Neoadjuvant Chemoradiotherapy Plus Local Excision vs. Radical Surgery in T1-2N0M0 Rectal Carcinoma: A SEER-based Propensity Matching Study.

BACKGROUND

Neoadjuvant chemoradiotherapy (NCRT) following local excision (LE) presents a potential strategy for organ preservation in the treatment of early-stage rectal carcinoma that is better than radical surgery (RS), such as Total Mesorectal Excision (TME). Nonetheless, the precise oncological outcomes associated with this approach remain uncertain.

METHOD: We collected clinicopathological data from the Surveillance, Epidemiology, and End Results (SEER) database. To rectify the imbalances in baseline covariates (P < 0.05) between the local excision group and the radical resection group, we employed a 1:1 propensity score matching (PSM) approach. A logistic proportional odds ratio (OR) was employed for the purpose of discerning the factors linked to cancer-specific survival (CSS) and overall survival (OS).

RESULTS

A total of 4882 eligible patients were enrolled in this research. 72 patients were treated with NCRT plus local excision (NCRT + LE group), and 4810 were treated with radical surgery (RS group). Before propensity score matching (PSM), the 3-year, 5-year, and 10-year overall survival (OS) rates were 98.5%, 83.3%, and 46.3% in the NCRT + LE group and 88.8%, 81.8%, and 67.1% in the RS group (P = 0.487, HR = 0.843, 95% CI 0.522-1.363), respectively. In the CSS cohort before PSM, the 3-year, 5-year, and 10-year OS rates were 98.2%, 92.5%, and 45.6% in the NCRT + LE group and 96.8%, 93.3%, and 76.1% in the RS group (P = 0.374, HR = 0.719, 95% CI 0.347-1.492). After 1:1 PSM, each group had 65 patients. The 3-year, 5-year, and 10-year survival rates were 98.3%, 92.5%, and 70.8% in the NCRT + LE group and 94.8%, 92.2%, and 86.4% in the RS group (P = 0.350, HR = 0.701, 95% CI 0.331-1.483), respectively. In the CSS cohort after PSM, the 3-year, 5-year, and 10-year survival rates were 98.2%, 92.3%, and 70.1% in the NCRT + LE group and 96.8%, 93.3%, and 91.1% in the RS group (P = 0.219, HR = 0.560, 95% CI 0.176-1.179), respectively.

CONCLUSION

In this SEER-based analysis, no statistically significant difference in OS or CSS was detected. However, due to limited statistical power, these findings do not indicate equivalence and are hypothesis-generating only.