Fax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com

Technical Note

Pancreatology 2011;11:233–239 DOI: 10.1159/000324273

Use of the Recanalised Umbilical Vein for Islet Autotransplantation following Total Pancreatectomy

Cristina Pollard   a Gianpiero Gravante   a M’Balu Webb   a Wen Yuan Chung   a Severine Illouz   a Seok Ling Ong   a Patrick Musto   b Ashley R. Dennison   a

Departments of a   Hepatobiliary and Pancreatic Surgery, and b   Anesthesia and Intensive Care Unit, LeicesterGeneral Hospital, Leicester , UK

safe and allows for venous sampling and postoperative mea-surements of the portal pressure. Under local anaesthetic, the umbilical vein can be approached above the umbilicus and exteriorised if repeated transplants are required for al-lograft patients.

Copyright © 2011 S. Karger AG, Basel and IAP

Introduction

Intractable pain is the most disabling symptom of chronic pancreatitis. Although the majority of patients can be controlled with analgesics and regional pain blocks, a number of them will require additional mea-sures such as endoscopic stenting, drainage procedures or even resectional surgery. Total pancreatectomy is the most effective treatment but inevitably results in diabetes due to the complete removal of all functioning endocrine tissue [1] . In an attempt to prevent or abrogate the endo-

Key Words

Islet autotransplantation � Umbilical vein � Pancreatectomy  � Chronic pancreatitis

Abstract

Introduction: Islet autotransplantation requires access to the portal vein or tributaries. We originally infused islets into the liver via the middle or right colic veins, but since 2005we have used the recanalised umbilical vein. Here, we de-scribe the technique, the advantages and the early results achieved. Materials and Methods: After removal of the pan-creas and restoration of the biliary and enteric continuity, the ligamentum teres is transected. The obliterated umbilical vein is identified and recanalised with Bakes dilators giving access to the left portal vein. A Vygon � Nutricath ‘S’ 11-Fr catheter is inserted and used for the islet infusion. If the liga-mentum teres is to be exteriorised for postoperative mea-surements or subsequent access, it is pulled through a 10-mm laparoscopic port in the epigastrium, sutured to the skin and covered with a dressing. Results: We have used this ap-proach in 17 patients and exteriorised the falciform ligament in 4. There have been no intra- or postoperative complica-tions. Conclusions: The recanalised umbilical approach is

Received: October 14, 2010 Accepted after revision: January 12, 2011 Published online: May 17, 2011

Mr. Ashley Dennison Department of HPB Surgery, Leicester General Hospital Gwendolen Road, Leicester LE5 4PW (UK) E-Mail ashley.dennison   @   uhl-tr.nhs.uk

© 2011 S. Karger AG, Basel and IAP1424–3903/11/0112–0233$38.00/0

Accessible online at:www.karger.com/pan

This article was originally presented at the 33rd Annual Meeting of the Pancreatic Society of Great Britain and Ireland, October 30–31, 2008, Derby, UK.

Pollard   /Gravante   /Webb   /Chung   /Illouz   /Ong   /Musto   /Dennison  

Pancreatology 2011;11:233–239234

crine effects of total pancreatectomy, a number of units employ islet autotransplantation immediately following surgery. This consists in the reinfusion of the patient’s own islets following the pancreatic resection in patients who are not diabetic prior to the surgery or those with C-peptide-positive diabetes. In this second group, the rein-fusion of islets still reduces the requirements for exoge-nous insulin and helps achieve a better metabolic control. The ability of an islet autotransplant to prevent the onset of diabetes or reduce the requirements for exogenous in-sulin is a valuable addition to the resectional surgery and can dramatically improve the quality of life in those pa-tients affected by severe chronic pancreatitis.

Since the introduction of the Edmonton Protocol, the number of islet allografts being performed has dramati-cally increased. The liver is the main hosting site although a number of sites have been alternatively used, including the spleen, intraperitoneally, wrapped in omentum, un-der the renal capsule and in the brachioradialis muscle [2] . Hepatic islet infusions require percutaneous and fre-quently transhepatic access to the portal vein under local anaesthesia, procedures occasionally associated with ma-jor complications [3] , or infusion through the portal vein via its tributaries. When using the middle colic, the right hepatic or the mesenteric vein have to be accessed intra-operatively because they cannot be exteriorised. This ap-proach prevents any postoperative repeated infusions of pancreatic digest [4] unless a second or a third surgical procedure is conducted under general anaesthesia [5] . A different method involves the use of a large omental vein which is then exteriorised through the abdominal wall by creating an omental tongue fashioned as a pedicle around the vein which is used. This allows access to the portal vein in the postoperative period and is useful if further infusions are required [6] .

We originally used the middle or right colic vein for islets infusions and observed an immediate significant rise of the portal pressure following the infusion. Al-though no histological data are available, it is legitimate to speculate that the rise derives from the islets embolisa-tion and occlusion of the hepatic sinusoids that increases the organs vascular resistances. Such changes are even greater following infusions of non-purified digest possi-bly due to the larger diameter of the infused islets com-pared to purified digest. Still, very little data exists about the postoperative duration of these changes and whether these pressures eventually return to preinfusion levels af-ter surgery. In fact, echogenic nodules are present in 25% of the livers receiving autotransplanted islets after 6– 12 months following surgery, and these changes may in

some cases result in persistent elevations of portal pres-sure [7] . As postoperative portal pressure measurements are not possible when the islets have been infused via a colic vein, we decided to employ the umbilical vein to ac-cess the portal venous system. This technique is not new and was used as early as 1991 to infuse islets into the liv-er [7–13] , but no specific paper has been devoted to the technique itself in islet transplantation or described the exteriorisation of the umbilical vein for the measurement of postoperative portal pressures. For both these reasons, in the present article, we describe the technique adopted and present the results achieved in a group of 17 patients.

Materials and Methods

Preoperative Patient Selection Total pancreatectomy for the unremitting pain from chronic

pancreatitis is a major undertaking only considered when patients fail to respond to maximal medical management over a very pro-longed period. Maximal management generally includes regional nerve blocks and not infrequently endoscopic approaches, surgi-cal drainage or partial resections. Our approach to these patients consists of an extensive workup which includes basic laboratory tests of exocrine and endocrine function (glucose tolerance test, serum HbA 1c level and oral butterfat test), a review or repeat of their cross-sectional imaging and magnetic resonance cholangio-pancreatography and then referral to the pancreatic multidisci-plinary team for an independent assessment. The team consists of a pancreatic surgeon, a diabetologist, a gastroenterologist, a pain specialist, an anaesthetist and a medical psychologist. During the assessments, the patient is followed by a clinical nurse specialist who organises the referrals and appointments, and provides ad-equate counselling regarding the procedure and the postoperative management and follow-up. Once the multidisciplinary team members approve surgery and islet transplantation as an appro-priate treatment, the patient is then formally consented.

Surgical Technique Criteria for patient selection and the surgical technique of total

pancreatectomy with partial duodenectomy have been previously described [14, 15] . We now use the umbilical vein to infuse islets in all patients undergoing total pancreatectomy unless any previ-ous surgery near the falciform ligament (and involving the um-bilical vein) prevent its use. The blood supply of the pancreas (splenic artery and vein and the gastroduodenal artery) is pre-served for as long as possible to minimise the warm ischemic time. Meticulous dissection and attention to haemostasis is important to avoid any intraoperative bleeding that can be exacerbated by the heparinisation prior to islet cell infusion (patients received 5,000 IU of heparin immediately preceding the infusion of pan-creatic islet cells) and the temporary portal hypertension that fol-lows the infusion. Preservation of the duodenum was performed in the first 6 cases in our series, but was abandoned due to duode-nal complications which occurred following interruption of the pancreaticoduodenal arcade in 2 patients [16, 17] . It is theoreti-cally also important to leave the pancreatic capsule intact and

Islet Autotransplantation Infusion Technique

Pancreatology 2011;11:233–239 235

avoid transection of the pancreas intraoperatively to minimise the period of warm ischemia and facilitate the digestion process. Splenic conservation is possible in the majority of patients ( 1 85%) by preservation of the splenic artery and vein and only rarely is it necessary to rely on the short gastric vessels.

After removal of the pancreas and restoration of biliary and gastrointestinal continuity, the ligamentum teres is transected and the obliterated umbilical vein identified ( fig. 1 ). This is reca-nalised with Bakes dilators which gives excellent access to the left portal vein ( fig. 1 ). A Vygon � Nutricath ‘S’ 11-Fr catheter is then inserted into the umbilical vein and used for the islet infusion ( fig. 1 ). The ligamentum teres is subsequently exteriorised in the epigastrium (using a 10-mm laparoscopic port), sutured to the skin and covered with a sterile dressing.

Statistical Analysis All data were inserted into an Excel database (Microsoft, Red-

mond, Wash., USA) and analyzed with the Statistical Package for the Social Sciences Windows version 13.0 (SPSS, Chicago, Ill., USA). Descriptive statistics used for continuous variables were mean and standard deviation when parametric, median and range (min–max) for non-parametric and frequencies for cate-gorical variables. Normality assumptions were demonstrated with histograms and the Kolmogorov/Smirnov test. The Student’s t test was used to compare the means between the groups for age, weight, height and BMI. The ANOVA test for repeated measures was performed on parametric data to compare metabolic results over time and between groups (umbilical vs. previous non-umbil-ical vein patients – data not reported), the Friedman test on non-parametric data. A p value of ! 0.05 was considered statistically significant.

Results

To date, we have performed 57 islet autotransplants and the recanalised umbilical vein has been used in the last 17 patients ( table 1 ), 4 of whom had also their falci-form ligament exteriorised for portal pressure measure-ments. Since the introduction of this new method, we have only had to revert to using the middle colic vein in 1 patient because he had previously undergone surgery for a perforated duodenal ulcer that involved the division of the falciform ligament and of the umbilical vein. The maximal amount of digest that has been infused in our series with the umbilical vein approach is 32 ml and the volumes infused are not significantly different compared to the previous technique.

Pressure results are presented in figure 2 . During the islet cell infusion, the portal pressures have been continu-ously monitored to ensure that the levels reached did not exceed 20 cm water. In most cases, the portal vein pres-sures at the end of the infusion were below 20 cm al-though a significant rise was present compared to base-line levels. Portal pressures then progressively decreased until the fourth postoperative day (n = 4; fig. 2 ). Central venous pressure was always lower than portal pressure and had a delayed rise only on the second and third post-operative day, while on the fourth they decreased to base-line levels (n = 4; fig. 2 ).

a b

c d

Fig. 1. Surgical technique. Falciform liga-ment identified ( a ), umbilical vein isolated and opened ( b ), umbilical vein recanalised with Bakes dilators ( c ) and insertion of Nutricath into umbilical/portal vein ( d ).

Colo

r ver

sion

ava

ilabl

e on

line

Pollard   /Gravante   /Webb   /Chung   /Illouz   /Ong   /Musto   /Dennison  

Pancreatology 2011;11:233–239236

Comparison of metabolic results among umbilical versus non-umbilical islets infusion revealed no signifi-cant differences between techniques for the fasting glu-cose tolerance test, glycosylated haemoglobin, fasting and after 2 h from stimulation C-peptide at 36 months of follow-up ( fig.  3 ). Following the transplant, a transient rise in the alkaline phosphatase values was observed ( fig. 4 ) corresponding to a grade 1 toxicity according to the Common Terminology Criteria for Adverse Events version 3.0 (National Cancer Institute). However, again, this was not any different from the results recorded with the non-umbilical approach. With our previous tech-nique of islets infusion, 1 patient experienced a postop-erative splenic infarct (during the period when we in-fused some of the digest into the spleen via the short gas-tric veins) and a second patient a portal vein thrombosis. This second patient had no clinical or laboratory signs to

suggest the portal vein thrombosis that was detected dur-ing an ultrasound scan conducted as part of a postopera-tive clinical trial investigating the portal venous velocities in the right and left portal veins. Following heparinisa-tion, the thrombus resolved completely with no long-term sequelae. However, familiarity with the umbilical vein technique has demonstrated its safety and reliability as no intra- or postoperative complications were recorded with this approach.

Discussion

Prior to the umbilical vein, we used the middle or right colic vein (identified and cannulated within the transverse mesocolon) to infuse the islet cell digest [4] . Islets can also be infused into the portal vein following access to the mesenteric vein using a minilaparotomy with a Rocky-Davis incision. All these approaches re-quired the patient to remain anaesthetised during the islet delivery and repeated infusions were not feasible as the veins could not be safely exteriorised for prolonged periods even using an omental tube. The umbilical vein as a method of portal venous access has been long used

Table 1. Clinical characteristics of patients examined (n = 17)

Age, years 39810Height, cm 16883Weight, kg 6783BMI 20.080.9Preoperative GTT, mmol/l

0 min30 min

120 min

580.67.682.37.382.6

Preoperative HbA1c 5.880.8Preoperative C-peptide, ng/ml

0 min30 min

120 min

1.4 (0–5.9)3.7 (0–12.5)6.4 (0–11.9)

Preoperative BTT difference 113 (0–600)Duration of surgery, min 463862Blood loss, ml 700 (0–1,500)Islet transplanted

Total volume infused, mlTotal islets countTotal IEQIEQ/kgIEQ g pancreas

19.987.7237,500 (37,800–461,461)195,660 (42,832–491,260)

3,031 (607–7,038)3,038 (769–8,932)

ITU stay, days 4.5 (0–7)Hospital stay, days 16 (0–45)

G TT = Glucose tolerance test; HbA1c = glycosylated haemo-globin; BTT = butterfat test: following a fasting blood sample, patients are given 1 g/kg b.w. butter. A second sample is then tak-en at 2 h. The change in the light scattering intensity of the serum measured with a nephelometer is then calculated by simple sub-traction. A difference of <20 is abnormal, between 20 and 60 is borderline, and >60 is normal. ITU = Intensive therapy unit; IEQ = islet equivalents.

0

5

10

15

20

25

30 Central venous pressure

Pres

sure

val

ues

(mm

Hg)

Beforeinfusion

End ofinfusion

1 2Timepoint (days)

Portal venous pressure

3 4

Colo

r ver

sion

ava

ilabl

e on

line

Fig. 2. Results of the postoperative measurements for the central and portal venous pressures of the last patient (transplanted in October 2010).

Islet Autotransplantation Infusion Technique

Pancreatology 2011;11:233–239 237

for total parenteral nutrition in neonates (where the vein is still patent) or occasionally in adults with difficult vas-cular access problems [18] . The use of the recanalised umbilical vein overcame the obstacles associated with the middle or right colic vein. In addition, the umbilical vein can be approached both during open surgery for to-

tal pancreatectomy and islet autotransplant and also un-der local anaesthetic above the umbilicus for allograft patients that require repeated transplants [7] . This meth-od would extend the indications for this methodology of infusion in situations where pancreatectomy is not re-quired including xenogenic islets [19] or stem cells trans-

0

5

10

15

20

Fast

ing

gluc

ose

tole

ranc

e te

st (m

mol

/l)

0 6 12 24 36Timepoint (months)a

NoUmbilical vein

Yes

0

6

3

9

12

15

HbA

1c (%

)

0 6 12 24 36Timepoint (months)b

0

2

1

3

4

5

6

7

Fast

ing

C-pe

ptid

e (n

g/m

l)

0 6 12 24 36Timepoint (months)c

NoUmbilical vein

Yes

0

2

4

6

10

8

12St

imul

ated

C-p

eptid

e at

120

min

(ng/

ml)

0 6 12 24 36Timepoint (months)d

Colo

r ver

sion

ava

ilabl

e on

line

Fig. 3. Comparison of metabolic results among umbilical versus non-umbilical islets infusion over time. Fast-ing glucose tolerance test ( a ), glycosylated haemoglobin ( b ), fasting ( c ) and after 2 h from stimulation ( d ) C-peptide.

Pollard   /Gravante   /Webb   /Chung   /Illouz   /Ong   /Musto   /Dennison  

Pancreatology 2011;11:233–239238

plantations [20, 21] for the treatment of type 1 or type 2 diabetes. Furthermore, the recanalised umbilical vein provides excellent access for postoperative venous sam-pling, pharmacological manipulations and portal pres-sure measurements as it can be safely exteriorised for at least 5 days following surgery. Although these early en-couraging findings need to be confirmed in a larger se-ries, the approach to the portal vein via the umbilical vein appears to be very safe and reliable. For this reason, the umbilical vein approach is now the preferred method of hepatic islet infusion in our unit unless previous sur-gery has involved the falciform ligament (or the vein it-self) and it cannot be accessed or recanalised.

Conclusions

The umbilical vein approach is associated with no dif-ferences in terms of postoperative metabolic function compared to our previous method utilising the middle or right colic vein. This is a safe technique with a number of specific advantages, particulary the ability to monitor postoperative pressures serially for up to 5 days and ac-cess the portal system for sampling or pharmacological manipulation. In patients receiving islet allografts, the procedure can be conducted under local anaesthesia al-lowing repeated infusions when necessary.

0

50

100

150

200

250Alb

Pre-op.

Post-op. (0)

1Days

ALPALTBil

102 3 4 5 6 8

Colo

r ver

sion

ava

ilabl

e on

line

References

1 Argo JL, Contreras JL, Wesley MM, Chris-tein JD: Pancreatic resection with islet cell autotransplant for the treatment of severe chronic pancreatitis. Am Surg 2008; 74: 530–537.

2 Pattou F, Kerr-Conte J, Wild D: GLP-1-re-ceptor scanning for imaging of human � cells transplanted in muscle. N Engl J Med 2010; 363: 1289–1290.

3 Ryan EA, Paty BW, Senior PA, Shapiro AM: Risks and side effects of islet transplantation. Curr Diab Rep 2004; 4: 304–309.

4 Osama Gaber A, Chamsuddin A, FragaD, Fisher J, Lo A: Insulin independence achieved using the transmesenteric ap-proach to the portal vein for islet transplan-tation. Transplantation 2004; 77: 309–311.

5 Vantyghem MC, Kerr-Conte J, Arnalsteen L, Sergent G, Defrance F, Gmyr V, Declerck N, Raverdy V, Vandewalle B, Pigny P, Noel C, Pattou F: Primary graft function, metabolic control, and graft survival after islet trans-plantation. Diabetes Care 2009; 32: 1473–1478.

6 Nath DS, Kellogg TA, Sutherland DE: Total pancreatectomy with intraportal auto-islet transplantation using a temporarily exterio-rised omental vein. J Am Coll Surg 2004; 199: 994–995.

7 Ong SL, Pollard C, Rees Y, Garcea G, Webb M, Illouz S, Berry DP, Dennison AR: Ultra-sound changes within the liver after total pancreatectomy and intrahepatic islet cell autotransplantation. Transplantation 2008; 85: 1773–1777.

8 Warnock GL, Kneteman NM, Ryan E, Seelis RE, Rabinovitch A, Rajotte RV: Normogly-caemia after transplantation of freshly iso-lated and cryopreserved pancreatic islets in type 1 (insulin-dependent) diabetes melli-tus. Diabetologia 1991; 34: 55–58.

9 Atkison PR, Zucker P, Hramiak I, Paul TL, Dupre J, Behme MT, Scharp DW, Lacy PE, Olack BJ, Stiller CR: Continued insulin de-pendence despite normal range insulin sen-sitivity and insulin connecting peptide levels in a kidney/islet transplant patient. Diabetes Care 1996; 19: 236–240.

10 Farkas G, Pap A: Management of diabetes in-duced by nearly total (95%) pancreatectomy with autologous transplantation of Langer-hans’ cells (in Hungarian). Orv Hetil 1997; 138: 1863–1867.

Fig. 4. Liver function tests of the last patient (transplanted in Oc-tober 2010). A rise in the values of the alkaline phosphatase (ALP) is observed following the operation, suggesting a mechanism of cholestasis. However, bilirubin (Bil) values remain within normal levels. Alb = Albumin; ALT = alanine transaminases.

Islet Autotransplantation Infusion Technique

Pancreatology 2011;11:233–239 239

11 Keymeulen B, Ling Z, Gorus FK, Delvaux G, Bouwens L, Grupping A, Hendrieckx C, Pipeleers-Marichal M, Van Schravendijk C, Salmela K, Pipeleers DG: Implantation of standardized � -cell grafts in a liver segment of IDDM patients: graft and recipients char-acteristics in two cases of insulin-indepen-dence under maintenance immunosuppres-sion for prior kidney graft. Diabetologia 1998; 41: 452–459.

12 Movahedi B, Keymeulen B, Lauwers MH, Goes E, Cools N, Delvaux G: Laparoscopic approach for human islet transplantation into a defined liver segment in type-1 dia-betic patients. Transpl Int 2003; 16: 186–190.

13 Hyon SH, Ceballos MC, Barbich M, Groppa R, Grosembacher L, Vieiro MM, Barcan L, Algranati S, Litwak L, Argibay PF: Effect of the embolization of completely unpurified islets on portal vein pressure and hepatic bio-chemistry in clinical practice. Cell Trans-plant 2004; 13: 61–65.

14 Robertson GS, Dennison AR, Johnson PR, London NJ: A review of pancreatic islet au-totransplantation. Hepatogastroenterology 1998; 45: 226–235.

15 Gruessner RW, Sutherland DE, Dunn DL, Najarian JS, Jie T, Hering BJ, Gruessner AC: Transplant options for patients undergoing total pancreatectomy for chronic pancreati-tis. J Am Coll Surg 2004; 198: 559–569.

16 Wahoff DC, Papalois BE, Najarian JS, Kend-all DM, Farney AC, Leone JP, Jessurun J, Dunn DL, Robertson RP, Sutherland DE: Autologous islet transplantation to prevent diabetes after pancreatic resection. AnnSurg 1995; 222: 562–579.

17 Farney AC, Najarian JS, Nakhleh RE, Lloveras G, Field MJ, Gores PF, Sutherland DE: Autotransplantation of dispersed pan-creatic islet tissue combined with total or near-total pancreatectomy for treatment of chronic pancreatitis. Surgery 1991; 110: 427–439.

18 Lopriore E, Verheij GH, Walther FJ: Mea-surement of the ‘shoulder-umbilical’ dis-tance for insertion of umbilical catheters in newborn babies: questionnaire study. Neo-natology 2008; 94: 35–37.

19 Prochorov AV, Tretjak SI, Goranov VA, Glinnik AA, Goltsev MV: Treatment of insu-lin-dependent diabetes mellitus with intra-vascular transplantation of pancreatic islet cells without immunosuppressive therapy. Adv Med Sci 2008; 53: 240–244.

20 Parekh VS, Joglekar MV, Hardikar AA: Dif-ferentiation of human umbilical cord blood-derived mononuclear cells to endocrine pan-creatic lineage. Differentiation 2009; 78: 232–240.

21 Dominguez-Bendala J, Inverardi L, Ricordi C: Stem cell-derived islet cells for transplan-tation. Curr Opin Organ Transplant 2010 (in press).