Background: We designed a 21-center study to test the hypothesis that immunosuppression does not accelerate HIV disease progression in transplant recipients with relatively intact immune systems and suppressed viremia. We further hypothesized that HIV+ patients would be higher risk but acceptable transplant candidates, similar to other higher risk groups. We previously described 3-year outcomes in 150 kidney (KTR) and 125 liver recipients (LTR). We now describe 5-year outcomes compared with HIV-negative controls.
Methods: We examined time to graft failure, death-censored graft failure, and death. Controls were identified from Scientific Registry of Transplant Recipients (SRTR). We fit 4 proportional hazards (PH) regression models: risk-matched and demographic-matched models examining HIV status, a demographic-matched model adjusting for risk score, and an unmatched model. We calculated case risk scores from a PH model adjusting for predictors among SRTR controls, identifying 4 controls per case after ranking by risk score. We created demographic-matched sets from randomly selected controls matched for age, gender, race, donor type, and time of transplant.
Results: In KTR, risk-matched and unmatched analyses indicated a marginally significant hazard ratio (HR) for graft loss (HR 1.4 [p=0.052] and 1.3 [p=0.07]), but no significant increase in risk of other events among HIV-positive KTR. All models demonstrated a statistically significantly higher relative hazard of graft loss or death (except risk-matched and death) in HIV-positive LTR. The absolute difference in the proportion of deaths was 6.7% in the risk-matched control analysis. HIV status was not significant in death-censored graft failure models.
Conclusions: In the risk-, demographic- and unmatched analyses, HIV-negative KTR had outcomes that were not statistically different compared with controls, suggesting that renal transplantation should be standard of care for HIV-positive patients with end stage renal disease. The increased risk for HIV-positive liver recipients was modest, supporting transplant among this higher risk population as a viable option. Patient selection should be informed by prior analyses identifying low BMI, dual liver-kidney transplant, and HCV co-infection as factors associated with poor outcome. However, the availability of interferon-free regimens and direct acting antivirals are anticipated to improve transplant outcomes among LTR with HIV-HCV co-infection.