Association between cytomegalovirus infection and cardiac allograft vasculopathy in heart transplant recipients: a single-center retrospective cohort study

Background/Introduction Cytomegalovirus (CMV) infection has been implicated in the pathogenesis of cardiac allograft vasculopathy (CAV) through endothelial injury, smooth muscle cell proliferation, and immune-mediated vascular inflammation. CAV remains the leading cause of late graft loss after heart transplantation. However, the independent contribution of CMV viremia to CAV risk in the contemporary era of modern prophylaxis remains uncertain. Purpose To determine whether CMV viremia >1,000 IU/mL independently increases the risk of CAV development in adult heart transplant recipients using competing-risk methodology. Methods A retrospective cohort study of 180 adult heart transplant recipients (age ≥14 years) transplanted between 2010 and 2021, surviving ≥12 months with documented CAV screening. CMV infection was defined as viremia >1,000 IU/mL. CAV was diagnosed and graded per ISHLT criteria by coronary angiography. The primary analysis used Fine-Gray competing-risk regression with death without CAV as the competing event. Secondary analyses included Kaplan-Meier survival analysis, time-dependent covariate models, landmark analyses at 6, 12, and 24 months, and marginal structural models with inverse probability weighting. Results Over a median follow-up of 7.0 years (IQR 4.5–9.5), 49 patients (27.2%) developed CAV and 51 (28.3%) died without CAV. CMV viremia occurred in 86 patients (47.8%), with 94.2% of infections occurring within the first 6 months. In multivariable competing-risk regression adjusted for age, sex, ischaemic aetiology, rejection, donor age, and transplant era, CMV viremia was not associated with CAV (adjusted sub-distribution hazard ratio [SHR] 0.78, 95% CI 0.46–1.33, p=0.359). Transplant era 2016–2021 was the only independent predictor (SHR 2.27, 95% CI 1.31–3.95, p=0.004). No dose-response relationship was observed with peak viral load (SHR 0.74, p=0.256) or episode frequency (trend p=0.148). Landmark analysis at 12 months showed a paradoxical protective CMV effect (SHR 0.59, 95% CI 0.36–0.99, p=0.044). Acute rejection predicted CAV severity (OR 3.05, 95% CI 1.24–7.50, p=0.015) but not occurrence. CMV did not predict mortality (HR 1.03, p=0.926) or graft failure (HR 0.99, p=0.973). Results were robust across alternative CMV thresholds, CAV definitions, and after Bonferroni correction. Conclusion(s) CMV viremia was not independently associated with increased CAV risk in this contemporary cohort. Transplant era was the dominant predictor, likely reflecting enhanced surveillance. These findings challenge the paradigm that CMV is a modifiable CAV risk factor and suggest that CAV prevention strategies should prioritise rejection management over CMV-directed interventions.