Immunosuppression is a known independent predictor of chronic kidney disease (CKD) in HIV-positive persons, but the nature of the association between different measures of immunosuppression and CKD is unknown.
D:A:D study participants without CKD and with >2 Cockcroft Gault (estimated glomerular filtration rate) eGFR measurements after 1/1/2004 (baseline) were followed until the earliest of CKD (eGFR<60, confirmed >3 months apart), last eGFR plus 6 months or 1/2/2014. Measures of immunosuppression included baseline, current and nadir CD4, 6-months’ time-lagged CD4, % of follow-up time (%FU) with CD4<200, time-averaged AUC for CD4 and CD4 recovery (baseline CD4<200 followed by current CD4>200). Poisson regression models were used to determine the relationship between CKD and each measure of immunosuppression accounting for relevant confounders, and tested for interactions with the D:A:D CKD risk score, demographics, HCV and HIV-related factors. Akaike Information Criteria (AIC) was used to indicate which measures were better CKD predictors.
Of the 33,144 persons included in analyses 1,588 developed CKD (incidence rate (IR) 7.2 [95%CI 6.8-7.5]/1000 PYFU) during a median 7.2 years FU (IQR 5.0-8.9). Those included were predominately white (47.6%), male (74.0%) with a baseline median age of 41 years (IQR 35-47) and median CD4 of 440 (292-626). The crude CKD IR varied for different measures of immunodeficiency (Fig). Univariately, all measures of immunosuppression were significantly associated with CKD, most strongly for nadir CD4 (>500 vs. <50, IR 0.39 [0.31–0.48]) and %FU CD4<200 (>25% vs. 0%, IR 1.86 [1.62-2.13]). Multivariately, the strongest CKD predictor was %FU CD4<200 (>25% vs. 0%, 1.29 [1.11-1.30]). There was a significant (p<0.0001) interaction between %FU CD4<200 and the D:A:D CKD risk score; those at lowest estimated CKD risk had a significantly higher CKD IR (>25% vs. 0%, 3.57 [2.23-5.70]) compared to those at highest CKD risk (>25% vs. 0%, 1.24 [1.05-1.46]). There was no significant interaction between measures of immunosuppression and ethnicity, age, HIV-RNA, ART status or use of nephrotoxic antiretrovirals including tenofovir, indinavir, atazanavir/r and lopinavir/r.
The strongest association between CKD and immunosuppression was observed for the relative duration of severe immunosuppression, which was of greatest importance in persons at low estimated CKD risk. These new data support aggressive ART to maintain/restore immune function.