6.333333
#7
COMPUTERIZED MATCHING ALGORITHMS TO OPTIMIZE MATCHES AND INCREASE OPPORTUNITIES FOR LIVE DONOR KIDNEY EXCHANGE.
Susan L. Saidman Ph.D 1, Alvin E. Roth Ph.D 2, Tayfun Sonmez Ph.D 2,3, M. Utku Unver Ph.D 3 and Francis L. Delmonico M.D 1,4. 1 Massachusetts General Hospital, Boston, MA ; 2 Harvard University, Cambridge, MA ; 3 Koc University, Istanbul, Turkey and 4 New England Organ Bank, Newton, MA .
Paired kidney donation between incompatible (due to ABO or positive XM) donor/recipient (D/R) pairs can increase transplant (Tx) opportunities for such patients (pts). In an effort to expand the opportunity for such donations, we designed computerized matching algorithms to identify maximal sets of compatible D/R pairs from a registry of incompatible pairs. To demonstrate the effectiveness of the program, data (ABO, HLA, PRA, recipient antibody specificity) from pts and their incompatible donors (45 pts, 68 D/R pairs) were submitted for computer analysis and matching. A compatible donor was identified for 29/45 pts (64%), but only 15 D/R pairs (33%) had another pair that was mutually compatible with them. If only exchanges involving 2 D/R pairs were allowed, 8 pairs could exchange kidneys. If 3-way exchanges were allowed, 11 D/R pairs could exchange kidneys. Simulations based on OPTN/SRTR data were used to further assess the importance of multiple exchange combinations. Incompatible D/R pairs were randomly generated and Monte-Carlo simulations demonstrated that the number of Txs could be substantially increased if 3-way exchanges were allowed. Exchanges were identified even for relatively disadvantaged ABO-O or high PRA pts, especially when using 3-way exchanges. In conclusion, a computer algorithm designed to optimize matches was used to identify D/R pairs who can potentially enter into donor exchanges. Such matching algorithms will be important in the development of regional and/or national exchange programs. A substantial number of additional pts can receive Txs if 3-way exchanges are allowed.