Novel computed tomography-based nomograms for the pretherapeutic prediction of response to neoadjuvant chemotherapy with S-1 and oxaliplatin with or without the addition of docetaxel in patients with advanced gastric cancer

Ke-Shan Wu; Ke-Ying Li; Yan Gui; Ning-Pu Li; Hai-Ying Zhou; Xiao-Ming Zhang; Tian-Wu Chen

doi:10.21037/qims-24-748

Novel computed tomography-based nomograms for the pretherapeutic prediction of response to neoadjuvant chemotherapy with S-1 and oxaliplatin with or without the addition of docetaxel in patients with advanced gastric cancer

Quant Imaging Med Surg. 2024 Sep 1;14(9):6711-6723. doi: 10.21037/qims-24-748. Epub 2024 Aug 19.

Authors

Ke-Shan Wu^#^{1

2}, Ke-Ying Li^#¹, Yan Gui^#³, Ning-Pu Li³, Hai-Ying Zhou¹, Xiao-Ming Zhang¹, Tian-Wu Chen⁴

Affiliations

¹ Medical Imaging Key Laboratory of Sichuan Province, and Department of Radiology, Affiliated Hospital of North Sichuan Medical College, Nanchong, China.
² Department of Radiology, Jinshan Hospital Affiliated of Fudan University, Shanghai, China.
³ Department of Oncology, Affiliated Hospital of North Sichuan Medical College, Nanchong, China.
⁴ Department of Radiology, The Second Affiliated Hospital of Chongqing Medical University, Chongqing, China.

^# Contributed equally.

Abstract

Background: Selecting the appropriate preoperative neoadjuvant chemotherapy (NACT) regimen for patients with advanced gastric cancer (GC) is critical to effective treatment. The aim of this study was to develop nomograms based on pretherapeutic computed tomography (CT) features to predict response to NACT with S-1 and oxaliplatin (SOX) or that with docetaxel and SOX (DOS) in patients with advanced GC.

Methods: This study enrolled 311 consecutive patients with confirmed advanced GC undergoing contrast-enhanced CT before and after the three cycles of NACT with DOS (n=152) or SOX (n=159), who were randomized into a training cohort (TC) (NACT with DOS: n=111; NACT with SOX: n=120) and validation cohort (VC) (NACT with DOS: n=41; NACT with SOX: n=39). The objective response rate (ORR) was used to evaluate the response to NACT. In the TC, ORR was compared between the DOS and SOX regimens, and independent predictors including CT features and tumor differentiation were determined by univariate and binary logistic regression analyses. Individual nomograms were constructed for the SOX and DOS regimens in the TC, and the predictive accuracy was validated in the VC.

Results: After NACT, the percentage of ORR was higher in patients receiving DOS than in those receiving SOX in TC (P value <0.05). The independent predictors after DOS and SOX were pretherapeutic cT stage [odds ratio (OR) =7.364; OR =8.848], cN stage (OR =1.027; OR =1.345), degree of differentiation (OR =7.127; OR =7.835), and gross tumor volume (OR =8.960; OR =8.161) (all P values <0.05). The concordance indexes of the individual nomograms developed using these predictors were 0.940 and 0.932 after DOS or SOX in the TC, respectively, which was validated by calibration plots with a slope close to 45° in the TC and VC.

Conclusions: Despite there being a superior response to DOS compared with SOX, nomograms for predicting response to both NACT regimens were similar, with each demonstrating good predictive performance.

Keywords: Gastric cancer (GC); computed tomography (CT); neoadjuvant chemotherapy (NACT); nomogram; prediction.