Purpose: The standard chemoradiation therapy currently used for locally advanced cervical cancer (LACC) patients does not reflect the biological heterogeneity of this disease, and there is an increasing need for the development of biomarkers that can help guide the individualized treatment regimens. The purpose of this study was to investigate the prognostic value of the integration pattern of human papillomavirus (HPV) in LACC patients.
Methods and materials: The HPV integration pattern was determined by in situ hybridization and polymerase chain reaction, and the tumors were classified as the episomal pattern (group A), as the single-copy integrated or multicopy tandem repetition-integrated pattern (group B), or as undetectable HPV (group C). Ninety-eight LACC patients were included in a development dataset and 106 independent patients in a validation dataset. The multivariate Cox model was used to examine the effect of the HPV integration pattern on disease-free survival (DFS). The model was validated internally by the leave-one-out cross-validation method and externally by an independent dataset.
Results: After adjustment for significant prognostic factors (stage, histologic grade, histologic type, and tumor size), the HPV integration pattern was significantly associated with DFS in the development (P=.032) and validation (P=.023) datasets. Survival was worst in group C and best in group A. The multivariate model with HPV integration pattern as an explanatory variable showed good discrimination ability and could separate patients with different risk profiles.
Conclusions: This study identified the HPV integration pattern, as determined by in situ hybridization and polymerase chain reaction, as a strong prognostic biomarker for DFS in LACC patients treated by chemoradiation therapy. This finding may open the possibility of personalized treatment of these patients.
Copyright © 2017. Published by Elsevier Inc.