Background:
People with HIV (PWH) and CMV have higher cardiometabolic disease risk than those without HIV despite antiretroviral therapy (ART), but mechanisms are unclear.
Methods:
We performed a randomized trial of letermovir (CMV terminase inhibitor, 480 mg daily) in ART-suppressed CMV-seropositive PWH to assess whether letermovir therapy for 48 weeks reduced plasma sTNFR2 levels (primary endpoint) and other inflammation and cardiometabolic indices vs. no anti-CMV treatment. A planned futility analysis was performed after 40 (of planned 180) participants reached week 8. Continuous changes in plasma biomarkers (ELISA and Olink Inflammation and Cardiometabolic Explore panels) and binary CMV shedding in genital secretion and saliva were compared between arms with linear mixed or logistic models, adjusting proteomic analyses for False Discovery Rate (Benjamini-Hochberg).
Results:
Of 42 participants enrolled, 40 contributed to the week 8 per-protocol analysis, stratified by CD4 count (45% <350 cells/mm3) and sex at birth (29% female). Adverse events were similar between arms. Letermovir suppressed CMV DNA at all on-treatment timepoints. Unexpectedly, plasma sTNFR2 levels increased in the letermovir arm at week 8 (P<0.001, between-arms P=0.059) and the trial was stopped for futility. We hypothesized that suppression of CMV's immunoregulatory viral IL-10 may have caused the increase in plasma sTNFR2. Consistent with this hypothesis, plasma IL-10RA – a marker of IL-10 receptor signaling – declined 28% more in the letermovir than control arms (P=0.009), while human IL-10 levels were unchanged. While sTNFR2 increased, 30 plasma inflammation (e.g., IL-6R, IL-1b) and cardiometabolic proteins (e.g., VEGF-D, VCAM-1, Neuropilin-1, FABP-4) declined more in the letermovir than control arms (FDR-adjusted P<0.05 [red] and P<0.10 [blue], Figure). Letermovir caused a greater decline in 3 of 5 proteins recently causally linked to vascular events in treated HIV by Mendelian Randomization (AXL, GAS6, and IL-6R; nominal P≤0.02).
Conclusions:
Suppression of asymptomatic CMV in PWH on ART increased plasma sTNFR2, but decreased markers of IL-10 receptor signaling and several inflammation and cardiometabolic proteins, some of which are causally linked to cardiovascular risk. While the clinical significance remains unclear, this is the first study to show that a specific inhibitor of CMV – without direct activity against other herpesviruses – is safe and has broad impact on immunologic and cardiometabolic biomarkers in PWH.
