Survival analysis of early-onset locally advanced rectal cancer: a retrospective study based on the Surveillance, Epidemiology, and End Results (SEER) database

Abstract

Objective: To investigate the factors influencing tumor-specific survival of early-onset locally advanced rectal cancer. Methods: All-age patients with primary locally advanced rectal cancer from the Surveillance, Epidemiology, and End Results (SEER) database (2010 to 2019) were included in this study. Early- and late-onset locally advanced rectal cancer was defined according to age of 50 years at diagnosis. Early-onset locally advanced rectal cancer was divided into five age groups for subgroup analyses. Age, sex, tumor-specific survival time and survival status of patients at diagnosis, pathological grade, TNM stage, perineural invasion, tumor deposits, tumor size, pretreatment CEA , radiotherapy, chemotherapy, and number of lymph node dissections were included. Progression-free survival (PFS) was analyzed and compared between patients with early- and late-onset rectal cancer. Results: A total of 5,048 patients with locally advanced rectal cancer were included in the study (aged 27-70 years): 1,290 (25.55%) patients with early-onset rectal cancer and 3,758 (74.45%) patients with late-onset rectal cancer. Patients with early-onset rectal cancer had a higher rate of perineural invasion (P<0.001), more positive lymph nodes dissected (P<0.001), higher positive lymph node ratios (P<0.001), and a higher proportion receiving preoperative radiotherapy (P=0.002). Patients with early-onset rectal cancer had slightly better short-term survival than those with late-onset rectal cancer (median (IQR ): 54 (33-83) vs 50 (31-79) months, χ2=5.192, P=0.023). Multivariate Cox regression for all patients with locally advanced rectal cancer showed that age (P=0.008), grade of tumor differentiation (P=0.002), pretreatment CEA (P=0.008), perineural invasion (P=0.021), positive number (P=0.004) and positive ratio (P=0.001) of dissected lymph nodes, and sequence of surgery and radiotherapy (P=0.005) influenced PFS. This suggests that the Cox regression results for all patients may not be applicable to patients with early-onset cancer. Cox analysis showed tumor differentiation grade (patients with low differentiation had a higher risk of death, P=0.027), TNM stage (stage III patients had a higher risk of death, P=0.025), T stage (higher risk of death in stage T4, P<0.001), pretreatment CEA (P=0.002), perineural invasion (P<0.001), tumor deposits (P=0.005), number of dissected lymph nodes (patients with removal of 12-20 lymph nodes had a lower risk of death, P<0.001), and positive number of dissected lymph nodes (P<0.001) were independent factors influencing PFS of patients with early-onset locally advanced rectal cancer. Conclusion: Patients with early-onset locally advanced rectal cancer were more likely to have adverse prognostic factors, but an adequate number of lymph node dissections (12-20) resulted in better survival outcomes.