Evaluating the response of gastric carcinomas to neoadjuvant chemotherapy using iodine concentration on spectral CT: a comparison with pathological regression

Aim: To investigate the potential of iodine concentration (IC) determined using virtual monochromatic spectral computed tomography (CT) to predict the response of gastric carcinomas to preoperative neoadjuvant chemotherapy (NC).

Materials and methods: A total of 20 patients were enrolled who underwent two spectral CT examinations (1 week before and two cycles after NC). The percentage change in tumour thickness (%ΔCWT) and in IC on the arterial phase (%ΔIC-a) and venous phase (%ΔIC-v) after NC were calculated and compared for different histopathological regression grades and response groups. The diagnostic efficacies to discriminate good response (GR) and poor response (PR) of the above three parameters were evaluated using receiver operating characteristic (ROC) curves.

Results: The decrease rate of %ΔIC-a for the GR group was higher than that for the PR group (-0.59 [-0.76, -0.20] versus -0.11 [-0.75, 0.92], p=0.012). There was no significant difference in the %ΔIC-v and %ΔCWT values between the GR and PR groups (p=0.076 and p=0.779, respectively). The areas under the ROC curve (AUC) values were 0.857, 0.762, and 0.542 for %ΔIC-a, %ΔIC-v, and %ΔCWT, respectively, in the response prediction. The cut-off value for identifying PR was a decrease rate of <52.9% for %ΔIC-a, and the sensitivity and specificity values were 0.857 and 0.833.

Conclusion: Changes in the IC for gastric carcinomas following NC were detected using spectral CT and correlated with histopathological regression. The prediction efficacy for IC was better than that for tumour thickness, with IC on the arterial phase being a better predictor than IC on the venous phase.