*Purpose: The current liver transplantation (LT) selection paradigm for hepatocellular carcinoma (HCC) patients ignores the variable risks of waitlist dropout and post-LT recurrence/death, with the same priority assigned to all eligible patients. We developed a novel risk model to identify HCC patients with the greatest survival benefit from LT.

*Methods: Clinicoradiologic and outcomes data of consecutively waitlisted HCC patients (2006-2015) at a single center were evaluated. Multivariate Fine and Gray competing-risk analysis identified independent predictors of HCC tumor progression, liver disease related waitlist dropout, and post-LT HCC recurrence, both at baseline and at the landmark 6 month time point. Survival benefit was defined as the predicted risk ratio of an adverse outcome following LT vs remaining on the waitlist, with patients stratified into 5 groups based on predicted 5-yr post-LT recurrence-free-survival (RFS) and survival benefit of LT > or < 25%.

*Results: Of 881 HCC patients, 366 (41.5%) had wait-list dropout (median 16.0 months) and 515 (58.5%) underwent LT (median 15.6 months), with tumor progression (n=190) and non-HCC related liver complications (n=170) accounting for 51.9% and 46.4% of dropouts, respectively (Figure 1A). Of patients who underwent LT, 64 (12.4%) recurred and 102 (19.8%) died with no recurrence (Figure 1B, median follow-up 43.2 months). At baseline, AFP and radiologic size criteria independently predicted HCC-related dropout (c-stat 0.64), and MELD predicted liver related dropout (c-stat 0.65). At 6 months following diagnosis, AFP response to LRT, mRECIST, and radiologic size criteria improved HCC related dropout and post-LT recurrence model performance (c-stat 0.78 and 0.77, respectively). Model identified futile outcomes (< 50% RFS) in 23.7% of patients. In the remaining patients, multivariate predictors of greatest survival benefit (> 25% risk reduction of adverse event with LT vs waitlist) included MELD at baseline and change in MELD at 6 months, AFP max and AFP response, radiologic size criteria at baseline, and mRECIST response (Figure 1C).

*Conclusions: We report a transplant survival benefit model for individualized risk prediction of HCC and non-HCC related mortality with and without liver transplantation, allowing for differential prioritization of HCC candidates.

« Back to 2022 American Transplant Congress