Photon counting computed tomography in rectal cancer : associations between iodine concentration, histopathology and treatment response : a pilot study

Abstract

Rationale and Objectives: Common computed tomography (CT) investigation plays a limited role in characterizing and assessing the response of rectal cancer (RC) to neoadjuvant radiochemotherapy (NARC). Photon counting computed tomography (PCCT) improves the imaging quality and can provide multiparametric spectral image information including iodine concentration (IC). Our purpose was to analyze associations between IC and histopathology in RC and to evaluate the role of IC in response prediction to NARC. Materials and Methods: Overall, 41 patients were included into the study, 14 women and 27 men, mean age, 65.5 years. PCCT in a portal venous phase of the abdomen was performed. In every case, a polygonal region of interest (ROI) was manually drawn on iodine maps. Normalized IC (NIC) was also calculated. Tumor stage, grade, lymphovascular invasion, circumferential resection margin, and tumor markers were analyzed. Tumor regression grade (absence/presence of tumor cells) after NARC was analyzed. NIC values in groups were compared to Mann–Whitney-U tests. Sensitivity, specificity, and area under the curve values were calculated. Intraclass correlation coefficient (ICC) was calculated. Results: ICC was 0.93, 95%CI = (0.88; 0.96). Tumors with lymphovascular invasion showed higher NIC values in comparison to those without (p = 0.04). Tumors with response grade 2–4 showed higher pretreatment NIC values in comparison to lesions with response grade 0–1 (p = 0.01). A NIC value of 0.36 and higher can predict response grade 2–4 (sensitivity, 73.9%; specificity, 91.7%; area under the curve, 0.85). Conclusion: NIC values showed an excellent interreader agreement in RC. NIC can predict treatment response to NARC.