Concomitant cytomegalovirus DNAemia is associated with sustained TNF-α elevation in HIV/AIDS: a multivariable and longitudinal analysis of immune restoration.

BACKGROUND

Despite successful viral suppression with antiretroviral therapy (ART), incomplete immune reconstitution and chronic inflammation persist in people with HIV, contributing to an increased burden of non-AIDS-defining comorbidities. Emerging evidence points to cytomegalovirus (CMV) co-infection as a critical driver of this residual inflammation. Consequently, this study aims to elucidate the specific role of CMV in sustaining inflammatory pathways in the context of HIV.

METHODS

This prospective observational study included 200 participants divided into three groups: 50 with HIV/CMV co‑infection, 100 with HIV mono‑infection, and 50 healthy controls. Baseline immunophenotyping and cytokine profiling were assessed using flow cytometry and enzyme-linked immunosorbent assays. To account for baseline HIV disease severity and demographic imbalances, cross-sectional comparisons were evaluated using multivariable linear regression adjusted for age, sex, and baseline absolute CD4⁺ T-cell counts, supplemented by CD4-stratified analyses (< 100 and > 200 cells/µL). To evaluate the longitudinal impact of CMV viral load dynamics, co‑infected participants were classified after 6 months follow‑up into "CMV responders" (≥ 68% reduction in CMV DNA load) and "CMV non‑responders" (< 68% reduction or increase).

RESULTS

The overall cohort was predominantly CMV IgG positive (97.5%). At baseline, the unadjusted expansion of CD8⁺ T-cells observed in HIV/CMV co-infected patients lost statistical significance after multivariable adjustment for baseline CD4⁺ counts (adjusted p = 0.150), indicating this was primarily related to advanced HIV immunosuppression. In contrast, CMV DNAemia remained independently associated with significantly elevated systemic TNF-α levels (adjusted p = 0.043). Stratified analysis revealed that this independent TNF-α elevation was particularly pronounced in patients with relatively preserved immune function (CD4 > 200 cells/µL; adjusted p = 0.042), whereas these differences were obscured in severe immunosuppression (CD4 < 100 cells/µL). Longitudinal analysis revealed that effective reduction of CMV viral load was associated with significant HIV viral suppression in responders (p = 0.0039), whereas no significant reduction was observed in non-responders (p = 0.3894), and seemingly facilitated the resolution of systemic TNF-α levels. In contrast, significant recovery of CD4⁺ T-cell absolute counts occurred in both groups (p < 0.01) and the rate of recovery did not differ significantly between CMV responders and non-responders.

CONCLUSIONS

These findings suggest that CMV DNAemia is independently associated with persistent TNF-α elevation in PLWH, particularly in those with preserved CD4 + counts. Rather than directly causing broad T-cell exhaustion, CMV appears to act as a specific driver of the TNF-α inflammatory pathway. Addressing adherence to anti-CMV interventions alongside ART may offer a potential strategy for reducing long-term inflammatory risks.