A real-world study based on targeted next-generation sequencing: association of cytomegalovirus detection with co-detection of Pneumocystis jirovecii and Aspergillus fumigatus in respiratory patients

Background Lower respiratory tract infections (LRTIs) are among the most common and lethal infectious diseases globally, with approximately 50% lacking a definitive etiological diagnosis. Targeted next-generation sequencing (tNGS) shows promise in pulmonary infection diagnosis. Human herpesviruses, including Epstein–Barr virus (EBV) and cytomegalovirus (CMV), can reactivate in immunocompromised individuals and are closely linked to the co-detection of opportunistic pathogens. Objective Based on 1156 samples, we evaluated the detection rates of EBV and CMV, explored the co-detection associations between CMV positivity and Pneumocystis jirovecii and Aspergillus fumigatus , and examined the modifying effect of age. Methods A total of 1156 patients who underwent tNGS at the Respiratory Department of Zhangzhou Hospital (December 2024–April 2026) were enrolled (38 duplicate samples were excluded). Standard Pearson’s chi-square tests (without continuity correction; Fisher’s exact test when expected frequency < 5) were used. Odds ratios (OR) and 95% confidence intervals (CI) were calculated, and the Benjamini–Hochberg false discovery rate (FDR) correction was applied. Multivariable logistic regression (adjusted for age and sex) and interaction analyses were conducted to evaluate independent associations and age-modifying effects. Results EBV (42.1% vs. 33.2%, P = 0.0016) and CMV (29.3% vs. 16.3%, P < 0.0001) detection rates were significantly higher in the elderly group. In the overall population, P. jirovecii detection was significantly increased in CMV-positive patients (OR = 1.72, 95% CI 1.12–2.59, P = 0.011). The strongest CMV– A. fumigatus association was observed in the non-elderly group (OR = 2.46, 95% CI 1.32–4.42, P = 0.003). Interaction analysis showed that age significantly modified the CMV– A. fumigatus association (interaction P = 0.027), whereas the CMV– P. jirovecii association remained robust across age groups (interaction P = 0.194). CMV/EBV showed no significant association with non-tuberculous mycobacteria (NTM) (all adjusted P > 0.05). Conclusion CMV positivity was significantly associated with the co-detection of P. jirovecii in the overall population and A. fumigatus in non-elderly patients, suggesting that CMV detection may serve as a warning signal prompting opportunistic pathogen screening. Age is a significant effect modifier of the CMV– A. fumigatus association. Of note, the present findings reflect microbiological co-detection rather than clinical co-infection.