PRospective Imaging of CErvical cancer and neoadjuvant treatment (PRICE) study: role of ultrasound to predict partial response in locally advanced cervical cancer patients undergoing chemoradiation and radical surgery.

Abstract

Chemoradiation-based neoadjuvant treatment followed by radical surgery is an alternative therapeutic strategy for locally advanced cervical cancer (LACC), but ultrasound variables used to predict partial response to neoadjuvant treatment are not well defined. Our goal was to analyze prospectively the potential role of transvaginal ultrasound in early prediction of partial pathological response, assessed in terms of residual disease at histology, in a large, single-institution series of LACC patients triaged to neoadjuvant treatment followed by radical surgery. Between October 2010 and June 2014, we screened 108 women with histologically documented LACC Stage IB2-IVA, of whom 88 were included in the final analysis. Tumor volume, three-dimensional (3D) power Doppler indices and contrast parameters were obtained before (baseline examination) and after 2 weeks of treatment. The pathological response was defined as complete (absence of any residual tumor after treatment) or partial (microscopic and/or macroscopic residual tumor at pathological examination). Complete-response and partial-response groups were compared and receiver-operating characteristics (ROC) curves were generated for ultrasound variables that were statistically significant on univariate analysis to evaluate their diagnostic ability to predict partial pathological response. There was a complete pathological response to neoadjuvant therapy in 40 (45.5%) patients and a partial response in 48 (54.5%). At baseline examination, tumor volume did not differ between the two groups. However, after 2 weeks of neoadjuvant treatment, the tumor volume was significantly greater in patients with partial response than it was in those with complete response (P = 0.019). Among the 3D vascular indices, the vascularization index (VI) was significantly lower in the partial-response compared with the complete-response group, both before and after 2 weeks of treatment (P = 0.037 and P = 0.024, respectively). At baseline examination in the contrast analysis, women with partial response had lower tumor peak enhancement (PE) as well as lower tumor wash-in rate (WiR) and longer tumor rise time (RT) compared with complete responders (P = 0.006, P = 0.003, P = 0.038, respectively). There was no difference in terms of contrast parameters after 2 weeks of treatment. ROC-curve analysis of baseline parameters showed that the best cut-offs for predicting partial pathological response were 41.5% for VI (sensitivity, 63.6%; specificity, 66.7%); 16123.5 auxiliary units for tumor PE (sensitivity, 47.9%; specificity, 84.2%); 7.8 s for tumor RT (sensitivity, 68.8%; specificity, 57.9%); and 4902 for tumor WiR (sensitivity, 77.1%; specificity, 60.5%). ROC curves of parameters after 2 weeks of treatment showed that the best cut-off for predicting partial pathological response was 18.1 cm3 for tumor volume (sensitivity, 70.8%; specificity 60.0%) and 39.5% for VI (sensitivity; 62.5%; specificity, 73.5%). Ultrasound and contrast parameters differ between LACC patients with complete response and those with partial response before and after 2 weeks of neoadjuvant treatment. However, neither ultrasound parameters before treatment nor those after 2 weeks of treatment had cut-off values with acceptable sensitivity and specificity for predicting partial pathological response to neoadjuvant therapy. Copyright © 2017 ISUOG. Published by John Wiley & Sons Ltd.