needle-based confocal endomicroscopy for in vivo histology of intra-abdominal organs: first results...

sdtsr

Acd

DtK

C0d

R

1

ORIGINAL ARTICLE: Experimental Endoscopy

Needle-based confocal endomicroscopy for in vivo histology of intra-abdominal organs: first results in a porcine model (with videos)

Valentin Becker, MD, Michael B. Wallace, MD, MPH, Paul Fockens, MD, Stefan von Delius, MD,Timothy A. Woodward, MD, Massimo Raimondo, MD, Rogier P. Voermans, MD, Alexander Meining, MD

Munich, Germany; Jacksonville, Florida, USA; Amsterdam, The Netherlands

Background: Probe-based confocal laser scanning endomicroscopy enables in vivo real time histopathology ofthe mucosa layer. Recently, a prototype of a new confocal miniprobe has been developed that is small enoughto be introduced through a 22-gauge puncture needle.

Objective: The aim of this study was to evaluate the feasibility of such needle-based confocal laser scanningendomicroscopy (nCLE) for in vivo histology of various organs in a porcine model.

Design: Feasibility study.

Setting: Nonsurvival animal experiments with the animals under general anesthesia at three academic centers.

Interventions: Ten pigs were examined while they were under general anesthesia. Either EUS-guided organpuncture or natural-orifice transluminal endoscopic surgery (NOTES) procedure was used. The confocal mini-probe was inserted through the 22-gauge needle, and puncture of various intra-abdominal structures and organswas performed (lymph nodes, diaphragm, ovaries, liver, spleen, and pancreas) after intravenous injection offluorescein (5-10 mL 1% or 2 mL 10% solution). Real-time sequences were recorded. Biopsy specimens weretaken for standard histopathology.

Main Outcome Measurements and Results: It was technically feasible to introduce the needle-based confocalminiprobe into various organs at the time of EUS or NOTES procedures. The device enabled real-time in vivocollection of images at histologic resolutions and of acceptable image quality from several intra-abdominalorgans interrogated.

Limitations: Data were assessed in an experimental animal setting and on healthy organs only.

Conclusions: Needle-based confocal laser endomicroscopy (CLE) of intra-abdominal organs is feasible in aporcine model. This innovative, minimally invasive technique has the potential to facilitate in vivo histologyduring EUS punctures or NOTES peritoneoscopy. (Gastrointest Endosc 2010;71:1260-6.)

Confocal laser endomicroscopy provides accurate as-essment of in vivo histopathology of the mucosal layeruring ongoing endoscopy. Similarly to standard histopa-hology, cell-to-cell borders, single-cell structures, muco-al inflammation, and vessel structures can be visualized ineal time after administration of a contrast dye.1,2 Currently

bbreviations: CLE, confocal laser endomicroscopy; nCLE, needle-basedonfocal laser endomicroscopy; NOTES, natural-orifice transluminal en-oscopic surgery.

ISCLOSURE: All authors disclosed no financial relationships relevant tohis publication. (Material support for this study was provided by Maunaea Technologies, Karl Storz Company, and Olympus.)

opyright © 2010 by the American Society for Gastrointestinal Endoscopy016-5107/$36.00oi:10.1016/j.gie.2010.01.010

eceived June 2, 2009. Accepted January 1, 2010.

260 GASTROINTESTINAL ENDOSCOPY Volume 71, No. 7 : 2010

available confocal high-resolution miniprobes are pre-dominantly designed for superficial intraluminal applica-tion introduced via the instrumentation channel of stan-dard endoscopes.3 Imaging depth is probe dependent andlimited to 30 to 150 �m. Applying this originally superficialdiagnostic method for histopathologic analysis within par-

Current affiliations: II Medical Department, Klinikum Rechts der Isar,Technical University of Munich, Munich, Germany (V.B., S.D., A.M.); Divisionof Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, Florida,USA (M.B.W., T.W., M.R.); Department of Gastroenterology and Hepatology,Academic Medical Centre Amsterdam, Amsterdam, The Netherlands(R.P.V.).

Reprint requests: Prof. Dr. Med. Alexander Meining, II. Medizinische Klinik,Technische Universität München, Klinikum Rechts der Isar, Ismaningerstr.22, D-81675 Munich, Germany.

www.giejournal.org

etKcepp

acr(codcdbnos

tbiue

M

Sa

4avpmAt2ttta

E

ptsafSuE

Becker et al Needle-based confocal endomicroscopy

w

nchymatous organs or for diagnostic issue in suspiciousumors, a prototype of a new confocal miniprobe (Maunaea Technologies, Paris, France) has been developed re-ently. This prototype is of flexible structure and is smallnough to be introduced through a standard 22-gaugeuncture needle, usually applied for EUS-guidedunctures.EUS-guided FNA is widely accepted as a safe and reli-

ble diagnostic approach in infiltrative disorders or suspi-ious lesions, with a low postprocedural complicationate.4 Natural-orifice transluminal endoscopic surgeryNOTES) procedures allow diagnostic and therapeutic ac-ess to the abdominal cavity as well as to intraperitonealrgans.5 Enhancing EUS-guided FNA and NOTES proce-ures with the diagnostic value of confocal laser endomi-roscopy (CLE) promises an interesting minimally invasiveiagnostic tool. In the past, we successfully applied com-ined NOTES procedures with CLE for superficial exami-ation of the liver, spleen, and peritoneal layer.6 However,wing to the limited depth of penetration, imaging of onlyuperficial parenchymatous structures was possible.

The aim of the present study, therefore, was to evaluatehe feasibility of a newly developed miniaturized needle-ased approach (nCLE) for in vivo histology of variousntra-abdominal organs applied in a porcine model bysing EUS-guided FNA and NOTES techniques and tostablish protocols for this new imaging tool in vivo.

ATERIALS AND METHODS

tudy animals, preinterventional care, andnesthesiaTen pigs with body weights ranging between 20 and

4.5 kg were examined with the animals under generalnesthesia with endotracheal intubation and controlledentilation at 3 study sites. Preanesthesia sedation wasrovided with an intramuscular injection of azaperone (2.0g/kg), ketamine (10 mg/kg), and atropine (0.02 mg/kg).nesthesia was initiated with propofol 1% and maintained

hroughout with a continuous infusion of pentobarbitone5 � 35 mg/kg/h. After termination of the examination,he animals were euthanized using a lethal dose of pen-obarbitone. The experimental protocol was approved byhe local or regional governmental Ethics Commission fornimal protection at each site.

quipment and interventionFor NOTES peritoneoscopy, a transgastric approach,

erformed by M.W. (n � 2) and A.M. (n � 2), or aranssigmoid route, performed by A.M. (n � 3), was cho-en. The transgastric access was achieved via a PEG-likepproach.7 For the transsigmoid route, a modified deviceor transanal endoscopic microsurgery was used (Karltorz, Tuttlingen, Germany).8 Endoscopy was performedsing a Storz and an Olympus standard gastroscope. The
US-guided approach was performed on 3 animals (by
ww.giejournal.org Vo

P.F.) using an Olympus GF-UCT140 echoendoscope com-bined with an Olympus EU-M60 processor or the GFUC160p with an EUC60 processor (Olympus Medical Sys-tems Europe, Hamburg, Germany; Center Valley, Pa).

Needle-based confocal laser endomicroscopyIdentification of the organ of interest was achieved by

endoscopic maneuvres including rotation, torque, and ret-roflexion for the NOTES procedures or by echoendo-scopic localization for the EUS-guided procedures. Organpunctures were performed by using a 22-gauge needleinserted through the instrumentation channel of the gas-troscope or echoendoscope (Fig. 1).. The puncture wasdone under direct visualization or EUS guidance by ad-vancing the needle carefully into the organ (Fig. 2). Punc-ture depth ranged between 1 and 3 cm.

After organ puncture, 5 to 10 mL 1% or 2 mL/kg 10%fluorescein (AlconPharma, Freiburg, Germany; AK Fluor,Akorn Pharmaceutical, Buffalo Grove, Ill) was injected intra-venously. Fluorescein stays predominantly in the vessels andintensifies contrast enhancement. If necessary, staining wasoptimized by reinjection.9 The miniaturized confocal probe(Mauna Kea Technologies, Paris, France; diameter, 350 �m;lateral resolution, 3.5 �m; field of view, 300 � 300 �m) wasconnected to a laser scanning unit, advanced through thepuncture needle, and attached to organ structures with mod-erate compression. The miniprobe is similar to the Cholan-gioflex probe regarding rigidity/flexibility. In retroflexion ofthe endoscope, shearing force reduced the motion of theminiprobe but movements were still possible. For reliableimage acquisition, the puncture needle allowed a stable po-sition of the miniprobe relative to the tissue, reducing respi-ratory artefacts. Dynamic real-time sequences of respectivelocations were analyzed, and sequences of 60 seconds eachwere recorded and stored digitally. Thereafter, biopsy orresection specimens were obtained for standard histopathol-ogy. This procedure has been performed repeatedly in dif-ferent organs. Intra-abdominal structures and organs, includ-ing diaphragm, liver, spleen, pancreas, and ovaries, wereinspected.

Assessment of abdominal lymph nodes was performedaccording to the concept of sentinel lymph node map-ping.10 For staining of lymph nodes and vessels, patent

Take-home Message

● Puncture needle– guided confocal laser endomicroscopyof intra-abdominal organs by EUS guidance or vianatural-orifice transluminal endoscopic surgeryprocedures is feasible. This innovative technique is afurther step in enabling transmural in vivohistopathology and might improve furthercharacterization of suspicious organs or lymph nodes.

blue (3 mL) was injected submucosally in the anterior

lume 71, No. 7 : 2010 GASTROINTESTINAL ENDOSCOPY 1261

sfsidlt

R

aowtice

Fnr

Ft

Needle-based confocal endomicroscopy Becker et al

1

igmoid wall. Thereafter, NOTES peritoneoscopy was per-ormed according to the method described above toearch for blue-stained sentinel lymph nodes related to thenfiltrated sigmoid area. Virtual histology was obtained byirect puncture of the identified lymph node. Finally,ymph nodes were excised or biopsy specimens wereaken for standard histopathology.

ESULTS

Needle-based CLE was technically feasible in all studynimals without any relevant complications. Dependingn the position of the abdominal access, different organsere examined. EUS was used in 3 animals for examina-

ion of the pancreas, liver, and spleen. In each animal,nsertion of the puncture needle through the workinghannel as well as organ puncture itself could be achieved

igure 1. Image of the new miniprobe passed through the punctureeedle. The miniprobe is comparable to the Cholangioflex probe inigidity and flexibility.

igure 2. EUS image of needle probe introduced into the pancreas. Theip of the probe (arrows) is readily visible.

asily in each position of the endoscope.


LiverFluorescein is distributed throughout the whole liver

tissue and allows an overview of the architecture. Afterinjection of contrast dye, sinusoids became evident. Ow-ing to the pharmacologic properties of fluorescein, hepa-tocytes were visualized by negative staining only. In singlecases, the central vein and ultrastructure of lobules wasvisualized; however, image quality was partly compro-mised by leakage of blood or cellular debris (Fig. 3).Overstained images owing to fluorescein leakage couldeasly be reduced by increasing the contact pressureslightly. Livers were punctured 14 times overall. All at-tempts were successful. The needle was repositioned upto 3 times per animal within the same organ to change/improve image acquisition. Fifty to sixty percent of theacquired images could be evaluated.

PancreasA transsigmoid NOTES or EUS-guided access was chosen.

Histologically, the gland is covered by a thin layer of looseconnective tissue, which was penetrated by the punctureneedle. The acini are arranged around the connective tissuesepta that contain smaller vessels. Imaging of the tissue en-abled clear visualization of connective fibers, which appearas dark honeycomb structures. The acini were stained morebrightly. The single acinus cells and pancreatic duct systemwere not visible, but single vessels with erythrocytes rushingthrough them could be clearly identified (Fig. 4; Video 1,available online at www.giejournal.org). Imaging of the pan-creas was more complex than liver biopsies, because of theanatomic position. Eight of 10 attempts were successful. Ap-proximately 40% of the sequences were evaluatable, becauseof the unstable position of the probe.

SpleenThe spleen is characterized by 2 main compartments,

the white and the red pulp. It is surrounded by a capsulecomposed of dense fibrous tissue, elastic fibers, andsmooth muscle. After penetration of the capsule, confocalmicroscopy allows distinguishing the histologic micro-structures of the spleen and the red and white pulp. Inconfocal imaging, red pulp appears brighter owing to thegood blood supply with fine reticulum fibers. The whitepulp is composed of lymph follicles and periarterial lym-phatic sheaths rich in lymphocytes and was representedby a less fluorescent structure (Fig. 5; Video 2, availableonline at www.giejournal.org). Each spleen was biopsied10 times. Sufficient sequences could be acquired at firstattempt, and 50% of the images were evaluatable.

Muscle cellsMuscles are characterized by longitudinally running

myofibrils. These single muscle cells are of cylindric shapeand contain the contractile fibers and the peripherallylocated nuclei. Confocal laser endomicroscopy accents the
longitudinal structure of the single myofibrils. Bright-
www.giejournal.org

http://www.giejournal.org


slb

O

csc

Fi

e and


w

tained capillaries were apparent (Video 3, available on-ine at www.giejournal.org). Each muscle specimen wasiopsied 10 times, and 50% of the images were evaluated.

variesNeedle-based CLE images acquired from ovaries show the

ortical region containing primary follicles and ovariantroma with multiple blood vessels in between them. Oo-

igure 4. A, Needle-based confocal endomicroscopy image and B, correntraparenchymatous blood vessels can be identified.

Figure 3. A, Needle-based confocal endomicroscopy imag

ytes measure 20 to 30 �m and are surrounded by primordial


follicles. nCLE clearly shows blood vessels in the ovarianstroma. Nuclei cannot be visualized, but oocytes can beidentified as larger round structures surrounded by a darkergrey circle and a bright circle resembling the granulosa cellsand the developing thecal layers with vascularization of thislayer by small capillaries (Fig. 6; Video 4, available online atwww.giejournal.org). Each ovary was biopsied 10 times. Suf-ficient sequences could be acquired at first attempt, and 50%

ing histopathologic image of the pancreas. Acini, connective tissue, and

B, corresponding histopathologic image of a liver lobule.

spond

of the images could be evaluated.




L

mnlCinlb

Fw

Fv


1

ymph nodes—sentinel lymph node mappingSubmucosal injection of patent blue led, owing to the fast

igration of the dye, to a good visualization of lymph chan-els and a related lymph node close to the injection site. Theymph node could clearly be identified after about 5 minutes.LE resulted in visualization of lymphatic tissue correspond-

ng well to standard histopathology from the same lymphode (Fig. 7). Lymph nodes consist of multiple lymphoidobules surrounded by lymph-filled sinuses and are enclosed

igure 5. A, Needle-based confocal endomicroscopy and B, correspondihite pulp (less fluorescence) can be differentiated.

igure 6. A, Needle-based confocal endomicroscopy image and B, corrisible (arrows).

y a capsule also forming a reticular network with capillaries


expanding inside the node. The lobes consist of the folliclesand interfollicular cortex. Patent blue did not compromiseconfocal image quality. Lymph node mapping was per-formed in 2 animals only at 1 site (Munich). The probe had tobe repositioned once because of insufficient image quality.

DISCUSSION

In this study, we present in vivo images of intraabdomi-

topathologic image of the spleen. The red pulp (more fluorescence) and

nding histopathologic image from a pig’s ovary. Single oocytes become

ng his

espo

nal organs using nCLE. Visualized organs were readily

www.giejournal.org

ikc

plcdesdpecipippettcm

cwlnvitici

and


w

dentified and characterized by typical histologic aspectsnown from standard histopathology. No relevant compli-ations occurred.

In daily clinical practice, histologic examination of sus-ected pancreatic lesions, of structural alterations, eg, in

iver cirrhosis, or for staging of malignant disease is aontinuous challenge. To confirm definite diagnosis, stan-ard histopathology is still the criterion standard. How-ver, potential disadvantages of standard biopsies areampling errors due to the limited number of biopsies andelayed diagnosis because of time-consuming samplingrocessing. This often impedes treatment during the samendoscopic session. As demonstrated in earlier studies,onfocal endomicroscopy of the superficial mucosa facil-tates immediate in vivo diagnosis without prior technicalrocessing.1 Because of the easy and uncomplicated use,n vivo imaging permits a higher number of biopsies sam-led across the organ, which will more likely reveal anyathologic changes.11 Dynamic monitoring of tissue alsonables visualization of blood flow, blood cells runninghrough the vessels, as well as dynamic organ func-ions.3,12,13 Additionally, structural changes such as in-reased microvessel density as a potentially neoplasticarker or prognostic indicator can be quantified in vivo.14

A limitation of the currently available diagnostic confo-al miniprobes is the relatively short penetration depth,hich is important for visualizing suspected neoplastic

esions or the microstructure of organs. The presentedeedle-based approach overcomes this limitation by ad-ancing the confocal miniprobe within the structure ofnterest. Invasiveness is similar to standard organ punc-ure, which can cause bleeding complications, particularlyn patients with compromised coagulation, eg, in liverirrhosis. In the present study, bleeding was the most

Figure 7. A, Needle-based confocal endomicroscopy image

nterfering factor for image quality. Bleeding led to self-


limited leakage of fluorescein, resulting in overstainedimages. However, fluorescein leackage was only a minorproblem predominately in liver biopsies, and to someextent in biopsies of the spleen. By increasing the contactpressure, image quality could be restored. However,bleeding did not affect the image acquisition seriously inany experiment. Furthermore, image quality was some-what impaired with prolonged time after puncture of anorgan.

Nevertheless, we think that further differentiation anddetection of pancreatic lesions might be a promising ap-plication field of this technique. EUS allows reliable andaccurate visualization of the respective organ. Suspiciouslesions, inflammation, or neoplasia are differentiated bymorphologic alterations or according to the results of EUS-guided core needle biopsies or FNA. However, EUS-guided biopsies or aspirations fail in up to 10% of casesowing to quality of the aquired specimen15 and are ofincongruent diagnostic accuracy in detecting malignancy.16

Because of the increased number of biopsies and theimmediate evaluation of the acquired “virtual biopsies,”nCLE might result in a higher diagnostic yield when usedas a guidance for standard EUS-guided samples and mightalso provide diagnosis within the same procedure. In reg-ular pancreatic tissue, nCLE visualized the regular honey-comb structure of the connective tissue septa. In chronicpancreatitis we would expect progressive fibrosis and lossof acinar cell mass compared with indistinct hypervascu-larized tissue in pancreatic cancer. Differentiation betweenchronic pancreatitis and carcinoma might be the mostchallenging issue for this approach. Further reasonableindications might be further differentiation of pancreaticcystic tumors.

Confocal laser endomicroscopy was also applied for intra-

B, corresponding histopathologic image of a lymph node.

abdominal lymph node mapping. This technique is used to


rdsvcttmnd

pstwgsciomwacs

aacmtcso

R


1

educe the risk of micrometastasis in regional lymph nodesuring endoscopic or minimally invasive treatment of earlytage cancers. nCLE might optimize the nodal staging byisualizing metastasis in lymphatic tissue during peritoneos-opy10 or EUS.17 In patients with early tumor stages, nCLE hashe potential, therefore, to guide treatment decisions for fur-her endoscopic or transanal tumor resection without en blocesenteric resection. Furthermore, completeness of lymphode and tumor resection could be evaluated in real timeuring ongoing therapy.

The presented study has 2 main limitations. First, thesereliminary data were assessed in an experimental animaletting only. However, NOTES techniques have already beenransfered successfully to humans, and the focus of the studyas the feasibility of this new prototype device. The EUS-uided FNA approach can be already regarded as a useful,afe, and reliable technique. In the present study, addingonfocal microscopy to these procedures did not furtherncrease the risks of the puncture itself. Second, only regularrgans without any pathologic changes were examined. Oneight suppose that exact placement of the miniprobe into, asell as evaluation of, pathologic structures, such as pancre-tic carcinomas or suspect lymph nodes, might be morehallenging. These aspects have to be further evaluated intudies including patients with various diseases.

In conclusion, we demonstrated that nCLE of intra-bdominal organs via NOTES procedures or by EUS guid-nce is feasible, with acceptable image quality, in a por-ine model. This innovative minimally invasive techniqueight be the first step toward intraorgan in vivo histopa-

hology during ongoing EUS-FNA or NOTES peritoneos-opy. We speculate that nCLE might widen the diagnosticpectrum of future NOTES procedures, EUS, and the fieldf probe-based CLE.

EFERENCES

1. Meining A, Saur D, Bajbouj M, et al. In vivo histopathology for detectionof gastrointestinal neoplasia with a portable, confocal miniprobe: an
examiner blinded analysis. Clin Gastroenterol Hepatol 2007;5:1261-7.

2. Kiesslich R, Burg J, Vieth M, et al. Confocal laser endoscopy for diagnos-ing intraepithelial neoplasias and colorectal cancer in vivo. Gastroenter-ology 2004;127:706-13.

3. Becker V, Vercauteren T, von Weyhern CH, et al. High-resolutionminiprobe-based confocal microscopy in combination with video mo-saicing (with video). Gastrointest Endosc 2007;66:1001-7

4. Al-Haddad M, Wallace MB, Woodward TA, et al. The safety of fine-needleaspiration guided by endoscopic ultrasound: a prospective study. En-doscopy 2008;40:204-8.

5. Kalloo AN, Singh VK, Jagannath SB, et al. Flexible transgastric peritone-oscopy: a novel approach to diagnostic and therapeutic interventions inthe peritoneal cavity. Gastrointest Endosc 2004;60:114-7.

6. von Delius S, Feussner H, Wilhelm D, et al. Transgastric in vivo histologyin the peritoneal cavity using miniprobe-based confocal fluorescencemicroscopy in an acute porcine model. Endoscopy 2007;39:407-11.

7. Kantsevoy SV, Jagannath SB, Niiyama H, et al. A novel safe approach tothe peritoneal cavity for per-oral transgastricendoscopic procedures.Gastrointest Endosc 2007;65:497-500.

8. Wilhelm D, Meining A, von Delius S, et al. An innovative, safe and sterilesigmoid access (ISSA) for NOTES. Endoscopy 2007;39:401-6.

9. Becker V, Delius S, Bajbouj M, et al. Intravenous application of fluores-cein for confocal laser scanning microscopy: evaluation of contrast dy-namics and image quality with increasing injection-to-imaging time.Gastrointest Endosc 2008;68:319-23.

10. Bembenek A, Rosenberg R, Wagler E, et al. Sentinel lymph node biopsyin colon cancer. Ann Surg 2007;245:858-63.

11. Pohl H, Rösch T, Vieth M, et al. Miniprobe confocal laser microscopy forthe detection of invisible neoplasia in patients with Barrett’s oesopha-gus. Gut 2008;57:1648-53.

12. Goetz M, Vieth M, Kanzler S, et al. In vivo confocal laser laparoscopyallows real time subsurface microscopy in animal models of liver dis-ease. J Hepatol 2008;48:91-7.

13. Wallace MB, Fockens P. Probe-based confocal laser endomicroscopy.Gastroenterology 2009;136:1509-13.

14. Becker V, Vieth M, Bajbouj M, et al. Confocal laser scanning fluorescencemicroscopy for in-vivo determination of micro-vessel density in Bar-rett’s esophagus. Endoscopy 2008;40:888-91.

15. Binmoeller KF, Brand B, Thul R, et al. EUS-guided, fineneedle aspirationbiopsy using a new mechanical scanning puncture echoendoscope.Gastrointest Endosc 1998;47:335-40.

16. Voss M, Hammel P, Molas G, et al. Value of endoscopic ultrasoundguided fine needle aspiration biopsy in the diagnosis of solid pancreaticmasses. Gut 2000;46:244-9.

17. Fockens P. EUS and lymph nodes: look, needle, or trust your feelings?
Gastrointest Endosc 2009;69:904-5.
www.giejournal.org

needle-based confocal endomicroscopy for in vivo histology of intra-abdominal organs: first results...

Documents