Background: It is unclear whether lack of immunological response despite viral suppression and relatively preserved CD4 T-cell count is associated with increased risk of AIDS or severe non-AIDS events.
Methods: Patients initiating first combination antiretroviral therapy (cART) were studied from first viral load 80 copies/ml or less up to AIDS, serious non-AIDS events (malignancies, severe infections, acute kidney injury, cardiovascular events, liver decompensation) or death. Follow-up was right censored if viral load was more than 500. Immunological nonresponse (INR) was defined as current CD4 cell count less than 120% pre-cART. A Poisson regression analysis was used to investigate the association between INR and the outcome.
Results: Three thousand, three hundred and seventy-eight patients were followed for a median of 32 months (interquartile range: 15-67). Two hundred and twenty-two events (32 deaths, 39 AIDS-defining events, 48 malignancies, 32 severe infections, 47 acute kidney injuries, 12 cardiovascular events, 12 other nonfatal events) were observed. The rate of clinical events among INR and immunological responders was 4.41 [95% confidence interval (CI) 3.38-5.74] and 1.84 (95% CI 1.58-2.15) per 100 person years of follow-up, respectively, accounting for a crude rate ratio of 2.39 (95% CI 1.77-3.25; P < 0.001). INR remained an independent predictor of clinical progression after adjusting for baseline characteristics, including pre-cART CD4 cell count (adjusted rate ratio 2.93; 95% CI 2.06-4.16, P < 0.001) or current CD4 cell count (adjusted rate ratio 1.94; 95% CI 1.39-2.72, P < 0.001). The association did not vary by pre-cART CD4 cell counts (P for interaction = 0.93)
Conclusion: INR are at higher risk of severe clinical events than responders. The association was consistent across different CD4 cell counts at cART initiation and was only partially explained by current CD4 cell count. INR could be a marker of immune system malfunctioning, not completely captured by absolute CD4 cell count.