Transplantation of human islets of Langerhans is becoming an established procedure for treatment of type 1 diabetes mellitus (1). Islet transplantation can provide physiologic insulin control with euglycemia and correction of HbA1c.
With recent improvements to protocols and techniques in all aspects of the islet isolation and transplantation processes, teams are now able to isolate transplantable islets from donors that were previously considered of marginal quality. Although transplant statistics have increased exponentially on a worldwide basis since the release of the Edmonton Protocol (2), transplant efficiency, which is the percentage of islet preparations transplanted, remains close to 50%. At the Clinical Islet Isolation Lab (CIIL) 40.2% (131/326) of all pancreases processed fulfill the release criteria for transplantation. Processing a donor pancreas is costly often in excess of $25,000 CAD and transportation can be in excess of $30,000 for a sum total between $25,000–$50,000 CAD per isolation. Therefore donor selection is a critical component to ensure and sustain the facility’s success (3,4).
Donor variables have a significant impact on the outcome of the isolation. Variables such as age, cold ischemia time (CIT), body mass index (BMI), and others have all been shown to have diverse effects on isolation outcome (3–12). By careful selection of pancreases for processing, higher transplant efficiencies can be obtained.
The aim of this study was to design and test an objective screening process to determine the potential of given pancreatic donors for islet isolation. After analysis of the donor information and pancreas condition, we assigned a composite numerical score to each organ, rated from 0–100 donor points (DP).
MATERIALS AND METHODS
Human pancreases were obtained with consent from relatives of 326 multiorgan brain-dead donors from January 1999 to March 2004. The Human Ethics Committee of the University of Alberta Hospital approves all tissue procurement and isolation protocols used by the CIIL and distant organ procurement organizations and staff. A total of 18 pancreases were excluded from the analysis because of incomplete data. Pancreases that were initially accepted but upon receiving the pancreas not processed were also excluded (n=31). The reason for not processing varied from positive serology after acceptance, high HbA1c, acute or chronic pancreatitis, pancreas consistency, or substantial damage sustained to the pancreas during procurement or preservation.
Pancreases were transported to the laboratory using either University of Wisconsin (UW) solution or the two-layer method (TLM) (13–15) preservation technique. Islet isolations were completed according to an established protocol (2) based on that first described by Ricordi et al. (16). For an islet transplant to proceed, the islet preparation had to meet certain guidelines set forth by the CIIL and Health Canada. These included a minimum islet mass of 4,000 islet equivalents (I.E.) per kilogram of recipient weight, a viability of 70% or higher, an endotoxin level of less than 5 endotoxin units per kilogram of recipient weight, and a tissue volume of less than 5 cc.
The donor charts were reviewed retrospectively to assess donor score. Variables included in the screening process were age, CIT, BMI, cause of death, hospital stay, serum amylase or lipase levels, blood glucose, organ procurement team, vasopressor levels, and social and medical information that might influence the outcome of the isolation such as hypertension, alcohol abuse, and warm ischemia time. Other included medical and social history information was assessed not in terms of influence on the isolation but rather in accordance of regulatory standards for the acceptance of pancreases for clinical transplantation.
All variables were weighted depending on the variable’s perceived effect on islet isolation success (Fig. 1A) (3–10). It was thus determined that age was the most influential, accounting for 20 DP, followed by cause of death (18 DP), CIT (15 DP), BMI (11 DP), procurement team (9 DP), length of hospital stay (5 DP), and serum amylase and/or lipase levels (5 DP). Vasopressor levels and blood glucose levels both accounted for 4 DP while donor medical and social history accounted for 5 and 4 DP, respectively. For every negative aspect of the donor’s history, the assigned amount was deducted from the total of the aforementioned variables.
A subjective assessment of the physical properties of each pancreas was performed at time of processing (Fig. 1B). Characteristics evaluated were tissue consistency, fat content, size, quality of procurement, flush, packaging and storage, and damage to the pancreas that would have occurred during procurement, cold storage, or decontamination.
Within each variable, the senior staff of the CIIL established specific ranges. The age range of 35–55 years was perceived to be the most desirable and therefore was assigned the highest value of 20 DP. Donors whose age fell into the ranges of 25–34 and 56–65 received 10 DP, and donors whose ages were <25 or >65 were assigned 5 DP. Cause of death was then assigned the next highest value (18 DP allocated). The donor’s cause of death was categorized into one of four groups: nontraumatic (anoxia, carbon monoxide poisonings, tumors, and drug overdoses) (0 DP), traumatic with injury (motor vehicle accidents, motor cycle accidents, and gunshot wounds where injury occurred to the thoracic or abdominal) (8 DP), traumatic without injury (motor vehicle accidents, motor cycle accidents, and gunshot wounds where no injury occurred to the thoracic or abdominal) (14 DP), or cerebrovascular (18 DP).
The CIT was divided into the following time ranges with the corresponding DP values in brackets; <3 hr (10 DP), ≥3; <8 hr (15 DP), ≥8; <12 (7.5 DP), and ≥12 hr (4 DP). BMI was merited 11 DP and split into the following ranges with accompanied values in brackets: <20 (2 DP), 20-24.9 (7 DP), 25-29.9 (11 DP), 30-40 (5 DP), and >40 (0 DP).
The organ procurement team variable was assigned 9 DP. If our local liver procurement team (which has significant experience procuring pancreases for islet isolation) procured the pancreas, 9 DP were awarded. If a distant procurement team procured the pancreas with no one from the CIIL present then 2 DP were awarded.
The donor’s hospital stay, serum amylase and lipase levels, and blood glucose levels were assigned the next highest value with each accounting for 5 DP. Hospital stay was divided into four ranges and had the following values. For hospital stays of <4 days, 5 DP were assigned. Longer time periods were sectioned in the following way: 5-6 days (3 DP), ≥7; <14 days (2 DP), and any length ≥14 days (0 DP).
Serum amylase and lipase levels used to determine the exocrine function of the pancreas were obtained prior to the procurement. Amylase levels were considered normal if the concentration was less than 200 U/L and lipase levels if less than 100 U/L. If the donor’s amylase and/or lipase levels fell within these limits, then 5 DP were assigned; if either amylase or lipase levels exceeded these limits by < 2 times, 2 DP were allocated. If the levels were ≥2 to ≤5 times normal levels, 1 DP was allocated while any levels ≥5 times garnered a deduction of 5 DP.
Blood glucose was divided into four ranges, including normal (3.3–11.1 mM) (4 DP), normal treated (3.3–11.1 mM with aid of insulin therapy) (3 DP), high treated (>11.1 with aid of insulin therapy) (1 DP), and high untreated (>11.1 without aid of insulin therapy) (0 DP). Vasopressors administered in high levels may have a deleterious effect on the pancreas viability and eventual outcome of the islet isolation (8). Vasopressor levels were defined as a sum of the dosage. For example, when dopamine with 15 μg/kg/min, norepinephrine with 20 μg/min, and vasopressin 2 units/hr were administered, the level was defined as 37 units. Vasopressor levels that had a total sum of <20 units were considered low, moderate vasopressor levels ranged from 20–40 units, and high vasopressor level was >40 units. Vasopressor levels administered at normal levels resulted in a score of 4 DP, whereas vasopressor levels ≤2 times normal levels resulted in 2 DP, >2 times normal levels 0 DP.
If the donor had other medical complications known to have a detrimental effect on the outcome of an islet isolation, then those were taken into consideration and further points were subtracted from the score. On the other hand, the absence of additional medical complications or history resulted in an additional score of 5 DP. Hypertension, chronic alcohol abuse, blood transfusions, and cardiac and pulmonary arrests are all believed to damage the pancreatic function as well as the ability to isolate islets from these pancreases. If the donor had a history of hypertension (controlled or uncontrolled), 2 DP were deducted. If the donor was shown to be a chronic alcoholic, a further 3 DP were deducted. The administration of blood transfusions resulted in a deduction of 1 DP. Cardiac and/or pulmonary arrests introduce warm ischemia time. If the donor had sustained any arrest time longer than 5 minutes, then 5 DP were subtracted.
An additional seven physical properties were observed once the pancreas was accepted and arrived at the CIIL. The quality of procurement, integrity and quality of packaging, quality of UW flush, pancreas damage, size, fat content, and consistency are all suspected or known to have an influence on the islet isolation. Ensuring that our protocol for procurement of the pancreas is adhered to is challenging, as distant teams procure a large majority of our organs and can only be gauged by information provided to us by the distant and local organ procurement organizations. Quality of procurement was rated qualitatively and scored as follows: very poor, –3 DP; poor, –2 DP; adequate, –1 DP; and good/excellent, no subtraction. Quality of flush was determined by the amount of blood observed in the transport UW, as well as the color of the collected Liberase HI postperfusion.
Blood has an inhibitory affect on the Liberase HI activity due to the protein present and thus can interfere with the digestion of the pancreas. Quality of flush can be very poor, poor, adequate, good and excellent having values of –3, –2, –1, 0, and 0, respectively.
Pancreases destined for the CIIL can be preserved using the TLM or solely on UW. Quality of packaging determination takes into consideration many different factors. If the TLM is used then volumes of both UW and perfluorocarbon (PFC) solutions were examined as well as verifying that the pancreas was suspended at the interface of the PFC and UW. If only UW has been used then we ensure that a sufficient volume of UW was used to preserve the pancreas. For both the preservation methods utilized by our laboratory, temperature, sufficient preservation volumes, adequate ice and package integrity were also examined. Quality of packaging evaluation is divided into five groups and had the following values. Very poor had a value of –3, poor had a value of –2, adequate had a value –1, good and excellent do not account for any point deduction. If there was any damage suffered during the procurement or transport of the pancreas then that again had a negative affect on the pancreas rating and ultimately the islet isolation. If there was an unacceptable amount of damage incurred then 2 points were deducted. If the damage was deemed minimal then 1 point was subtracted.
Unlike whole pancreas transplants, fatty pancreases are desirable for islet isolation procedures (28). Very lean pancreases had 5 points deducted, whereas lean pancreases had a 3-point deduction. If an average fat content or a fatty pancreas was observed then no points were deducted while pancreases that were grossly fatty had 2 points subtracted. The staff of the CIIL determines the consistency of the pancreas. Soft pancreases have historically not digested as well as firmer pancreases; however, pancreases that are too firm due to fibrotic tissue also do not digest. A pancreas consistency that is neither soft nor hard is a more desirable property. Soft pancreases could be attributed to many different factors including self-digestion, warm ischemia, or degranulization. A firm pancreas could be documented to factors such as pancreatitis and chronic alcoholism. If a pancreas was determined to be very soft then 5 points were subtracted, if the pancreas was deemed soft then 4 points were deducted. If the pancreas was hard, 1 point was retracted; if it was extremely hard, then 3 points were taken off.
Once all donor and pancreas information was gathered then the score from the donor variable section was tallied, this score was expressed as a percentage with the total number out of 100. Section B, which entailed the pancreases physical properties was tallied and deducted from section A, as all of section B scores were 0 or negative. The sum of section A and section B then resulted in the overall pancreas rating, from this numerical value a decision based on an established algorithm was employed. Logistic regression analysis was used to analyze all data and the statistical significance level was set at P < 0.05. Additionally the data was analyzed using the receiver-operator characteristics (ROC) curve to determine the highest positive differential rate, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. All data was analyzed using SPSS software version 11.5 (SPSS, Chicago, IL).
After analysis of 326 donor charts and isolation records, the donor scores were classified into groups thus identifying the general pancreas types that were isolated (Table 1). Of the 326 cases, our overall islet isolation to transplant success rate is 131/326 (40.2%). Donor pancreases with a DP score of ≤49.5 were considered poor donors. The 22 pancreases within this group accounted for 6.8% of all donors and had a transplant frequency of 0%. Donor pancreases that had a DP score of 50–59.5 were considered marginal donors; in this group, there were 44 pancreases that were processed, which accounted for 13.5% of all donors accepted. Of these 44 pancreases, 13 produced islets that were transplanted (29.6% transplant efficiency). Pancreases where the DP score fell with in the range of 60–69.5 accounted for 24.2% of all donors with a transplant efficiency of 40.5% (32 transplants). Donor point scores between 70 and 79.5 accounted for 30.7% (n=100) of all donors analyzed and had a transplant frequency of 35.0% (35 transplants). Donors with DP scores of 80 or above were considered optimal, Donors with scores between 80–89.5 had a donor frequency of 20.3% (n=66) and a transplant frequency of 54.6% (n=66), whereas donors with a score of 90–100 had a donor frequency of 4.6% (n=15) and a transplant frequency of 100% (n=15). Logistic regression analysis showed a significant association between the DP score and outcome of the isolation (odds ratio = 1.056, 95% CI 1.035–1.078, P < 0.001).
The data was also analyzed using ROC curve to determine the most appropriate cutoff levels. The optimum cutoff value that produced the highest differential positive rate was a DP score of 79. With this cutoff value, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 43%, 82%, 62%, 68%, and 66%, respectively.
In addition to determining the statistical significance of the DP scoring system, we were able to delineate the quality of pancreases that have been processed in our laboratory over the last 5 years (Fig. 2). It was determined that since 1999 the number of optimal pancreases have decreased from 35.3% of all pancreases to 16.3%, whereas poor-to-marginal pancreases have increased from 13.7% to 22.1%. Using this DP scoring system, we have identified 22 pancreases that in the past 5 years should have not have been processed. With the increasing costs of managing an islet isolation facility, these 22 isolations have carried a cost of $550,000 CAD plus transportation costs. Since developing the scoring system, we have generally declined all pancreases with scores less than 49.5. Since the development of this system, we have declined a total of 14 pancreases while accepting and not processing 11 more pancreases because of a low DP score. The cost of 25 isolations is $625,000 CAD plus transportation costs; using this system, we have been able to alleviate these expenses from our processing expenditures.
Ensuring success of every islet isolation remains a major hurdle. The ubiquitous application of the transplantation of islets of Langerhans to treat someone with labile diabetes is a goal that lies beyond several hurdles, including improvement of isolation success and cost effectiveness. The described multivariate system of determining donor potential for isolation success has the potential to enhance isolation success and thus avoid costs resulting from failed isolations. We have shown a strong correlation between donor score and donor success rate in islet isolation. Other groups as well as our team have established the impact of donor selection on the success of a given isolation, and moreover identified some key variables.
Donor age was weighted most heavily as a marker for donor potential in our system. Its role in isolation success is well established. Lakey et al. (3) found that the average age of donors for successful isolations was 40.9 ± 1.4 years (compared to 29.1 ± 1.7 years for unsuccessful isolations). Researchers have confirmed that isolation success suffers when donor are young (often defined as <20 years) (3,6,12), and some studies have found significant decrease in insulin content and response in older pancreases (often defined as >50 years) (3,7,18). Although Lakey et al. (3) observed lower stimulation index in islets from older pancreases (51-65 years), the isolation success rate was increased in this age group; Toso et al. have had success with pancreases >50 years (11). The employed age ranges were based both on review of these findings as well as our laboratory’s experience with varied donor ages. The major obstacle that is involved with isolating islets from young donors is the inability to liberate islets from the exocrine tissue. This in turn results in packed cell volume that is greater than the 5 cc, our current guideline. Young donor pancreases can and will play an integral role in the future of islet isolation and the ability of these donors to produce successful single donor islet transplants. In an recent study by Hering et al., successful single donor transplants (4/6) were achieved with an average donor age of 25 years and low CIT (< 6 hrs), these islets produced high stimulation indexes (4.6–27.1) and the ability to normalize diabetic nude mice with a marginal islet mass. The evolution of surgical transplant procedures will help programs to transplant islet preparations that were previously unusable because of the large tissue volume and the threat of complications using the percutaneous transhepatic approach.
Cause of death was weighted significantly as it has a strong influence on the out come of an islet isolation (3,6). Cerebrovascular causes of death scored the highest and were considered most ideal for islet isolations as older donors primary cause of death is from cerebrovascular events. We have had the greatest isolation success rate in this age category and thus merited 18 DP. Traumatic death was further separated into with or with out injury depending on the severity of trauma the donor sustained. Traumatic causes of death such as motor vehicle accidents generally occur to younger donors so this is taken into consideration, as the ability to isolate islets effectively from young donors is limited right now. Furthermore, if the donor has sustained any injury to the abdominal area that might injure the pancreas the DP amount scored was reduced. Nontraumatic causes of death have been proven in the past to be detrimental to the success of the isolation so this warranted a greater deduction.
CIT was furthermore weighted heavily as a marker for isolation success. Prolonged CIT time has been associated with damage to the donor pancreas and thus decreased islet yield (3,6,11,19,20,27). The development of the TLM has done much to reduce the negative effect of increased CIT on islet isolation (13,21), potentially offering resuscitation to the ischemically damaged pancreas (22). However, the significant impact of CIT is undeniable, and thus it merited heavy weighting within our system.
BMI has been shown to affect isolation success, such that pancreases from overweight donors (generally BMI >25) provide higher yield (3,5,11). However, obese donors (BMI >30) are at risk for type II diabetes mellitus (23,24), which would result in the impairment of islet function. We therefore chose to give a higher score to donor pancreases with BMI 25–30 than those <25 and >30. The procurement team has moreover been shown to affect the islet isolation success (3). In addition to being assigned 9 DP, procurement is given further consideration during the qualitative assessment of the pancreas upon being received. The quality of procurement, integrity and quality of packaging, quality of UW flush, and pancreas damage were considered during the initial stages of isolation, all of which are intrinsically related to the procurement team. Other donor characteristics such as vasopressor (3,8,26), amylase (26), and glucose (3,26) levels, known to affect isolation recovery and/or islet insulin response, were all considered.
Many isolation laboratories use donor variables to subjectively qualify the potential of a donor preisolation. The described scoring system provides an opportunity for more objective and consistent donor selection, thus standardizing the donor-screening process. The scoring system also provides an opportunity to further optimize isolation outcomes from more marginal organs. Presently, pancreases with a perceived low qualitative potential are immediately processed or discarded. By establishing that a given pancreas has a low quantitative donor potential score, the laboratory might choose, for instance, to employ techniques for resuscitating the pancreas, as has been proposed using the TLM (14,21,22,28).
Although the tool is promising, we clearly recognize that a number of limitations exist. For instance, although the score is numerical, it is still subjective, and the scoring process was determined arbitrarily based on the experience of our isolation team. Furthermore, the weight of different factors was based on a review of the literature as well as individual experience with various types of donors, and is thus also subjective. It is also important to note the systems dynamic nature, as new technologies and techniques are developed to further refine islet isolation, so too will the scoring system be refined to increase the sensitivity. The sensitivity of this donor screening process will improve as we come to better understand the dynamic nature of islet isolations and the effect donor variables have on the outcome of isolations. The ideal pancreas donor for islet isolation has not changed but has been more clearly defined, so it is important to note that using a more stringent selection criteria will not increase competition between islet isolation facilities and whole pancreas transplant programs.
Limitations aside, this scoring system has proven to be a useful guide to help determine the likelihood of successful islet isolation. Future studies may further refine this method and test its success at other islet isolation centers. Scoring for donor potential has proven to be extremely useful to our laboratory, and may offer a similar effect to other centers seeking to maximize cost efficiency and improve isolation success.
The authors would like to thank the staff of the CIIL for their expertise and the HOPE coordinators for the identification of all donors.
1. Kobayashi N, Okitsu T, Lakey JR, Tanaka N. The current situation in human pancreatic islet transplantation: problems and prospects. J Artif Organs
2004; 7(1): 1.
2. Shapiro AMJ, Lakey JRT, Ryan EA, et al. Islet Transplantation in Seven Patients with Type 1 Diabetes Mellitus Using a Glucocorticoid-Free Immunosuppressive Regimen. N Engl J Med
2000; 343(4): 230.
3. Lakey JR, Warnock GL, Rajotte RV, et al. Variables in organ donors that affect the recovery of human islets of Langerhans. Transplantation
1996; 61(7): 1047.
4. Lakey JR, Rajotte RV, Warnock GL, Kneteman NM. Human pancreas preservation prior to islet isolation. Cold ischemic tolerance. Transplantation
1995; 59(5): 689.
5. Brandhorst H, Brandhorst D, Hering BJ, et al. Body mass index of pancreatic donors: a decisive factor for human islet isolation
. Exp Clin Endocrinol Diabetes
1995; 103 Suppl 2: 23.
6. Zeng Y, Torre MA, Karrison T, Thistlethwaite JR. The correlation between donor characteristics and the success of human islet isolation
1994; 57(6): 954.
7. Benhamou PY, Watt PC, Mullen Y, et al. Human islet isolation
in 104 consecutive cases. Factors affecting isolation success. Transplantation
1994; 57(12): 1804.
8. Lakey JR, Warnock GL, Rajotte RV, et al. Factors in cadaveric donors that affect recovery of human islets of Langerhans. Transplant Proc
1995; 27(6): 3265.
9. Sabek O, Hathaway D, Fraga D, Gaber AO. Influence of Human Donor Factors on Pancreatic Collagenase Digestion. Transplant Proc
1998; 30(2): 353.
10. Chiarelli AG. Donor and organ suitability: risk due to donor social behavior. Transplant Proc
2004; 36(3): 426.
11. Toso C, Oberholzer J, Ris F, et al. Factors affecting human islet of Langerhans isolation yields. Transplant Proc
2002; 34(3): 826.
12. Brandhorst D, Hering BJ, Brandhorst H, et al. Influence of donor data and organ procurement on human islet isolation
. Transplant Proc
1994; 26(2): 592.
13. Ricordi C, Fraker C, Szust J, et al. Improved human islet isolation
outcome from marginal donors following addition of oxygenated perfluorocarbon to the cold-storage solution. Transplantation
2003; 75(9): 1524.
14. Tsujimura T, Kuroda Y, Kin T, et al. Human islet transplantation
from pancreases with prolonged cold ischemia using additional preservation by the two-layer (UW solution/perfluorochemical) cold-storage method. Transplantation
2002; 74(12): 1687.
15. Hering BJ, Kandaswamy R, Harmon JV, et al. Transplantation of cultured islets from two-layer preserved pancreases in type 1 diabetes with anti-CD3 antibody. Am J Transplant
2004; 4(3): 390.
16. Ricordi C, Lacy PE, Scharp DW. Automated islet isolation from human pancreas. Diabetes
1989; 38 Suppl 1: 140.
17. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta Diabetol Lat
1990; 27(3): 185.
18. Sodoyez J, Van Schavendijk C, Vercammen M, et al. Decreased content in B cell secretory peptides in the human pancreas after the fifth decade. Diabetes
19. Kinasiewicz A, Juszczak M, Pachecka J, Fiedor P. Pancreatic islets isolation using different protocols with in situ flushing and intraductal collagenase injection. Physiol Res
2004; 53(3): 327.
20. Kneteman NM, Lakey JRT, Warnock GL, Rajotte RV. Human islet isolation
after prolonged cold storage. Diab Nutr Metab
1992; 5 (Suppl. 1): 33.
21. Tsujimura T, Kuroda Y, Avila JG, et al. Influence of pancreas preservation on human islet isolation
outcomes: impact of the two-layer method. Transplantation
2004; 78(1): 96.
22. Tsujimura T, Kuroda Y, Avila JG, et al. Resuscitation of the ischemically damaged human pancreas by the two-layer method prior to islet isolation. Transplant Proc
2003; 35(7): 2461.
23. Norris SL, Zhang X, Avenell A, et al. Efficacy of pharmacotherapy for weight loss in adults with type 2 diabetes mellitus: a meta-analysis. Arch Intern Med
2004; 164(13): 1395.
24. Reynolds LR, Anderson JW. Practical office strategies for weight management of the obese diabetic individual. Endocr Pract
2004; 10(2): 153.
25. Fiedor P, Goodman ER, Sung RS, et al. The effect of clinical and biochemical donor parameters on pancreatic islet isolation yield from cadaveric organ donors. Ann Transplant
1996; 1(1): 59.
26. Watt PC, Mullen Y, Benhamou PY, et al. Donor factors affecting successful isolation of islets from the human pancreas. Transplant Proc
1994; 26(2): 594.
27. Matsumoto S, Kuroda Y, Hamano M, et al. Direct Evidence of pancreatic tissue oxygenation during preservation by the two layer method. Transplantation
1996; 62(11): 1167.
28. Matsumoto I, Sawada T, Nakano M, et al. Improvement in islet yield from obese donors for human islet transplants. Transplantation
2004; 78(6): 880.