Development of a dual-energy spectral computed tomography-based nomogram for the preoperative discrimination of histological grade in colorectal adenocarcinoma patients

Yuntai Cao; Guojin Zhang; Haihua Bao; Jialiang Ren; Zhan Wang; Jing Zhang; Zhiyong Zhao; Xiaohong Yan; Yanjun Chai; Junlin Zhou

doi:10.21037/jgo-20-368

Development of a dual-energy spectral computed tomography-based nomogram for the preoperative discrimination of histological grade in colorectal adenocarcinoma patients

J Gastrointest Oncol. 2021 Apr;12(2):544-555. doi: 10.21037/jgo-20-368.

Authors

Yuntai Cao^{1

2

3

4}, Guojin Zhang^{1

2}, Haihua Bao⁴, Jialiang Ren⁵, Zhan Wang⁶, Jing Zhang^{1

2}, Zhiyong Zhao^{1

2}, Xiaohong Yan⁷, Yanjun Chai³, Junlin Zhou^{2

3}

Affiliations

¹ Second Clinical School, Lanzhou University, Lanzhou, China.
² Key Laboratory of Medical Imaging of Gansu Province, Lanzhou, China.
³ Department of Radiology, Lanzhou University Second Hospital, Lanzhou, China.
⁴ Department of Radiology, Affiliated Hospital of Qinghai University, Xining, China.
⁵ Department of Pharmaceuticals Diagnosis, GE Healthcare, Beijing, China.
⁶ Department of Biomedical Engineering, Tsinghua University, Beijing, China.
⁷ Department of Critical Medicine, Affiliated Hospital of Qinghai University, Xining, China.

Abstract

Background: The usefulness of a dual-energy spectral computed tomography (DESCT)-based nomogram in discriminating between histological grades of colorectal adenocarcinoma (CRAC) is unclear. This study aimed to develop such a nomogram and assess its ability to preoperatively discriminate between histological grades in CRAC patients.

Methods: Primary tumors monochromatic CT value, iodine concentration (IC) value, and effective atomic number (Eff-Z) in the arterial (AP) and venous phases (VP) were retrospectively compared between patients with high-grade (n=65) and low-grade (n=108) CRAC who underwent preoperative abdominal DESCT. Univariate analysis was used to compare the DESCT parameters and clinical factors between these two patient groups. Statistically significant features in the univariate analysis were included in the multivariate logistic regression model to identify the indicators for building a nomogram that could discriminate between histological grades in CRAC patients. The clinical usefulness of the nomogram and its value for predicting overall survival were statistically evaluated.

Results: The logistic regression analysis showed that age, clinical T stage, clinical N stage, and IC values in AP and VP were significant independent predictors for high-grade CRAC. A quantitative nomogram developed based on these predictors showed excellent performance for discriminating between the histological grades, with an area under the curve (AUC) of 0.886 and excellent agreement in the calibration curve. The Kaplan-Meier curve for overall survival showed that our nomogram identified a significant difference between the high- and low-risk groups [hazard ratio (HR), 2.188; 95% CI, 1.072-4.465; P=0.027).

Conclusions: This study presents a nomogram that incorporates DESCT parameters and clinical factors and can potentially be used as a clinical tool for individual preoperative prediction of CRAC histological grade.

Keywords: Computed tomography (CT); colorectal adenocarcinoma (CRAC); histologic grade; nomogram.