Ex-Vivo Lung Perfusate Cell Death Markers May Predict Long Term Outcomes After Transplantation

Purpose: Ex-Vivo Lung Perfusion (EVLP) is an emerging technology to reas- sess marginal donor lungs for transplantation. We have previously reported that cell death related perfusate bio-markers in EVLP are potentially predic- tive for primary graft dysfunction (PGD). Our objective was to investigate whether EVLP perfusate cell death markers could be used to predict post- transplant long term survival. Methods: Caspase-cleaved cytokeratin 18 (M30) and HMGB-1 levels in the perfusate of EVLP of marginal donor lungs were measured by quantitative enzyme linked immunosorbent assay, and were correlated with short and long term clinical outcome. Results: From 100 sequential clinical EVLP cases, 79 lungs were selected to be transplanted. Recipients that were bridged with pre-transplant extra- corporeal life support system were excluded from this analysis. 11 patients (PGD group) developed PGD 3 at any time within 72 h after transplant following bilateral lung transplant (BLTx). 34 patients (Control group) did not develop PGD grade 3 after BLTx. Levels of M30 were signifi- cantly higher in PGD group compared to Control group at 1 h (Control 53.9±15.9 vs. PGD 73.3± 24.9 U/L, p= 0.004) and 4 h (Control 72.4±40.0 vs. PGD 137.0±146.6 U/L; p= 0.04) of EVLP . The increase of HMGB-1 from 1 h to 4 h of EVLP (delta HMGB-1) in PGD group was signifi- cantly higher than Control (Control 7.2±16.8 vs. PGD 37.0±25.4 ng/ml; p= 0.0002). Uni variate Cox regression analysis revealed that higher M30 at 1 h (HR= 1 .03 (95%CI 1.004-1.058) per 1 U/L; p= 0 .025) and 4 h (HR= 1.009 (95%CI 1.002-1.016) per 1 U/L; p= 0.009), and higher delta HMGB-1 (HR= 1.033 (95%CI 1.006-1.061) per 1 ng/ml; p= 0.016) are sig- nificantly predictive for survival. After adjustment for PGD using bivariate cox regression analyses, M30 at 4h remained significant (p= 0.04) and the other markers were not statistically significant (M30 at 1 h; p= 0.13; delta HMGB-1; p= 0 .10). The significance of M30 at 4 h seen after adjustment for PGD may suggest that the marker provides additional predictive power for survival above that predicted by PGD status alone. Conclusion: We have demonstrated a correlation of cell death markers in the EVLP perfusate with PGD outcome and long term survival. Further research and validation is required to determine which of these biomarkers will be important in repertoire for real time predictive evaluation of the ex vivo perfused donor lung.