医学
内科学
化疗
放射治疗
放化疗
结直肠癌
肿瘤科
淋巴细胞
诱导化疗
胃肠病学
新辅助治疗
回顾性队列研究
前瞻性队列研究
外科
完全响应
子群分析
单变量分析
多元分析
探索性分析
作者
Rie Sasaki,Senzo Taguchi,Hikaru Miyauchi,Yasuo Yoshioka,Eiji Shinozaki,Kensei Yamaguchi,Tomohiro Yamaguchi,Takashi Akiyoshi,Seno Satoshi,Takeaki Ishihara,Daisuke Miyawaki,Ryohei Sasaki
摘要
ABSTRACT Total neoadjuvant therapy (TNT) improves oncological outcomes in locally advanced rectal cancer (LARC); however, treatment-induced lymphopenia remains a concern. We analyzed 74 patients undergoing three TNT regimens: long-course chemoradiotherapy with consolidation chemotherapy (LCCRT–CNCT), short-course radiotherapy with CNCT (SCRT–CNCT), and induction chemotherapy with LCCRT (INCT–LCCRT). Severe radiation-induced lymphopenia (RIL, Grade ≥ 3) occurred in 48%, 24%, and 54%, respectively (P = 0.126). In the LCCRT–CNCT group, large bowel irradiation (V35 Gy > 46 cc) was significantly associated with severe RIL in univariable analysis but not in multivariable models (P = 0.227), and in an exploratory combined analysis of LCCRT–CNCT and INCT–LCCRT, this showed a trend (P = 0.093). Pre-TNT absolute lymphocyte count (ALC) was an independent predictor of RIL. Small bowel irradiation (V15 Gy > 104 cc) predicted severe lymphopenia during chemotherapy in the univariable analysis; but multivariable analysis suggested pre-TNT ALC as the main factor, showing a trend toward significance (P = 0.051). In the SCRT–CNCT group, pre-TNT ALC was the only significant factor for severe lymphopenia in both the RT and chemotherapy phases in univariable analysis. Severe RIL significantly prolonged lymphocyte recovery time (median, 283 vs. 76 days, P < 0.001), whereas immune recovery did not differ according to the TNT regimen. The median ALC at the last follow-up was 86% of the baseline value, indicating incomplete recovery. While pre-TNT ALC correlated with lymphopenia risk, minimizing bowel irradiation may help mitigate treatment-induced immunosuppression. Prospective studies are required to validate these findings.
科研通智能强力驱动
Strongly Powered by AbleSci AI