Tumor size measured by multidetector CT in resectable colon cancer: Correlation with regional lymph node metastasis

Anna Mou,Hang Li,Xiaoli Chen,Fanghua Fan,Hong Pu
DOI: https://doi.org/10.21203/rs.2.24612/v1
2020-01-01
Abstract:Abstract Background : The aim of our study was to determine whether tumor size measured by multidetector computed tomography (MDCT) could be used to evaluate lymph node metastasis (LNM) in patients with resectable colon cancer. Methods: This retrospective study consisted of 106 consecutive patients with colon cancer who underwent radical surgery within 1 week after contrast enhanced-CT scan. Tumor size including tumor length (Tlen), tumor maximum diameter (Tdia), t umor maximum cross-sectional area (Tare) and tumor volume (Tvol) were measured on contrast enhanced-CT images and correlated with pathologic LNM and N stages using univariate analysis, logistic regression analysis and receiver operating characteristic (ROC) curve analysis. Results: The inter - (intraclass correlation coefficient [ICC]=0.94, 0.81, 0.97, 0.99) and intraobserver (ICC=0.95, 0.93, 0.91 and 0.99) reproducibility of Tlen, Tdia, Tare and Tvol parameters measurement is excellent. Univariate analysis showed Tlen, Tdia, Tare, and Tvol could predict LNM (all P <0.05), whereas Tvol was an independent risk factor for LNM (odds ratio =1.09; 95% confidence interval, 1.02-1.17; P =0.017) by logistic regression analysis. Tlen, Tdia, Tare and Tvol could distinguish between N0 and N1, N0 and N2, N0 and N1-2, and N0-1 and N2 disease (all P < 0.05). The area under the ROC (AUC) was higher for Tvol than for Tlen, Tdia and Tare in identifying LNM (AUC =0.83, 0.82, 0.69, 0.79, respectively) and distinguishing N0 from N1 (AUC =0.79, 0.78, 0.63, 0.74, respectively), N0 from N2 (AUC =0.92, 0.89, 0.80, 0.89, respectively), and N0-1 from N2 (AUC =0.84, 0.79, 0.76, 0.83, respectively). Conclusion: Tlen , Tdia, Tare and Tvol measured with MDCT correlated with regional LNM in resectable colon cancer. In particular, Tvol showed the most potential for noninvasive preoperative evaluation of regional LNM.
What problem does this paper attempt to address?