RESUMO
OBJECTIVES: To analyse the role of DW-MRI in early prediction of pathologically-assessed residual disease in locally-advanced cervical cancer (LACC) treated with neoadjuvant chemoradiotherapy followed by radical surgery. METHODS: Between October 2010-June 2014, 108 women with histologically-proven cervical cancer were screened; 88 were included in this study. Tumour volume (TV) and ADCmean were measured before (baseline-MRI) and after 2 weeks of chemoradiotherapy (early-MRI). According to histopathology, treatment response was classified as complete (CR) or partial (PR). Comparisons were made with Mann-Whitney, Wilcoxon and χ2 tests. ROC curves were generated for statistically significant parameters on univariate analysis. RESULTS: CR and PR were documented in 40 and 48 patients. At baseline-MRI, TV did not differ between groups. At early-MRI, TV was higher in PR than in CR (p=0.001). ΔTV reduction after treatment was lower in PR than in CR (63.6% vs. 81.1%; p=0.001). At baseline-MRI and early-MRI, ADCmean did not differ between PR and CR. ROC curve showed best cut-off for predicting pathological PR was ΔTV reduction of 73% with sensitivity, specificity, accuracy, NPV, PPV of 73%, 72.5%, 72.7%, 76%, 69%. CONCLUSIONS: TV evaluated before and early after treatment could predict pathological response in LACC. ADCmean did not correlate with treatment outcome. KEY POINTS: ⢠Early-MRI tumour volume assessment could predict pathological response to nCRT in LACC. ⢠Best cut-off for predicting pathological PR was ΔTV reduction of 73 %. ⢠Early-MRI ADC mean measurements did not correlate with treatment outcome.