RESUMO
PURPOSE: We sought to build prediction models for organ transplantation and recipient survival using both biomarkers and clinical information. MATERIALS AND METHODS: We abstracted clinical variables from a previous randomized trial (nâ¯=â¯556) of donor management. In a subset of donors (nâ¯=â¯97), we measured two candidate biomarkers in plasma at enrollment and just prior to explantation. RESULTS: Secretory leukocyte protease inhibitor (SLPI) was significant for predicting liver transplantation (C-statistic 0.65 (0.53, 0.78)). SLPI also significantly improved the predictive performance of a clinical model for liver transplantation (integrated discrimination improvement (IDI): 0.090 (0.009, 0.210)). For other organs, clinical variables alone had strong predictive ability (C-statistic >0.80). Recipient 3-years survival was 80.0% (71.9%, 87.0%). Donor IL-6 was significantly associated with recipient 3-years survival (adjusted Hazard Ratio (95%CI): 1.26(1.08, 1.48), Pâ¯=â¯.004). Neither clinical variables nor biomarkers showed strong predictive ability for 3-year recipient survival. CONCLUSIONS: Plasma biomarkers in neurologically deceased donors were associated with organ use. SLPI enhanced prediction within a liver transplantation model, whereas IL-6 before transplantation was significantly associated with recipient 3-year survival. Clinicaltrials.gov: NCT00987714.