Nomograms combining clinical factors and apparent diffusion coefficient to predict downstaging and progression-free survival after concurrent chemoradiotherapy in patients with cervical cancer.

Abstract

Background: Concurrent chemoradiotherapy (CCRT) is used as the primary treatment modality for currently limited cervical cancer and lacks non-invasive quantitative parameters to assess clinical outcomes of treatment for cervical cancer treatment.

Purpose: To develop nomograms based on clinical prognostic factors and apparent diffusion coefficient (ADC) in predicting downstaging and progression-free survival (PFS) after CCRT for cervical cancer.

Material and methods: X-tile was used to calculate the optimal threshold for ΔADC_mean(%) for prognostic stratification. Kaplan-Meier curves were used to calculate the difference in PFS between high- and low-risk groups. Univariate and multivariate Cox proportional risk regression models were used to identify clinical and radiological risk factors for prognosis and construct a prognostic nomogram model.

Results: ΔADC_mean(%) was significantly correlated with tumor downstaging; the area under the receiver operating characteristic curve (AUC) was 0.868. X-tile showed that the optimal threshold for ΔADC_mean(%) to diagnose prognosis was 40.8. Kaplan-Meier curves showed that the low-risk population in the training group had significantly longer PFS within 3 years (P < 0.001). Multivariate Cox regression showed that ΔADC (%) is independent risk factor for PFS. The C-index of ΔADC(%) predicting 3-year PFS in the training set is 0.761 and the C-index of the nomogram model is 0.862.

Conclusion: ΔADC_mean(%) is a non-invasive biomarker for predicting tumor downstaging in cervical cancer after CCRT. The nomograms based on ΔADC_mean(%) predict PFS of patients with cervical cancer with moderate accuracy.