A nomogram for predicting 10-year cancer specific survival in patients with pathological T3N0M0 rectal cancer

Shuang Liu; Shanfei Yang; Haina Yu; Huilong Luo; Gong Chen; Yuanhong Gao; Rui Sun; Weiwei Xiao

doi:10.3389/fmed.2022.977652

A nomogram for predicting 10-year cancer specific survival in patients with pathological T3N0M0 rectal cancer

Front Med (Lausanne). 2022 Aug 22:9:977652. doi: 10.3389/fmed.2022.977652. eCollection 2022.

Authors

Shuang Liu^{1

2

3}, Shanfei Yang^{1

2

3}, Haina Yu^{1

2

3}, Huilong Luo^{1

2

3}, Gong Chen^{2

3

4}, Yuanhong Gao^{1

2

3}, Rui Sun^{1

2

3}, Weiwei Xiao^{1

2

3}

Affiliations

¹ Department of Radiation Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China.
² State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Guangzhou, China.
³ Collaborative Innovation Center for Cancer Medicine, Sun Yat-sen University Cancer Center, Guangzhou, China.
⁴ Department of Colorectal Surgery, Sun Yat-sen University Cancer Center, Guangzhou, China.

Abstract

Background: The pathological T3N0M0 (pT3N0M0) rectal cancer is the earliest stage and has the best prognosis in the locally advanced rectal cancer, but the optimal treatment remains controversial. A reliable prognostic model is needed to discriminate the high-risk patients from the low-risk patients, and optimize adjuvant chemotherapy (ACT) treatment decisions by predicting the likelihood of ACT benefit for the target population.

Patients and methods: We gathered and analyzed 276 patients in Sun Yat-sen University Cancer Center from March 2005 to December 2011. All patients underwent total mesorectal excision (TME), without preoperative therapy, and were pathologically proven pT3N0M0 rectal cancer with negative circumferential resection margin (CRM). LASSO regression model was used for variable selection and risk factor prediction. Multivariable cox regression was used to develop the predicting model. Optimum cut-off values were determined using X-Tile plot analysis. The 10-fold cross-validation was adopted to validate the model. The performance of the nomogram was evaluated with its calibration, discrimination and clinical usefulness.

Results: A total of 188 patients (68.1%) had ACT and no patients had adjuvant radiotherapy. Age, monocyte percentage, carbohydrate antigen 19-9, lymph node dissection numbers and perineural invasion (PNI) were identified as significantly associated variables that could be combined for an accurate prediction risk of Cancer Specific Survival (CSS) for pT3N0M0 patients. The model adjusted for CSS showed good discrimination with a C-index of 0.723 (95% CI: 0.652-0.794). The calibration curves showed that the nomogram adjusted for CSS was able to predict 3-, 5-, and 10-year CSS accurately. The corresponding predicted probability was used to stratify high and low-risk patients (10-year CSS: 69.1% vs. 90.8%, HR = 3.815, 95%CI: 2.102-6.924, P < 0.0001). ACT improved overall survival (OS) in the low-risk patients (10-year OS: 91.9% vs. 83.3%, HR = 0.338, 95% CI: 0.135-0.848, P < 0.0001), while it did not exhibit a significant benefit in the high-risk patients.

Conclusion: The present study showed that age, monocyte percentage, carbohydrate antigen 19-9, lymph node dissection numbers and PNI were independent prognostic factors for pT3N0M0 rectal cancer patients. A nomogram based on these prognostic factors effectively predicts CSS in patients, which can be conveniently used in clinical practice. ACT may improve overall survival in the low-risk patients. But the benefit of ACT was not seen in the high-risk patients.

Keywords: CSS; adjuvant chemotherapy; nomogram; pT3N0M0; rectal cancer.