Current Diabetes Reports

, Volume 10, Issue 5, pp 326–331

Pediatric Islet Autotransplantation: Indication, Technique, and Outcome


    • Department of Pediatrics, Schulze Diabetes InstituteUniversity of Minnesota
  • David E. R. Sutherland
    • Department of Surgery, Schulze Diabetes InstituteUniversity of Minnesota

DOI: 10.1007/s11892-010-0140-4

Cite this article as:
Bellin, M.D. & Sutherland, D.E.R. Curr Diab Rep (2010) 10: 326. doi:10.1007/s11892-010-0140-4


Chronic pancreatitis is a rare disease in childhood. However, when severe, a total pancreatectomy may be the only option to relieve pain and restore quality of life. An islet autotransplant performed at the time of pancreatectomy can prevent or minimize the postsurgical diabetes that would otherwise result from pancreatectomy alone. In this procedure, the resected pancreas is mechanically disrupted and enzymatically digested to separate the islets from the surrounding exocrine tissue, and the isolated islets are infused into the portal vein and engraft in the liver. Because patients are receiving their own tissue, no immunosuppression is required. Islet autotransplant is successful in two thirds of children—these patients are insulin independent or require little insulin to maintain euglycemia. Factors associated with a more successful outcome include a younger age at transplant (<13 years), more islets transplanted, and lack of prior surgical procedures on the pancreas (partial pancreatectomy or surgical drainage procedures).


Islet transplantAutotransplantPancreatitisChronic pancreatitisHereditary pancreatitisAutoislet transplantPancreatectomyTotal pancreatectomyIslet


Chronic pancreatitis is often associated with incapacitating abdominal pain, frequent hospitalization, and risk of narcotic dependence. Medical therapies and endoscopic pancreatic duct drainage procedures are generally the initial approaches in attempts to relieve the pain of chronic pancreatitis [1, 2]. When these conservative therapies fail to relieve the pain, surgical intervention may be needed. One surgical option is total pancreatectomy, to remove the root cause of the pain. However, total pancreatectomy alone will always be complicated by surgical diabetes mellitus, from complete β-cell loss. Thus, one chronic illness is traded for another. The goal of islet autotransplantation is to preserve β-cell mass in patients who undergo total pancreatectomy to treat chronic pancreatitis. In many patients, islet autotransplant will prevent or minimize postsurgical diabetes mellitus [3].

Islet autotransplantation should not be confused with islet allotransplantation. In islet allotransplantation, a patient with pre-existing type 1 diabetes mellitus receives islets isolated from a deceased donor and the recipient must be maintained on immunosuppression to prevent rejection of the foreign tissue. In contrast, in islet autotransplantation, patients receives their own cells and do not require immunosuppression. In this procedure, the patient’s pancreas is removed and is mechanically disrupted and enzymatically digested to separate the islets from the surrounding exocrine tissue. The isolated islets are infused into the portal vein and lodge in the liver sinusoids, where they engraft and function, potentially long term. In approximately two thirds of islet autograft recipients, diabetes mellitus is prevented or partially mitigated by the islet autotransplant [3, 4•, 5••, 6].

Chronic Pancreatitis in Children

Islet autotransplantation is indicated in children with painful chronic pancreatitis that is refractory to relief by other measures with pain severe enough to warrant pancreatectomy. Although the exact prevalence remains unknown, chronic pancreatitis is rare in childhood. Chronic pancreatitis may manifest initially as recurrent episodes of acute pancreatitis, with subsequent evolution to pancreatic fibrosis and intractable abdominal pain. In children, disease is most often attributed to predisposing genetic mutations or unknown causes (idiopathic) [7]. Hereditary pancreatitis most often results from mutations in the cationic trypsinogen gene (PRSS1), which permits unregulated activity of the pancreatic enzyme trypsin within the pancreas, resulting in pancreatic inflammation and injury. Transmission is autosomal dominant, so a family history of pancreatitis is common. Affected patients often present with abdominal pain at a young age (≤10 years) and have a high lifetime risk of exocrine insufficiency, diabetes mellitus, and pancreatic adenocarcinoma (up to 40%) [8, 9]. Other gene mutations associated with chronic pancreatitis include those of the serine protease inhibitor, Kazal type 1 (SPINK1) gene, and the cystic fibrosis transmembrane conductance regulator (CFTR, “pancreatic sufficient” cystic fibrosis) [10].

Selecting Appropriate Candidates for Total Pancreatectomy and Islet Autotransplant

The current practice at our institution is that all pediatric patients referred for possible pancreatectomy and islet autotransplantation are discussed by a team of medical and surgical specialists, which includes gastroenterologists, endocrinologists, and transplant surgeons. Pancreatitis should be documented by at least one of the following: 1) multiple episodes of spontaneous elevations in amylase and lipase consistent with recurrent acute pancreatitis; 2) confirmed hereditary pancreatitis (by family history or genetic testing); 3) clear morphologic or histopathologic evidence of chronic pancreatitis (eg, calcifications on CT scan, histopathology from prior surgical intervention); or 4) diagnostic studies supporting the diagnosis of chronic pancreatitis (endoscopic ultrasound, magnetic resonance cholangiopancreatography with secretin, or pancreatic function tests). Patients with pain, narcotic dependence, and/or impaired quality of life are considered surgical candidates. Although daily narcotic use is common, it is not necessary to wait until the point of daily narcotic dependence to consider surgery, particularly for those children who are frequently hospitalized and not attending school secondary to pancreatic disease. In fact, chronic narcotic use can result in opioid-induced hyperalgesia, a condition in which the body’s pain sensitivity is increased, which may make postoperative narcotic weaning more difficult [11].

We assess islet function prior to surgery, with a hemoglobin A1c level and fasting and stimulated glucose and C-peptide levels. Diabetes is only rarely present in children with chronic pancreatitis (~5% of pediatric patients referred to our institution from 2006–2009). With hereditary pancreatitis, 5% of patients will develop diabetes by 10 years after symptom onset and 18% by 20 years [9]. Children with painful chronic pancreatitis who are fully diabetic are considered candidates for total pancreatectomy alone to relieve the pain, but if the diabetes is mild (C-peptide positive) and there is no evidence for autoimmune diabetes (negative autoantibodies against islet antigens) then an islet autotransplant should be done in attempt to retain C-peptide positivity and avoid conversion to labile diabetes.

Surgical Procedure and Islet Isolation Technique

Islet isolation and autotransplantation can be performed at the time of a total pancreatectomy, partial pancreatectomy, or Whipple procedure. Total pancreatectomy with islet autotransplant is the most common procedure in children at our institution, as pancreatic disease is often diffuse and partial resection is unlikely to result in significant or sustained pain relief. During pancreatectomy, care is taken to preserve the blood supply to the pancreas for as long as possible, to minimize the duration of warm ischemia. Because of the close proximity of the splenic vessels to the pancreas, the splenic vessels are ligated and splenectomy is often necessary. When the spleen is left in place, there is a risk of painful splenomegaly necessitating reoperation or variceal formation in the gastric veins draining the spleen [3]. With total pancreatectomy, a pylorus-sparing partial duodenectomy is performed with transection of the bile duct and reconstruction via duodenoenterostomies and choledochoenterostomies. Cholecystectomy and incidental appendectomy are performed to avoid unnecessary future reoperations.

Islet isolation must be performed in a facility meeting US Food and Drug Administration standards for Good Manufacturing Practice for processed tissue. Thus, islet isolations are performed at only a limited number of academic institutions in the United States that possess such a facility. The pancreas is distended with a cold collagenase enzyme solution infused through the pancreatic duct using a pressure-controlled system [12], and then the pancreas is enzymatically and mechanically digested using the semi-automated method of Ricordi [13]. During this process, the islets are freed and separated from the exocrine tissue. However, a small amount of exocrine tissue remains following collagenase digestion. Islets can be further purified by continuous iodixanol (OptiPrep; Axis-Shield, Oslo, Norway) density gradient on a COBE 2991 cell separator [14]. Although purification reduces the residual exocrine tissue, islets may be lost in the process. Thus, the decision to purify must be weighed against the potential for islet loss. At our institution, if the digest volume is large (>20 mL), then purification is typically undertaken, because large tissue volumes may precipitate a rise in portal pressure that may prevent successful intraportal infusion of all islets. For preparations in which the digest volume is small or a significant proportion of the islet cells are embedded/mantled by exocrine tissue, purification is not performed, to minimize islet loss. Islet number is counted by an aliquot stained with dithizone. Islet yield is quantified in terms of islet equivalents (IE), which is islet mass standardized to an islet size of 150 μm.

Islets are suspended in an albumin-based media (Connaught Medical Research Labratories-1066 medium; Mediatech, Manassas, VA) and transported immediately back to the operating room for transplant. The islets are infused into the portal vein. Heparin is given at the time of infusion as prophylaxis against portal vein thrombosis. Portal pressures are monitored during the infusion. If the pressure elevates to greater than 25 to 30 cm H2O, then the infusion is stopped. The remaining islets can be transplanted into alternative sites at the surgeon’s discretion. Alternative sites have included the gastric subserosa, renal capsule, and peritoneal cavity [5••, 15]. There is one case report of intramuscular islet autotransplant into the brachioradialis muscle in a 7-year-old child, with low insulin requirements and C-peptide positivity for more than 2 years [16].

Islet isolation disrupts the normal vasculature of the islet. Immediately following transplant, islets are dependent on diffusion to supply nutrients and oxygen to the islet core. Thus, islets are not fully functional initially and care must be taken to avoid undue stress on the islets until neovascularization occurs, a process that takes 2 to 4 weeks [17, 18]. In addition, animal models suggest that early hyperglycemia results in impaired revascularization, islet apoptosis, and a substantial reduction in islet mass post transplant [1922]. Therefore, all patients are placed on an insulin drip immediately after surgery, to target blood sugars in the 80- to 120-mg/dL range. Patients are subsequently transitioned to a subcutaneous insulin regimen. In the first several weeks post transplant, the goal is to maintain euglycemia as much as possible. Long term, we wean insulin to maintain the following minimum goals: fasting blood sugar less than 126 mg/dL, postprandial blood sugars less than 180 mg/dL, and hemoglobin A1c ≤ 6.5%.

Liver Function After Intraportal Islet Transplant

Although theoretically a risk of intraportal islet infusion, portal vein thrombosis is rare, and the risk is minimized by heparinization and halting islet infusion for elevated portal pressures [3]. Transient elevations in liver enzymes are commonly observed following intraportal islet infusion. However, these return to normal or near-normal by 1-month post transplant, and are not associated with any adverse long-term sequelae in liver function [23]. However, one should note that the presence of intrahepatic islets may change the “normal” appearance of the liver on routine imaging. In one series, 25% of patients had echogenic nodularity on liver ultrasound at 6 to 12 months post transplant. These were considered benign findings, not associated with any differences in liver function tests compared with a control group of total pancreatectomy alone patients [24•].

Pain Relief After Pancreatectomy

The bulk of experience with pancreatectomy and islet autotransplantation has been in adults. Overall, more than half of patients successfully wean off narcotic medications [6, 2528]. At our institution, 15% of surgeries are performed in children. Over 60% of pediatric patients discontinue and remain off narcotic medication; those who continue to use narcotic analgesics have better pain control on more mild agents or lower doses [5••]. Significant improvement in overall health-related quality of life, as assessed by the RAND short-form-36 general health survey (SF-36), are seen by 1-year postoperatively (Bellin, Unpublished data).

Insulin Use After Islet Autotransplantation

Overall, approximately one third of patients will achieve insulin independence and another one-third will require minima insulin—often just one injection per day of basal insulin [3]. At the institutions with the greatest experience with islet autotransplantation, insulin independence rates range from 26% to 41% [25, 2931•]. The longevity of islet autografts is less clearly defined, but long-term maintenance of insulin independence and graft function is frequent. In one analysis, of those patients achieving insulin independence, 74% remain so at 2 years post transplant and 46% at 5 years [4•]. Insulin independence has been observed for over 16 years in one case [32]; this patient continues to demonstrate islet graft function with minimal insulin requirements (once-daily glargine at <10 units/d) at over 23 years post transplant (Sutherland, Unpublished observations). In another publication from the University of Leicester (Leicester, United Kingdom), all patients receiving an islet autotransplant demonstrated graft function, defined as measurable C-peptide production, at their most recent assessment and up to 10 years post transplant [31].

The first pediatric islet autotransplant was performed in 1989 in a 12-year-old boy who subsequently achieved insulin independence for nearly 2 years [33]. Insulin independence rates may be slightly higher in children than in adults. Among the first 24 pediatric patients undergoing islet autotransplant at the University of Minnesota (Minneapolis, MN), 56% were insulin independent at 1 year post transplant and an additional 22% required just once-daily basal insulin. Hemoglobin A1c levels were less than 6% in insulin-independent patients and less than 6.5% in those with partial islet graft function. Children under the age of 13 years had a particularly high likelihood of insulin independence or minimal insulin requirements [5••]. Since that initial report, an additional 21 pediatric patients have been treated at our institution. Cumulative islet graft outcomes for those patients with sufficient follow-up data are reported in Fig. 1. The vast majority of pediatric cases have been performed within the past 10 years, resulting in very limited long-term follow-up data. There are several pediatric patients who remain insulin independent at ≥5 years post transplant (Bellin, Unpublished data).
Fig. 1

Islet yield and insulin requirements in 36 children (5–18 years of age) undergoing total pancreatectomy and islet autotransplant from 1989 to 2009 at the University of Minnesota, (Minneapolis, MN). Child indicates children less than 13 years of age. Adolescent indicates patients 13 to 18 years of age, based on age at the time of transplant. Dependent indicates required a basal-bolus (typical type 1 diabetic patient) insulin regimen. The circles indicate minimal or no insulin and the triangles indicate insulin dependent. IE/kg—islet equivalents per kilogram recipient body weight transplanted; min/no insulin—insulin independent or required once-daily basal insulin alone or intermittent correction scale alone to maintain euglycemia (hemoglobin A1c ≤ 6.5%)

Factors Associated with Successful or Unsuccessful Islet Autotransplant

No single factor is predictive of insulin independence. Most case series suggest that the number of IE transplanted is key to transplant success, although there is no islet yield that guarantees insulin independence [4•, 6, 25]. Among 164 patients transplanted at the University of Minnesota from 1977 to 2007, insulin independence was observed at 1 year post transplant in 63% of patients receiving greater than 5,000 IE/kg, 27% in those receiving 2,501 to 5,000 IE/kg, and 7% in those receiving ≤2,500 IE/kg [4•]. In pediatric patients transplanted between 1989 and mid-2006, islet yield greater than 2,000 IE/kg was an important predictor of insulin independence [5••]. In contrast to these findings, the group in Leicester found no relationship between islet yield and insulin independence in the 46 patients undergoing total pancreatectomy and islet autotransplant. In fact, this series included one patient who remains insulin independent 4 years after transplant with just 955 IE/kg [28, 31•].

Prior pancreatic surgery appears to be an important factor negatively impacting islet yield and insulin independence rates if subsequent total (completion) pancreatectomy and islet autotransplant is performed [29]. The pancreatic surgeries most strongly associated with decreased islet yield and less successful islet transplant outcomes are distal pancreatectomy, Puestow drainage procedure (lateral pancreaticojejunostomy), and Frey drainage procedure (partial resection of the pancreatic head combined with lateral pancreaticojejunostomy of the distal pancreas). These procedures may result in β-cell loss from direct removal of endocrine tissue or associated scarring. In addition, because the Puestow and Frey procedures involve opening up the pancreatic duct along the length of the pancreas, the islet isolation process, which involves intraductal perfusion of collagenase, is compromised. Whipple procedures (resection of the pancreatic head along with the duodenum while leaving the distal pancreas intact) do not appear to have the same untoward effect on later islet transplant outcomes [6, 29]. Pediatric patients with a prior history of distal pancreatectomy or Puestow/Frey procedures have higher rates of insulin dependence, and those with a prior pancreaticojejunostomy (Puestow/Frey) in particular have more severe fibrosis, a lower islet yield (IE/kg), and lower IE per gram of pancreatic tissue [5••, 34•]. In addition, a longer duration of disease (>7 years), moderate to severe fibrosis, and severe acinar atrophy were associated with lower islet yields in children [34•].

Unfortunately, we currently have no way to predict islet yield or subsequent insulin needs prior to pancreatectomy and islet autotransplant. Thus, patients must be willing to risk postoperative diabetes as a trade-off for pain relief. Preliminary data from 10 pediatric patients suggest that a simple fasting glucose and patient weight may be predictive of total islet yield, and that higher stimulated C-peptide levels and lower hemoglobin A1c at baseline are associated with greater islet yields [15]. However, these results need to be verified in a larger group of patients before they may be clinically valuable. In a small series of adult patients at our institution, the acute insulin response to glucose from an intravenous glucose tolerance test was strongly correlated with islet yield [35]. This is consistent with previous data that post-transplant glucose and arginine-induced insulin secretory responses correlate with the β-cell mass transplanted, and may be useful as a surrogate measure of surviving β-cell mass [36].


Total pancreatectomy may provide pain relief for pediatric patients with severe chronic pancreatitis when other more conservative therapies have failed. Because total pancreatectomy alone will always result in diabetes, an intraportal islet autotransplant should be performed whenever possible to prevent or minimize postsurgical diabetes. All patients will require insulin therapy in the immediate postoperative period, to protect islets during the period of engraftment and compromised function. Subsequently, about two thirds of patients will wean completely off insulin or maintain good glycemic control with low-dose insulin. Substantial islet graft function and insulin independence have been demonstrated for over 5 years in pediatric patients, and longer in adults. Islet yield is best when duration of disease is short or modest (<7 years). Younger patients and those without any prior pancreatic surgery appear to have the best outcomes.


No potential conflicts of interest relevant to this article were reported.

Copyright information

© Springer Science+Business Media, LLC 2010