Validation of the Scoring System for Standardization of the Pancreatic Donor for Islet Isolation as Used in a New Islet Isolation Center P. Witkowski, Z. Liu, S. Cernea, Q. Guo, E. Poumian-Ruiz, K. Herold, and M.A. Hardy ABSTRACT Background. The aim of this study was to evaluate the effectiveness of the Edmonton Donor Scoring System for use in our much less active islet center. Because the ability to recognize an appropriate donor may help to achieve consistent and predictable success of pancreatic islet isolation, it should lead to increased effectiveness and lower cost. Material and Methods. Charts of 36 consecutive pancreas donors were reviewed to assess the donor points (DP). DP ranged from 0 to 100 based on donor age, body mass index, cause of death, social and medical history, hospital stay, vasopressor dosages, laboratory tests, cold ischemia time and procurement team, as well as pancreas size, consistency, fat content, damage, and quality of procurement and packing. Results. Successful isolation was achieved in 39% of donors (14 of 36), a value similar to that achieved in Edmonton (40%). We used the optimal cutoff value (DP ⫽ 79) proposed by the Edmonton group. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 66%, 75%, 57%, 82% and 72%, respectively. Successful islet isolation from poor or marginal donors (DP ⬍ 49.5 and 50 to 59.5) was 0% and 28.6% respectively; it was 63% and 100% in optimal donors (DP ⫽ 80 to 89.5 and 90 to 100). We concluded that islet isolation success correlated with the previously proposed donor scoring system. Conclusions. The Donor Scoring System can be successfully implemented regardless of the level of activity of an experienced isolation center. This system permits identification of a suitable donor prior to organ processing. It may guide a center’s donor selection strategy based on its goals and its budget.
C
ONSISTENT AND predictable success of pancreatic islet isolation remain a major hurdle to successful transplantation outcomes and to reduction of costs, even in the most experience centers. The Edmonton Donor Scoring System DSS1 has never been validated by other centers, especially those with a lower level of activity. We therefore evaluated the effectiveness of the Edmonton DSS to predict the success of islet isolation in our center.
Ricordi method. They were assessed as successful when the yields met all criteria for clinical transplantation (at least 280,000 IEQ). Pancreata were initially preserved using the two-layer method (PFC/UW) whenever it was available, otherwise, we used cold UW solution. We used chi-square tests for statistical analysis.
RESULTS
MATERIALS AND METHODS
Successful isolation, achieved in 39% of donors (14 of 36), was similar to that reported by Edmonton (40%). We used the optimal cutoff value (DP ⫽ 79) proposed by the Edmonton
The charts of 36 consecutive deceased donors of pancreatic islets from 2004 and 2005 were retrospectively reviewed to assess their donor points (DP), according to the previously described Edmonton DSS.1 The range of DP, which extended from 0 to 100, was based on donor age, body mass index (BMI), cause of death, social and medical history, hospital stay, vasopressor dosages, laboratory tests, cold ischemia time, and procurement team as well as pancreas size, consistency, fat content, damage, and quality of procurement and packing. Isolations were performed using the automated
From the Department of Surgery (P.W., Z.L., Q.G., M.A.H.), and the Department of Medicine (S.C., E.P.-R., K.H.), Columbia University, New York, New York and the Department of Surgery, Medical University of Gdansk (P.W.), Gdansk, Poland. Piotr Witkowski is a Naomi Berrie Fellow. Address reprint requests to Dr P. Witkowski, Columbia University, Department of Surgery, 520 west 218th St, Apt 2C, P&S 17-514, New York, NY 10034. E-mail:
[email protected]
© 2006 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/06/$–see front matter doi:10.1016/j.transproceed.2006.08.143
Transplantation Proceedings, 38, 3039 –3040 (2006)
3039
3040
WITKOWSKI, LIU, CERNEA ET AL
Table 1. Pancreatic Donor and Successful Isolation Frequency Stratified by the Donor Scoring System and Compared With Results From the Edmonton Group Score
Donor Quality
0–49.5 50–59.5 60–69.5 70–79.5 80–89.5 90–100 Total
Poor Marginal Intermediate Optimal
Number of Donors and Their Frequency (%)
Number of Successful Isolations
Frequency of Successful Isolations (%)
P
Frequency of Successful Isolations at Edmonton (%)
1 (2.8%) 7 (19.4%) 1 (2.8%) 15 (41.7%) 11 (30.6%) 1 (2.8%) 36 (100%)
0 2 0 4 7 1 14 (100%)
0 28.6 0.0 26.7 63.6 100.0 38.9
⬎.05 ⬎.05 — ⬎.05 ⬎.05 ⬎.05 ⬎.05
0 29.6 40.5 35 54.6 100 40.2
group, which led to the highest differential positive rate. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 66%, 75%, 57%, 82%, and 72%, respectively, values that were similar to the levels calculated in Edmonton, namely, 43%, 82%, 62%, 68%, and 66%, respectively (P ⬎ .05). Pancreatic donor frequency and successful isolation frequency determined according to DSS are presented in Table 1. All islet isolations from poor donor organs (DP ⬍ 50) were unsuccessful at our center, similar to the results obtained at Edmonton. The frequency of successful islet isolations from marginal donors (DP 50 to 59.5) was similarly low at 28.6% and 29.6% in our and in the Canadian center, respectively (P ⬎ .05). Results for intermediate quality donors (DP 60 to 69.5) were biased and unreliable; we performed only one such isolation. Procedures utilizing organs from optimal donors (DP 80 to 100) performed by us and by the Edmonton group were similarly successful with success frequencies of 66.6% and 62%, respectively (P ⬎ .05). Overall, islet isolation success versus failure rates in our center correlated well with the donor score described by Edmonton. DISCUSSION
Although the techniques for islet isolation have been standardized at most centers, the result and yields are frequently inconsistent and often unpredictable, partially owing to a wide range of donors and organ quality. To overcome this variable, the Edmonton group validated a donor scoring system based on its own high volume of isolations and extensive experience. This scoring system was based on a multivariable analysis of the factors, which most influenced the success of islet isolation. In our retrospective study, we compared the outcome of our isolations with those obtained at Edmonton using groups segregated by different donor quality as described by the DSS. Although our number of islet isolations was 3-fold lower than Edmonton’s, we obtained comparable results based on the similarity of donor quality as defined by DSS. The reproducibility and reliability of this scoring system stems from inclusion of several critical factors, including the quality of the organ along with the final donor maintenance, anatomy, age, and other features of the donor pancreas, as well as the techniques used in organ harvesting and shipping. It appears that the DDS analysis has a high negative predictive value. Using this approach, poor quality pancreas donors can be quickly identified and rejected by the islet center, saving both
time and money (25,000 to 40,000 USD) by avoiding the expense of an islet isolation that most likely would fail. DSS permits establishment of clearer, more objective criteria for organ acceptance by islet isolation centers, and more effective organ allocation by local Organ Procurement Organizations. Different quality organs could be distributed partially according to the needs of the centers. DSS also permits a rapid initial evaluation of the donor, leading to implementation of additional supportive procedures which may increase the final DP score for the organ and the chance for a successful isolation. Because it has already been shown that a procurement team affiliated with an islet isolation center offers a major advantage to the eventual quality of the organ used for islet isolation, it may be concluded that in cases of marginal or intermediate quality donors, a specifically designated harvesting team may significantly increase the organ score and probability of successful isolation. Similarly, with appropriate warning and analysis the two-layer method of preservation may be used instead, or UW organ preservation to improve the results.2,3 We strongly feel that inclusion into the DDS of the method of organ preservation may be a critical factor to increase accuracy and specifity as well as the positive predictive value of the DDS. The advantages of the two-layer method over UW cold storage for both short- and long-term organ preservation has been previously described.2,3 We were unable to establish an accurate number of DDS points for these two types of preservations, because we performed too few isolations to conduct a multivariable analysis. We concluded that DSS is a useful instrument for both donor selection and organ quality assessment; it can improve effectiveness of all islet isolation centers. In conclusion, the Edmonton DSS was successfully implemented regardless of the level of activity of an experienced isolation center. This system permits identification of a suitable donor prior to organ processing and may guide a center’s donor selection strategy based on its goals and its budget. REFERENCES 1. O’Gorman D, Kin T, Murdoch T, et al: The standardization of pancreatic donors for islet isolations. Transplantation 80:801, 2005 2. Witkowski P, Liu Z, Guo Q, et al: Two-layer method in short-term pancreas preservation for successful islet isolation. Transplant Proc 37:3398, 2005 3. Tsujimura T, Kuroda Y, Avila JG, et al: Influence of pancreas preservation on human islet isolation outcomes: impact of the two-layer method. Transplantation 78:96, 2004