Is allocation of liver grafts based on major Extended Donor Criteria plausible? Proposal of an allocation algorithm

Background:

Donor age > 65 years, biopsy proven macrovesicular steatosis (BPS) > 40%, and cold ischemia time (CIT) > 14 hours are major EDC (maEDC) that decrease short- and long-term graft survival, and long-term patient survival after liver transplantation (LT). We aimed to analyze the impact of maEDC in combination with different labMELD-scores on post-LT outcome and examine the plausibility of a maEDC-based allocation algorithm.

Methods:

All consecutive LTs (n = 465) since December 2006 were included in the study. We considered the graft to be good if none of the maEDC (donor age > 65y, BPS > 40%, CIT > 14h) were present, whereas grafts with ≥1 maEDC were considered to be marginal. Patients with labMELD score ≥20 were considered to be recipients at high risk, and those with labMELD score < 20 were considered as recipients at low risk. We compared the groups as follows: group A (good graft-to-low risk recipient; n = 122); group B (good graft-to-high risk recipient; n = 89); group C (marginal graft-to-low risk recipient; n = 156); group D (marginal graft-to-high risk recipient; n = 85). We analyzed graft failure and mortality rates and compared differences between the groups.

Results:

We observed lowest graft failure rates at 90 days, 1 year, and 5 years post-LT in group A, and highest in group D (p = 0.009, 0.040, and 0.010 adequately). Similarly, the highest 5-year patient mortality rate was that of group D (16.5%), and the lowest was that of group A (4.9%) (p = 0.042). Overall 5-year graft survival was significantly higher in recipients at risk (labMELD ≥20) who were transplanted with grafts with one maEDC compared to recipients at risk who received organs with two maEDC (p = 0.034), whereas no significant difference was observed in low risk patients (labMELD < 20) who received grafts with one or two maEDC (p = 0.140). Overall 5-year patient survival didn't differ between high risk recipients (labMELD ≥20) transplanted with grafts with one and two maEDC (p = 0.464). Similarly, no difference in patient survival was observed in low risk recipients (labMELD < 20) who received grafts with one or two maEDC (p = 0.165).

Conclusion:

Allocation algorithm based on maEDC is plausible and necessary as it may improve post-LT outcome. We suggest that donor organs with one maEDC could be allocated to recipients with labMELD ≥20, whereas grafts with two maEDC should be preferably allocated to recipients with labMELD < 20 or to oncologic patients. Grafts with three maEDC might be preferred for carefully selected oncologic patients with hepatocellular or cholangiocellular cancer, colorectal liver metastases or neuroendocrine tumors in rigorously controlled clinical trials.