POLYGENIC RISK STRATIFICATION OF CYTOMEGALOVIRUS INFECTION USING A FIVE-LOCUS SCORING MODEL: DIAGNOSTIC ACCURACY AND CLINICAL APPLICABILITY

Abstract

Cytomegalovirus infection (CMVI) remains one of the most prevalent infectious diseases worldwide, with seroprevalence reaching 85–95% in developing regions including Central Asia (1, 8). The clinical spectrum ranges from asymptomatic latency to severe multi-organ involvement in neonates, pregnant women, and immunocompromised patients (2, 6). Predicting which patients will progress from latent to manifest persistent infection remains a key clinical challenge, as traditional laboratory markers such as viral load and C-reactive protein reflect the current disease state rather than the inherent susceptibility of the host (3, 6). Single nucleotide polymorphisms (SNPs) in genes regulating immune response (TNF-α, IL-10, TLR4), oxidative stress defense (SOD2), and endothelial function (eNOS) have been individually associated with susceptibility to CMVI; however, no single SNP has demonstrated sufficient discriminative power for clinical application, with individual AUC values typically below 0.70 (4, 5, 7). The polygenic nature of susceptibility to chronic viral infections suggests that a composite approach integrating multiple genetic variants into a single quantitative measure may yield superior predictive accuracy.