig. 2), the left hand is used to perform the simple or doublesurgical knot.

Having tied the knot, it is sufficient to advance the instrument into theabdominal cavity so that the knot can be tied. In this manner, the instru-ment behaves very much as an extension of the surgeon’s finger (Figs. 3and 4).

Once the first knot has been tightened, the instrument is pulled out andthe procedure is repeated until the desired number of knots have beencreated.

As mentioned earlier, it is also possible to create a slip knot similar toa Roeder slip knot.

After tying a surgical knot, the thinner part of the instrument is put justunder the knot, holding the two ends of the suture material close to it.

The shorter end of the instrument is circled around the end with theneedle four times so that it is pegged together to the last coil and thenpassed in the hold (Fig. 5).

The instrument is then pulled out, which allows preparation of knotssimilar to a Roeder knot (Fig. 6).

Successively introducing the needle in the cavity, the knot can bepushed into the abdominal cavity and will position itself around the desiredstructure (Fig. 7).

Another similar slip knot is obtained by locating both ends of the threadparallel to the instrument and winding the shorter end round the end withthe needle five or six times with (Fig. 8) or without (Fig. 9) peggingtogether the third coil.

Passing the short end in the hole, the instrument is pulled out, thus tyingthe knot being made, which can be pushed into the cavity (Figs. 6and 7).

It is possible, in order to obtain extra hold, to execute a surgical knotin front of the slip-knot.

Moreover, by executing a slip knot it is possible to employ the newinstrument in the usual fashion as an endoloop, using the preferred threadand knotting and tightening the knot everywhere due to the length of theinstrument.

Results

We successfully use this instrument and these suturing tech-niques in any situation in which execution of an extracor-poreal knot is required.

In about 30 s, especially when employing the usual sur-gical knot, even an inexpert laparoscopic surgeon is able toCorrespondence to:S. Olmi, Via Solferino 42, 20121 Milano, Italy

Surg Endosc (1997) 11: 963–964

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

execute extracorporeal suturing extremely safely and muchmore simply, quickly, and cheaply than with any othermethod.

Discussion

One of the most problematic areas in the field of laparo-scopic surgery is knot tying. In certain instances, this hurdleis sufficiently great that laparoscopic procedures have to beabandoned. This is particularly true where organs such asthe proximal stomach require suturing in the proceduressuch as Nissen fundoplication; the method described heremakes it possible to execute blood vessel bindings, a cho-lecystectomy, a Heller-Dor, a gastric perforation closure, areinforcement of stapled-bowel anastomosis, and intestinalanastomosis. In addition, the length of the instrument allowsone to tighten a knot anywhere either in pelvic rectness or atesophageal hiatus level, even in obese patients. This paperdescribes an instrument which greatly facilitates the execu-tion of knot tying during laparoscopic interventions. Thisinstrument acts in essence as an extension of the surgeon’shand and, given its rigid structure, allows the surgeon fullcontrol of the process of knot tying. As a result, this tech-nical device simplifies knot tying and may help to reducethe frustration often associated with intracorporeal suturingduring laparoscopic surgery.

References

1. Buess G, Cuschieri A, Manuckle K (1993) Technique and preliminaryresults of laparoscopic cardiomyotomy. End Surg 1: 76–81

2. Cadiere G, Houben J, Bruyns J, Himpens J, Panzer J, Gelin M (1994)Laparoscopic Nissen fundoplication; technique and preliminary re-sults. Br J Surg 81: 400–403

3. Cuschieri A, Shimi S, Nathanson L (1992) Laparoscopic reduction,crural repair, and fundoplication of large hiatal hernia. Am J Surg 163:425–430

4. Meilahn JE (1992) The need for improving laparoscopic suturing andknot tying. J Laparoendosc Surg 2: 267

5. Nathanson LK, Nathanson PDK, Cuschieri A (1991) Safety of vesselligation in laparoscopic surgery. Endoscopy 23: 206–210

6. Nathanson LK, Easter DW, Cuschieri A (1991) Ligation of the struc-tures of the cystic pedicle during laparoscopic cholecystectomy. Am JSurg 161: 350–354

7. Roder H (1918) Die tecknik der mandelgesundungbestrebungen.Aertzl Rundshau 57: 169–171

8. Semm K (1978) Tissue puncher and loop ligation. New aids for sur-gical-therapeutic pelviscopy endoscopy intraabdominal surgery. En-doscopy 10: 119–124

9. Shimi S, Cuschieri A, Vander Velpen G (1994) Comparative study ofthe holding strength of slipknots using absorbable and nonabsorbableligature materials. Surg Endosc 8: 1285–1291

10. Soper NJ, Hunter JG (1992) Suturing and knot tying in laparoscopy.Surg Clin North Am 72: 1139–1152

11. Sung-Tao K, Airan MC (1994) Therapeutic laparoscopic suturingtechniques. Surg Endosc 6: 41–46

12. Szabo Z (1993) Laparoscopic suturing and tissue approximation. In:Minimally invasive surgery. McGraw Hill, New York, pp 141–155

Fig. 1. Laparoscopic knot-tying instruments.

Fig. 2. Needle is passed through the hole of the instrument to tie knot.

Fig. 3. Once tied, the instrument acts as an extension of the surgeon’sfinger.

Fig. 4. The instrument may now be advanced into the abdominal cavity.

Figs. 5–7. Technique for preparation of knots similar to a Roeder knot.

Figs. 8–9. Technique for preparation of slip knots with (Fig. 8) orwithout (Fig. 9) pegging together the third coil.

964