Abstract
An emergency autologous islet transplant after a traumatic Whipple operation and subsequent total pancreatectomy was performed for a 21-year-old patient who was wounded with multiple abdominal gunshot wounds. After Whipple pancreatectomy, the remnant pancreas (63.5 g), along with other damaged organs, was removed by the surgeons at Walter Reed Army Medical Center (WRAMC) and shipped to Diabetes Research Institute (DRI) for islet isolation. The pancreas was preserved in UW solution for 9.25 h prior to islet isolation. Upon arrival, the organ was visually inspected; the pancreatic head was missing, the rest of the pancreas was damaged and full of blood; the tail looked normal. A 16-gauge catheter was inserted into the main duct and directed towards tail of the pancreas after the dissection of main duct in the midbody of the pancreas. The pancreas was distended with collagenase solution (Roche MTF) through the catheter. During 10 min of intraductal delivery of enzyme, the gland was distended uniformly. No leakage of the solution was observed. The pancreas was transferred to a Ricordi chamber for automated mechanical and enzymatic digestion. Islets were purified using a COBE 2991 cell processor. Islet equivalents (IEQ; 221,250) of 40% purity and 90% viability were recovered during the isolation, which were shipped back to WRAMC and infused by intraportal injection into the patient. Immediate islet function was demonstrated by the rapid elevation of serum C peptide followed by insulin independence with near normal oral glucose tolerance test (OGTT) 1 and 2 months later. It is possible to restore near normal glucose tolerance with autologous islet transplantation after total pancreatectomy even with suboptimal number of islets while confirming that islets processed at a remote site are suitable for transplantation.
Introduction
Recent advantages in techniques for the isolation of human pancreatic islets of Langerhans have led to the introduction of clinical trials of allo- and autologous islet transplantation (1,20,24). In rare instances, where the pancreas is traumatized by gun shot or blunt abdominal injury, total pancreatectomy may be the only option. However, the metabolic consequences are severe; type 1 diabetes mellitus is inevitable and may result in long-term complications (19). Autologous islet transplantation is an option to prevent diabetes to reliably establish normoglycemia and achieve insulin independence (4,24,26). Another option is allograft of islets of whole pancreas; however these patients will require life-long immunosuppression and side effects of these medications are considerable (15,23).
We present a case report of autologous islet transplantation that was carried out in collaboration between Diabetes Research Institute (DRI), Miami and Walter Reed Army Medical Center (WRAMC), Washington, DC, during life saving surgery for a service man critically wounded with multiple abdominal gunshot wounds (10). This article describes the techniques used and early results in the patient who received intraportal autologous islet transplantation (18).
Materials and Methods
Islet Cell Processing
The processing began when the pancreas was removed from the shipping container and was placed into a sterile tray containing cold preservation solution. The pancreas was dissected from surrounding tissue and disinfected in decontamination solutions [betadine, fungizone/cefazolin, and Hanks balanced salt solution (HBSS)]; the organ was then cannulated and distended using a collagenase solution comprised of collagenase MTF (Roche, Indianapolis, IN) and thermolysin (Roche, Indianapolis, IN). This enzyme blend was dissolved in a HBSS and used to distend the gland and fill the isolation circuit.
The pancreatic duct was cannulated with a 16-gauge angiocatheter in the pancreatic duct in the remaining section. The dissociation solution containing the enzyme was then delivered throughout the pancreas via the catheter in the pancreatic duct. This was accomplished by perfusing the enzyme into the duct with the aid of a Biorep perfusion machine (mechanical distension) under controlled temperature (4–8°C) and pressure (80–180 mmHg) for a period of 10 min after which we proceeded to the digestion.
Pancreas Digestion Phase 1: Recirculation
The digestion of the pancreas was obtained by combining mechanical and enzymatic action. The introduction of the automated method made it possible to improve the efficacy of human islet isolation (22). The key element of this method is the digestion chamber also known as the Ricordi Chamber, which is part of a closed circuit in which the enzyme solution is circulated (22). The pancreas was placed in the lower transparent 500-ml chamber and separated by a stainless steel mesh from the upper portion. The lower cylindrical portion was loaded with the distended pancreatic tissue (7–9 large pieces) and marbles; the upper portion has an inverted funnel shape.
During the digestion phase, pancreatic fragments were freed from the pancreatic tissue and pushed through the mesh by the hydraulic pressure applied via a peristaltic pump. This allowed for the removal of fragments of defined size (limited by the mesh) to preserve their integrity. The circuit and chamber were filled up with the collagenase enzyme solution and the solution is recirculated at 37°C, as the chamber was shaken. Initially, the chamber was agitated manually, then an automated system was used that allowed for the input of defined parameters such as amplitude, stroke, and frequency of shaking as well as recording of temperature, pH, and other parameters that were of assistance in optimizing digestion conditions (6). The enzyme solution was progressively heated at a rate of 2°C/min until the target 37°C was reached, 5–8 min, by passing the solution through the stainless steel heating coil.
During the digestion procedure, samples of the digest were taken periodically to assess the progression of the dissociation and the quality (i.e., number, integrity, relation to acinar tissue) of islets. Islets were identified using diphenylthiocarbazone [dithizone (DTZ); Sigma-Aldrich] solution that confers a peculiar red color to endocrine cells by binding to the zinc of secretory granules (14), allowing to distinguish them from unstained acinar tissue under microscope. As the digestion progressed, the amount of digest in the samples increased. Recirculation was considered complete when an increase in total tissue was observed and most or all of the islets were free of the surrounding acinar tissue. The decision of stopping the digestion was made following guidelines described in Table 1.
Guidelines for Ending Digestion Phase 1
9–10 points = A: outstanding looking islets; 7–8 points = B: good looking islets; 4–6 points = C: fair looking islets; 2–3 points = D: poor looking islets; 0–1 point = F: very poor looking islets.
Pancreas Digestion Phase 2: Dilution
The digestion circuit was opened and the digest collected, beginning the dilution phase (Digestion Phase 2). The recirculation cylinder and the heating circuit were bypassed and the islet suspension was pooled into an open vessel containing chilled collection solution [RPMI containing 25% human serum albumin (HSA), 0.2 U/ml insulin, and 10 U/ml heparin] to neutralize the enzymatic activity. Additionally, the temperature of the heating coil was lowered to 30°C. Simultaneously, fresh dilution solution was continuously pumped through the collection flasks. The tissue suspension was collected, concentrated by centrifugation, and resuspended in smaller volumes of wash solution. The dilution phase lasted until islets were no longer detected in samples, over a period of 15 min. After collection and concentration of all the tissue, a sample was obtained for prepurification counts and to assess purity, percentage free islets, degree of islet fragmentation, and condition of acinar tissue.
Purification of Islets Using Density Gradients
Continuous density purification was performed by creating the decreasing density solution on top of which the pancreatic digest is loaded in cold University of Wisconsin (UW) solution (13). The continuous gradients are composed of Ficoll® 400 dissolved in amidotrizoic acid. Densities of 1.100 and 1.077 g/ml are used (Cedarlene Laboratories). The pancreatic digest (43 ml of packed pancreatic tissue) was gently resuspended in UW solution to total 100 ml and chilled for 30-min incubation to prevent the acinar tissue from becoming edematous. After collection of the resulting fractions, individual samples were taken to determine which fractions with comparable high purity (generally contained in the top fractions) can be combined. After the collection of different fractions, the tissue was washed, islets enumeration, and assessment performed before culture.
The purification step aims at enriching the final cell product with endocrine cluster while reducing the presence of nonendocrine tissue. This is meant to reduce the final volume that will be transplanted into the recipient's liver by embolization into the portal vein.
The purification procedure utilizes centrifugation of the pancreatic digest on density gradients using a semi-automated cell separation system for blood products (COBE 2991 Cell Processor; Gambro) that allows for the bulk separation of large volumes of tissue (1,6).
Islet Culture
Islet cell suspension was initially cultured at 37°C with 5% CO2, in culture media (CMRL-based medium supplemented with cytoprotective agents such as nicotinamide) with 0.5% HSA for 22 h prior to transplant at a concentration of ~20,000 islet equivalents (IEQ) per 30 ml media, in a 175-cm2 non-tissue-treated, tissue culture flask (Sarstedt). Representative aliquots of islets were removed for bacteriology assessment, viability, and endotoxin content.
Quality Control Tests and Lot Release
Diphenylthiocarbazone was utilized for discrimination of islets from acinar tissue and for islet enumeration (21). Two representative 100-μl aliquots were taken from the islet cell suspension, incubated with DTZ, and evaluated using a light microscope with an ocular micrometer to enumerate and assess islet purity. The size distribution of the islets is quantified within a range of 50 to >350 μm (Table 2). The islet volume was calculated based on the assumption that islets are spherical, and the number of islets is expressed in terms of IEQ using a conversion factor multiplied by the number of islets in the specific size range to convert them to the “ideal” one IEQ equal to a 150-μm islet (21) (Table 2). Additionally, the DTZ staining provides a rapid means for discrimination of islet from acinar tissue, and therefore, determining islet integrity and purity of the sample (5).
Equations to Calculate Total Islet Equivalent (IEQ), Islet Particulate Number (IPN) and Islet Cell Purity
DF, dilution factor; L1, layer 1; L2, layer 2; Vol, volume.
Product release assays for islet products for transplantation include cell identity, product yield, viability, gram stain, and endotoxin (16). Postrelease assays included sterility and potency tests. Cell identity was performed by assessing purity. Purity is the percentage of islet to nonislet tissue present in a preparation (Table 2). The purity of the islets was 40%. Viability was performed as an indirect measure of membrane integrity. A representative aliquot of the preparation was mixed with fluorometric inclusion and exclusion dyes. The estimated viability of the islets was 90%.
A quantitative test for gram-negative bacteria endotoxin was performed. The assay is known as the limulus amebocyte lysate test. A 1-ml sample of supernatant from the final product for transplant was tested. The endotoxin level present in final product was 0.28 EU/kg of recipient.
Pancreatic Islet Shipment to Distant Transplant Center
Prior to transplantation, the islets were collected from the tissue culture flasks, placed into 250-ml conical tubes, and the suspension was settled for 3–5 min. After removing supernatant, the islets were resuspended in transplant media (plain CMRL, 2.5% human serum albumin, 25 mM HEPES) and were washed one more time to get rid of cellular debris, tissue culture media, and soluble proteolytic activity. A sample was taken from the supernatant for microbiological analysis after the islets were suspended in the final product suspension of 200 ml of transplant media supplemented with heparin (70 U/kg of recipient body weight). The final product was loaded into a labeled, sterile transfer bag and shipped to WRAMC via commercial airline (17).
Islet Transplantation
The islets (221,250 IEQ of 40% purity and 90% viability) were shipped back to WRAMC, where by laparotomy they were injected back into the patient's main portal vein so as to seed in the liver. Portal pressures remained normal throughout the infusion.
Results
The pancreas was perfused for 10 min resulting in excellent distention. The initial weight of the pancreas was 63.5 g; only 8.7 g of undigested tissue remained in the chamber after the digestion. Major part of that undigested tissue contained connective tissue; only 10% of 8.7 g was pancreatic tissue. Islet yield after the digestion was 324,933 IEQ. At the completion of isolation procedure, a total of 221,250 IEQ were recovered in 2.0 ml of pellet volume. Microscopic appearance of postpurification islets was excellent and viability of the islet tissue was estimated at 90% using the membrane integrity test [fluorescein diacetate/propidium iodide (FDA/PI)].
During islet infusion procedure, portal pressure and blood glucose remained stable. After infusion, immediate islet function was demonstrated by the rapid elevation of serum C-peptide, confirmed by independence of exogenous insulin.
Basal/stimulated C-peptide and glucose levels following oral glucose tolerance test were 0.5/3.9 ng/ml and 80/184 mg/dl, respectively. Liver enzymes peaked on day 3 [aspartate aminotransferase/alanine aminotransferase (AST/ALT) 800/900] which normalized on day 8. Patient has been completely off insulin since day 24. Initially, he required small amounts of insulin (1–2 units/h) for total parenteral nutrition (TPN) and surgical procedures to serially close his abdomen (total 11 abdominal operations). The patient is currently eating a normal diet. Table 3 shows 5-h oral glucose tolerance test (OGTT) on 12/30/2009 and on 01/29/2010. Note that in both instances there is a prolonged decay period, which is known to occur after islet transplantation.
Oral Glucose Tolerance Test (OGTT) 12/30/2009 (Top) and 01/29/2010 (Bottom)
Discussion
To our knowledge, this is the first case report of islet transplantation obtained from the remnant pancreas after Whipple pancreatectomy in a patient who become insulin independent after autologous islet transplantation with 3000 IEQ/kg body weight of islets processed in a remote center and transferred to the recipient hospital. These results support the concept that insulin independence can be achieved in patients with suboptimal number of islets in autologous situation (2).
There are several unique features of this case. Pancreatectomy was carried out in an emergency setting after major life threatening trauma to the pancreas and several other organs. Two cases were reported from Canada in which patients underwent distal pancreatectomy for trauma followed by autologous islet transplantation. The first case received 88,000; she was insulin free at 20 months follow-up. The second case received distal pancreatectomy with splenectomy with infusion of 325,000 pancreatic islets into the portal vein on postoperative day 4; a few hours later, the patient's blood sugars normalized and he was taken off insulin (7). It should be noted that total pancreatectomy was not carried out in these cases and it is not possible to ascertain the function of implanted islets (as opposed to function of the native pancreas). We are also not certain that if there was a clear indication for autologous islet transplantation as only distal pancreatectomy was performed and most patients with the head and body of the pancreas will remain insulin free. Furthermore, intraportal islet transplantation carries the risk of portal hypertension and liver necrosis. Therefore, this procedure should not be undertaken without a clear indication, to prevent brittle diabetes and secondary complications of diabetes. Nonetheless, it has now been shown by us and others that it is feasible to perform autologous islet transplantation in cases of trauma requiring excision of the pancreas.
Islet yields are critical in determining islet function after both allogeneic and autologous islet transplantation. In our case, we obtained 3000 IEQ/kg body weight of islets, which is suboptimal for insulin independence. However, we showed that the patient was insulin independent and making adequate amounts of c-peptide. Several clinical trials report on the importance of islet yields in determining the need for insulin independence. Jindal et al. (11) reported their experience with 10 consecutive cases of autologous islet transplantation after pancreatectomy for chronic pain syndrome. Two patients underwent near-total, while eight underwent total pancreatectomy. The average weight of the excised pancreas was 53 g with islet yield of 230,000 equivalent islet numbers (EIN) and purity of 75%. At mean follow-up of 8.7 months, c-peptide values ranged from 1 to 4.3 ng/ml and glycated hemoglobin levels were within normal limits. Sutton et al. (25) reported their experience of total pancreatectomy and autologous islet transplantation in patients with several genetic mutations [protease serine 1 (PRSS1), cystic fibrosis transmembrane conductance regulator (CFTR), and serine peptidase inhibitor, Kazal type 1 (SPINK1)]. Sixteen patients had genetic mutation. No patients were taking insulin before operation. After resection and autologous islet transplantation, patients were discharged from the hospital with a daily average of 22 ± 4 units of insulin with six (38%) patients requiring fewer than 15 units of insulin at the time of discharge. At a mean follow-up of 22 months, mean insulin requirements decreased to 15 U/d. When comparing the islet yield per body weight in patients who were insulin independent versus those who still required insulin, patients who were insulin independent had similar islet cell yields at the time of surgery (7,062 IEQ/kg vs. 5,205 IEQ/kg). Webb et al. (27) from Leicester, UK reported their experience of 46 patients who had total pancreatectomy with autologous islet transplant receiving a median of 2,246 IEQ/kg body weight (range, 405–20,385 IEQ/kg body weight). Twelve patients showed periods of insulin independence and five remain insulin independent. Over the 10 years of follow-up, 100% of patients tested were c-peptide positive at their most recent assessment with normal renal function, although most patients required some exogenous insulin. Bellin et al. (3) compared islet function in eight autograft and eight allograft recipients (insulin independent or requiring minimal insulin), who were matched for similar duration posttransplant (mean 2.1 ± 1.2 years). These patients were compared with 11 healthy control subjects; age, gender, body mass index, duration posttransplant, and hemoglobin A1c levels were similar between groups. Glucose tolerance was worse in transplant recipients compared with controls. All islet recipients received significantly more islet equivalents than autograft recipients (9,958 ± 6,229 IE/ kg vs. 4,589 ± 1,232 IE/kg, p = 0.03). Insulin secretion and glucose excursion were similar in allograft and autograft recipients, despite the latter group receiving less than half as many islets. In our case, the remnant pancreas weighed 63.5 g, of which we obtained 3,000 IEQ/ kg body weight of islets. This number is certainly below the threshold required for insulin independence.
Kobayashi et al. (12) investigated the relationship between histopathologic findings and islet yield and graft function. They examined pancreatic histopathology in 105 adults who underwent pancreatectomy and autologous islet transplantation; histologic degree of fibrosis, acinar atrophy, inflammation, and nesidioblastosis were scored by a surgical pathologist. Patients received a median of 2,968 IEQ/kg; fibrosis and acinar atrophy correlated negatively with islet yield (p < 0.0001, r = 0.67), as did inflammation. There was a positive correlation of islet yield and a negative correlation of fibrosis (p = 0.006, r = 0.43) and acinar atrophy (p = 0.006, r = 0.42) with islet graft function. In our case, there was no evidence of chronic pancreatitis or fibrosis, an advantage over fibrotic pancreas. The adverse elements of unplanned surgery and isolation at a remote center added to the difficult logistics of the procedure.
Recently, it has been shown that pancreas can be processed at a remote center and islets shipped for transplantation. Ichii et al. (8) evaluated the use of gas-permeable bags for human islet preparation shipment from DRI to two remote transplant centers. Thirty-five islet preparations were shipped either immediately after isolation (n = 20) or following culture (n = 15). Islet recovery rate after shipment was higher in cultured preparations, when compared to those not cultured, though the overall recovery rate based on isolation and pretransplant counts was comparable. All preparations met product release criteria for transplantation. Ikemoto et al. (9) isolated from six cadaver donors and cultured until shipment in either gas-permeable bags or in non-gas-permeable bags and shipped from Baylor Research Institute (Dallas) to Fukuoka University, Japan. The distance of our shipment was 11,148.4 km, and the mean duration of the shipments was 48.2 ± 8.2 h. Islet viability and stimulation index did not change significantly between pre- and postshipping, in either gas-permeable bags or in non-gas-permeable bags.
In conclusion, our case had unique features of suboptimal islet yield from a remnant pancreas after traumatic Whipple operation; pancreas processed in a remote center and transferred to the recipient hospital with successful outcome.
Footnotes
Acknowledgments
The Authors wish to thank the members of the teams at the Diabetes Research Institute Cell Transplant Center and at Walter Reed Army Medical Center who participated and contributed to the success of this unprecedented procedure: Ryosuke Misawa, Andrea Curry, Sara Giunti, Carlo Ricordi, and Christopher Thome. This work was supported in part by the Diabetes Research Institute Foundation (DRIF), National Institutes of Health/National Center for Research Resources (U42 RR016603, 5U01DK070460-07, M01RR16587); Juvenile Diabetes Research Foundation International (4-2000-946 and 4-2004-361); National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases (5 R01 DK55347, 5 R01 DK056953, R01 DK025802, 1RO1 DK25802-21; 1RO1 D59993-04). The views expressed herein are those of the authors and do not necessarily reflect the policy or position of the US Army, Department of Defense, or the US Government.