Value of 18F-FDG PET/MR in evaluating the effect of neoadjuvant chemoradiotherapy in locally advanced rectal cancer / 中华核医学与分子影像杂志

ABSTRACT

Objective:To assess the value of 18F-fluorodeoxyglucose (FDG) PET/MR parameters and their changes in predicting and evaluating the curative effect in patients with locally advanced rectal cancer before and after neoadjuvant chemoradiotherapy (NCRT). Methods:From June 2017 to June 2020, 13 patients (9 males, 4 females; age (52.2±13.2) years) with locally advanced rectal cancer confirmed pathologically and underwent NCRT in Chinese PLA General Hospital were retrospectively enrolled. All patients performed the first PET/MR within one month before NCRT and the second PET/MR within one month before operation. PET/MR parameters including maximum standardized uptake value (SUV max), mean standardized uptake value (SUV mean), metabolic tumor volume (MTV) 2.5, total lesion glycolysis (TLG), minimum apparent diffusion coefficient (ADC min), and their changing percentage (Δ) before and after NCRT were collected. Patients were divided into pathologically complete remission (pCR) group and non-pCR group or response group and non-response group according to the postoperative pathological results as the gold standard. Mann-Whitney U test and logistic regression analysis were used for data analysis. The cut-off values of related parameters and their diagnostic efficiencies were determined by receiver operating characteristic (ROC) curve analysis. Results:Of 13 patients, 5 reached pCR and 8 had histological reaction (response). There were no significant differences in parameters (SUV max, SUV mean, MTV 2.5, TLG, ADC min) between different groups before treatment ( U values: 8.00-19.00, all P>0.05). There were significant differences in SUV max, SUV mean, MTV 2.5, TLG and ΔADC min between pCR group and non-pCR group after treatment ( U values: 0.00-6.00, all P<0.05), but only SUV max was correlated with pCR after treatment (odds ratio ( OR)=0.335, 95% CI: 0.123-0.917, P=0.033). The area under curve (AUC) was 0.95 and the cut-off value of SUV max was 3.055, with the sensitivity of 100%, the specificity of 80.0% and the accuracy of 92.3%. There were significant differences in SUV max, SUV mean, TLG, ADC min, ΔSUV max and ΔADC min between the response group and non-response group after treatment ( U values: 0.00-6.00, all P<0.05), but only ΔSUV max was correlated with the response results ( OR=2.022, 95% CI: 1.100-4.130, P=0.048). The AUC was 0.90 and the cut-off value of ΔSUV max was 69.0%, with the sensitivity of 87.5%, the specificity of 80.0% and the accuracy of 84.6%. Conclusions:PET/MR has high accuracy in evaluating NCRT for locally advanced rectal cancer. SUV max is an independent predictor of pCR after treatment, while ΔSUV max is an independent predictor of histological reaction (response).