RESUMEN
Persistent high-risk human papillomavirus (HPV) infection is strongly associated with the development of high-grade cervical intraepithelial neoplasia (CIN) or cancer. Not all persistent infections lead to cancer. Viral load measured at a single time-point is a poor predictor of the natural history of HPV infections. However the profile of viral load evolution over time could distinguish nonprogressive from progressive (carcinogenic) infections. A retrospective natural history study was set up using a Belgian laboratory database including more than 800,000 liquid cytology specimens. All samples were submitted to qPCR identifying E6/E7 genes of 18 HPV types. Viral load changes over time were assessed by the linear regression slope. Database search identified 261 untreated women with persistent type-specific HPV DNA detected (270 infections) in at least three of the last smears for a average period of 3.2 years. Using the coefficient of determination (R²) infections could be subdivided in a latency group (n = 143; R² < 0.85) and a regressing group (n = 127; R² ≥ 0.85). In (≥ 3) serial viral load measurements, serial transient infections with latency is characterized by a nonlinear limited difference in decrease or increase of type-specific viral load (R² < 0.85 and slopes between 2 measurements 0.0010 and -0.0010 HPV copies/cell per day) over a longer period of time (1553 days), whereas regression of a clonal cell population is characterized by a linear (R² ≥ 0.85) decrease (-0.0033 HPV copies/cell per day) over a shorter period of time (708 days; P < 0.001). Using serial HPV type-specific viral load measurements we could for the first time identify regressing CIN2 and CIN3 lesions. Evolution of the viral load is an objective measurable indicator of the natural history of HPV infections and could be used for future triage in HPV-based cervical screening programs.