Investigation of Practice Patterns in Elderly Patients with Locally Advanced Rectal Cancer and Prediction of Survival by a Validated Model

Abstract

Rectal cancer is one of the most common malignant tumors in China and the world, with more than half of the patients aged ≥70 years at the time of diagnosis. However, the optimal treatment strategies in elderly patients with locally advanced rectal cancer (LARC) are still controversial. We designed our study to delineate the pattern of treatment for elderly LARC and analyze the risk factors that affect their overall survival (OS), and provide a basis for further analysis of decision-making for treatment of elderly patients. We retrospectively analyzed 187 elderly patients (≥70 years ) with LARC, who had received surgeries in our institution between January 2008 and December 2018, to evaluate the OS of four subgroups of patients according to four different treatment modalities: surgery only, neoadjuvant (C)RT, adjuvant (C)RT and adjuvant chemotherapy (no RT) and compared with 11,347 elderly LARC patients from the Surveillance, Epidemiology, and End Results (SEER)- registered database between 2004 and 2016. We also established a nomogram for predicting OS, which was evaluated by Harrell’s concordance index (C-index) and calibration plots in both training and validation cohort. 187 elderly patients with LARC in our institution were more frequently treated with only received surgery (49.2%) ,9.1% of patients received neoadjuvant (chemo)radiotherapy, adjuvant (C)RT and adjuvant chemotherapy (no RT) accounted for 18.7% and 23%, respectively. In the SEER cohort, surgery alone (41.8%) was also the main treatment strategy. The treatment patterns of the two cohorts were significantly different (P <0.001), especially in the application of radiation therapy (P <0.001). The 5-year overall survival (OS) was 45.0% in surgery only, 77.1% in n(C)RT, 54.6% in adjuvant (C)RT and 61.7% in adjuvant chemotherapy (no RT), which had significant difference in univariate and multivariate Cox regression (P < 0.05). The C-index of the Nomogram model based on independent predictive factors and the calibration curve showed a high degree of agreement between the predicted and actual survival rates. The survival results of elderly LARC patients in our institution are similar to the clinical data of the SEER database during the same period, but there is a large gap in the treatment patterns, especially the application of radiation therapy. Radical surgery after neoadjuvant (chemo)radiotherapy may be the best treatment strategy for elderly patients with LARC, but individual analysis and appropriate adjustments still need to be made according to comorbidities and treatment benefits. The proposed nomograms based on independent clinicopathological variables may be practical and helpful for precise evaluation of patient prognosis, and guiding the individualized treatment for elderly LARC.