Purpose: To evaluate whether human papillomavirus (HPV) viral load measured in cervical smear and HPV type 18 are associated with radiotherapy outcomes in uterine cervical cancer. PATIENTS AND METHODS HPV DNA: was semiquantitatively measured in the cervical smears of 169 radiotherapy patients. HPV viral load was classified as low or high according to median HPV DNA titer and examined for its prognostic value. The multivariable Cox proportional hazards model was used to adjust for covariates. A relapse-predicting model was constructed to classify three risk groups for disease-free survival (DFS), which were used for internal validation.
Results: Patients with lower HPV viral load showed worse DFS in univariate analysis. HPV type 18, younger patient age, stage group, nodal status, histologic grade, and histologic type were other prognostic factors for poor DFS. Among these factors, all except stage group were associated with HPV viral load. Multivariate analysis showed the strong influence of HPV viral load for poor DFS. The prognostic model developed using our outcome data performed well in predicting the risk of relapse.
Conclusion: Our data suggest that HPV viral load is a strong independent prognostic factor for DFS. HPV type 18 showed a significant relationship with poor radiotherapy outcome in univariate analysis, but not in multivariate analysis.