Background:  HIV infection is associated with rapid progression of liver disease and may influence outcomes in liver-transplant candidates. While the model for end stage liver disease (MELD) score is accepted as a reliable predictor of mortality in HIV-uninfected transplant candidates, factors that predict mortality have not been established in HIV-infected candidates.

Methods:  HIV⁺ liver transplant candidates without hepatocellular carcinoma enrolled in the Solid Organ Multi-Site Transplant Study (HIVTR) were matched 1:5 with controls from the United Network of Organ Sharing (UNOS) on age, gender, race, time period, and hepatitis C virus (HCV) infection. Time to death was compared and predictors of pre-transplant mortality and elevation of MELD ≥25 were examined by proportional hazards models. Contrasts with controls were based on clustered sampling.

Results:  Of 167 HIVTR subjects, 58 (34.7%) were transplanted and 24 (14.4%) died prior to transplant. The pre-transplant mortality rate of 24/167 (14.4%) was similar to that of UNOS controls, 88/792 (11.1%), p = 0.30; there was no difference between HCV/HIV co-infected subjects, 18/125 (14.4%) and HCV-infected UNOS controls, 62/592 (10.5%), p = 0.28. Cumulative incidence of death (p = 0.15), transplant (p = 0.43), and elevation of MELD ≥25 (p = 0.50) were similar for HIVTR and controls. In both groups baseline MELD was a significant predictor of pre-transplant mortality by proportional hazards model (HR 1.27; p <0.001), with sepsis and multiple organ system failure the main causes of death. In the HIVTR group, those who died had lower median CD4 count at enrollment (237 cells/mm³) than those transplanted (315 cells/mm³) or non-transplanted (264 cells/mm³), p = 0.027. The proportion with detectable HIV RNA did not differ between those who died 5/24 (21.4%), were transplanted 9/57 (16.2%), or not transplanted 6/84 (7.1%), p = 0.09; but detectable HIV RNA was associated with an increased hazard of death (HR = 3.18; p = 0.02) and faster progression to MELD ≥25 (HR = 2.79, p = 0.005). After controlling for CD4 count, detectable HIV RNA and HAART use at enrollment, the only significant predictor of mortality was baseline MELD (HR 1.28; p <0.001).

Conclusions:  HIV⁺ liver transplant candidates have similar pre-transplant mortality characteristics as HIV^– controls. While lower CD4 counts and detectable HIV RNA are associated with death, baseline MELD appears to be the only significant predictor of pre-transplant mortality in HIV-infected liver transplant candidates.