The impact of surging transplantation of alcohol-associated liver disease on transplantation for HCC and other indications

Abstract

Background: Liver transplantation (LT) for alcohol-associated liver disease (ALD) is increasing and may impact LT outcomes for patients listed for HCC and other indications.

Methods: Using US adults listed for primary LT (grouped as ALD, HCC, and other) from October 8, 2015, to December 31, 2021, we examined the impact of center-level ALD LT volume (ATxV) on waitlist outcomes in 2 eras: Era 1 (6-month wait for HCC) and Era 2 (MMaT-3). The tertile distribution of ATxV (low to high) was derived from the listed candidates as Tertile 1 (T1): <28.4%, Tertile 2 (T2): 28.4%-37.6%, and Tertile 3 (T3): >37.6% ALD LTs per year. Cumulative incidence of waitlist death and LT within 18 months from listing by LT indication were compared using the Gray test, stratified on eras and ATxV tertiles. Multivariable competing risk regression estimated the adjusted subhazard ratios (sHRs) for the risk of waitlist mortality and LT with interaction effects of ATxV by LT indication (interaction p).

Results: Of 56,596 candidates listed, the cumulative waitlist mortality for those with HCC and other was higher and their LT probability was lower in high (T3) ATxV centers, compared to low (T1) ATxV centers in Era 2. However, compared to ALD (sHR: 0.92 [0.66-1.26]), the adjusted waitlist mortality for HCC (sHR: 1.15 [0.96-1.38], interaction p = 0.22) and other (sHR: 1.13 [0.87-1.46], interaction p = 0.16) were no different suggesting no differential impact of ATxV on the waitlist mortality. The adjusted LT probability for HCC (sHR: 0.89 [0.72-1.11], interaction p = 0.08) did not differ by AtxV while it was lower for other (sHR: 0.82 [0.67-1.01], interaction p = 0.02) compared to ALD (sHR: 1.04 [0.80-1.34]) suggesting a differential impact of ATxV on LT probability.

Conclusions: The high volume of LT for ALD does not impact waitlist mortality for HCC and others but affects LT probability for other in the MMAT-3 era warranting continued monitoring.

Graphical abstract

FIGURE 1

Study flowchart of the total study cohort.

FIGURE 2

Distribution of study subjects by ALD transplant volume percent and era. Alcohol transplant volume (ATxV) was quantified as the center-level percentage of adult primary LTs for ALD (without HCC) each year (a relative proportion within a center). The number of listings by ATxV and era are shown in this figure. Abbreviations: ALD, alcohol-associated liver disease; LT, liver transplantation.

FIGURE 3

Cumulative incidence within 18 months of listing for waitlist mortality. Cumulative incidence within 18 months of listing during Era 1 for (A) ALD, (B) HCC, and (C) other and Era 2 for (D) ALD, (E) HCC, and (F) Other is shown in Figure 2. Waitlist mortality was significantly higher among those with HCC (E) and other (F) in Era 2 in the highest ATxV tertile (T3). Abbreviations: ALD, alcohol-associated liver disease; ATxV, transplant volume for alcohol-associated liver disease.

FIGURE 4

Multivariable subhazard ratios for waitlist mortality by indication and listing era. Adjusted waitlist mortality subhazard ratios for candidates listed at high (T3) versus low (T1) ATxV centers did not differ significantly among those with ALD (Era 1: 0.90 [0.77–1.04]; Era 2: ALD 0.92 [0.66–1.26]), HCC (Era 1: 0.95 [0.80–1.13], interaction p = 0.35; Era 2: 0.95 [0.80–1.13], interaction p = 0.22), and other (Era 1: 0.95 [0.82–1.09], interaction p = 0.33; Era 2: 1.13 [0.87–1.46], interaction p = 0.16) suggesting no differential effect of ATxV on waitlist mortality by indication. Multivariable models were adjusted for characteristics at listing including etiology of liver disease, sex, race/ethnicity, public insurance, diabetes, dialysis, BMI, Child-Pugh class, MELD, region, age, and blood type. Abbreviations: ALD, alcohol-associated liver disease; ATxV, transplant volume for alcohol-associated liver disease.

FIGURE 5

Cumulative incidence within 18 months of listing for liver transplantation. Cumulative incidence of liver transplantation within 18 months of listing during Era 1 for (A) ALD, (B) HCC, and (C) Other and Era 2 for (D) ALD, (E) HCC, and (F) other is shown in the figure. The cumulative incidence of LT was significantly lower among those with HCC (B, E) and Other (C, F) in the highest ATxV tertile (T3) in Eras 1 and 2. The cumulative incidence of LT was significantly higher among those with ALD diagnosis (A) in the highest ATxV tertiles (T2, 3) in Era 1. Abbreviations: ALD, alcohol-associated liver disease; ATxV, transplant volume for alcohol-associated liver disease; LT, liver transplantation.

FIGURE 6

Multivariable subhazard ratios for liver transplant probability by indication and listing era. The effect of high (T3) versus low (T1) ATxV on LT differed by the etiology of liver disease. Compared to those with ALD (Era 1: 1.27 [1.07–1.52]; Era 2: ALD 1.04 [0.80–1.34]), the adjusted subhazard ratios were lower among those with HCC (Era 1: 1.00 [0.85–1.18], interaction p = 0.001; Era 2: 0.89 [0.72–1.11], interaction p = 0.08) and other (Era 1: 1.04 [0.84–1.24], interaction p = 0.006; Era 2: 0.82 [0.67–1.01], interaction p = 0.02). Multivariable models were adjusted for characteristics at listing including etiology of liver disease, sex, race/ethnicity, public insurance, diabetes, dialysis, BMI, Child-Pugh class, MELD, region, age, and blood type. Abbreviations: ALD, alcohol-associated liver disease; ATxV, transplant volume for alcohol-associated liver disease; LT, liver transplantation.