【目的】
分析局部晚期直肠癌盆腔侧方淋巴结 (lateral pelvic lymph nodes, LPLN)转移的发生概率、临床因素相关性，以及LPLN转移对预后的影响。将目前利用放疗治疗LPLN转移的模式进行比对，为未来的治疗选择提供参考。

【方法】
本研究定义盆腔LPLN范围包括：髂内血管区、闭孔血管区，盆腔LPLN转移定义为MRI评估短径＞7 mm。研究分两部分阐述，第一部分分析一项III期随机对照研究（STELLAR, NCT02533271）的数据，该研究纳入从2015年8月30日-2018年8月27日间、全国16家医疗中心连续收治的591例局部晚期直肠癌患者。该研究中所有患者均接受新辅助化放疗【(chemo)radiotherapy, (C)RT】联合直肠全系膜切除 (total mesorectal excision, TME)术治疗策略。本课题利用该数据评估：1.局部晚期直肠癌LPLN转移概率；2.分析LPLN转移的相关临床因素（T/N分期，肿瘤位置，肿瘤大小等）；3.评估LPLN转移对预后的作用；4.不同的新辅助治疗对LPLN转移的直肠癌患者的预后影响。
研究第二部分，利用本中心2015年8月30日-2021年11月30日的数据，将目前治疗直肠癌伴有LPLN转移患者的手段进行比对分析，具体为：(C)RT +TME术vs. TME术+侧方淋巴结清扫；(C)RT +TME术vs. (C)RT +TME术+侧方淋巴清扫；(C)RT +LPLN推量+TME术vs. (C)RT +TME。组间比较采用卡方检验，Kaplan-Meier法计算DFS、LRR、DM以及OS并Log-rank检验，多因素生存分析采用CoX比例风险模型，P<0.05为差异有显著性。

【结果】
研究第一步部分结果显示：在591例患者中，99例(16.8%)诊断为LPLN转移，大部分情况为单侧的LPLN转移(79.8%)，以髂内淋巴结转移更为常见(81.8%)。有、无LPLN转移与直肠分段(χ²=14.864, P=0.001)、N分期(χ²=25.078, P＜0.001)、临床分期(χ²=16.237, P＜0.001)、系膜淋巴结是否阳性(χ²=4.721, P=0.030)显著相关。LPLN转移组和未转移组的3年DFS率分别为51.4% vs. 68.2%(P＜0.001)，3年OS率分别为71.8% vs. 84.2%(P=0.006)，3年MFS率分别为60.8% vs 80.1%(P＜0.001)，3年LRR率分别为11.4% vs 8.5%(P=0.564)。多因素分析发现LPLN转移是影响DFS率(P=0.005)、OS率(P=0.036)及MFS率(P=0.001)的独立预后因素。未发现短程放疗联合化疗模式较常规CRT在LPLN转移患者治疗中的获益，3年DFS率分别为57.5% vs. 45.5%(P=0.087)，3年OS率分别为76.9% vs. 66.7%(P=0.166)，3年MFS率分别为66.9% vs. 54.9%(P=0.099)，3年LRR率分别为7.0% vs. 15.1%(P=0.237)。
研究第二部分，对比(C)RT +TME术vs. TME术+侧方淋巴结清扫，两种模式均是有效的治疗方法，对于LPLN病理阳性者，(C)RT +TME模式可获得相似的DFS率（1年DFS率，79.4% vs. 100.0%）和OS率（1年OS率，93.9% vs. 94.7%），需进一步延长随访时间。对比(C)RT +TME术vs. (C)RT +TME术+侧方淋巴清扫，3年DFS率分别为53.1% vs. 51.4%(P=0.746)，3年OS率分别为81.8% vs. 71.8%(P=0.535)，未发现侧方清扫对DFS率和OS率的影响，前瞻性研究和大样本量的数据是值得期待的。对比(C)RT +LPLN推量+TME术vs. (C)RT +TME，经过倾向评分配比（PSM）法配对，3年DFS率分别为78.6% vs. 83.3%（P=0.762），3年OS率分别为84.4% vs. 90.3%（P=0.803），行新辅助同步放化疗并侧方淋巴结推量初步被证实是安全有效的，仍需前瞻性研究并扩大样本量来证实。

【结论】
高分辨率MRI观察到的LPLN转移在局部晚期直肠癌患者中需要被关注，特别是对低位肿瘤、N2、直肠系膜淋巴结阳性人群，且MRI判断的LPLN转移是预后不良因素。(C)RT +TME术是有效的，避免了侧方清扫以及带来的损伤，不同的治疗方法如TME术+侧方清扫，(C)RT +TME术+侧方清扫以及(C)RT +LPLN推量+TME术，其优劣还需要前瞻性、大样本研究来评估。

【Objective】To analyze the probability of lateral pelvic lymph nodes (LPLN) metastasis in locally advanced rectal cancer, the correlation of clinical factors, and the impact of LPLN metastasis on prognosis. To compare different comprehensive therapies involving radiotherapy to provide reference for future treatment options.

【Methods】The scope of LPLN defined in this study included: internal iliac region and obturator region. LPLN metastasis was defined as MRI evaluation of short diameter>7 mm. The first part of the study analyzed data from a phase III randomized controlled study (STELLAR, NCT02533271) that included 591 patients with locally advanced rectal cancer from 16 medical centers in China from August 30, 2015 to August 27, 2018. In this study, all patients received neoadjuvant (chemo)radiotherapy ((C)RT) combined with total mesorectal excision (TME) surgery. This study uses this data to evaluate: 1. The probability of LPLN metastasis in locally advanced rectal cancer; 2. Analyzing the clinical factors related to metastasis of LPLN (T/N stage, tumor location, tumor size, etc.); 3. Evaluating the effect of LPLN metastasis on prognosis; 4. The prognostic impact of different neoadjuvant therapies on rectal cancer with LPLN metastasis. In the second part of the study, using the data from our center from August 30, 2015 to November 30, 2021, we compared and analyzed the current treatment methods for patients with rectal cancer accompanied by LPLN metastasis, specifically: (C)RT + TME vs. TME + lateral lymph node dissection; (C)RT + TME vs. (C)RT + TME + lateral lymph node dissection; (C)RT+ LPLN boost + TME vs. (C)RT + TME. Chi-square test was used for inter-group comparison, Kaplan Meier method was used to calculate DFS, LRR, DM, and OS, and Log-rank test was used. CoX proportional risk model was used for multivariate survival analysis, and P<0.05 was a significant difference.

【Results】Of the 591 patients, 99 (16.8%) were diagnosed with LPLN metastases, mostly with unilateral metastasis (79.8%), with internal iliac lymph node metastasis being more common (81.8%). The 3-year DFS rate in the LPLN metastatic group and the non-metastatic group were 51.4% vs. 68.2% (P<0.001), with the 3-year OS rate 71.8% vs. 84.2% (P=0.006), the 3-year MFS rate 60.8% vs. 80.1% (P<0.001), and the 3-year LRR rate 11.4% vs. 8.5% (P=0.564), respectively. Multivariate analysis found that LPLN metastasis was an independent prognostic factor affecting the DFS rate (P=0.005), OS rate (P=0.036), and MFS rate (P=0.001). No benefit was found in the treatment of patients with LPLN metastases using short-term radiotherapy combined with chemotherapy compared to CRT. The 3-year DFS rate were 57.5% vs. 45.5% (P=0.087), with the 3-year OS rate 76.9% vs. 66.7% (P=0.166), the 3-year MFS rate 66.9% vs. 54.9% (P=0.099), and the 3-year LRR rate 7.0% vs. 15.1% (P=0.237). Comparing (C)RT + TME vs. TME plus lateral lymph node dissection, both were effective treatment methods. For patients with pathologic LPLN metastases, (C)RT + TME mode could achieve similar DFS rates (1-year DFS rate, 79.4% vs.100.0%) and OS rates (1-year OS rate, 93.9% vs.94.7%), requiring further extension of follow-up time. Comparing (C)RT + TME vs. (C)RT + TME + lateral lymphadenectomy, it resulted in a 3-year DFS rate of 53.1% vs. 51.4% (P=0.746), and a 3-year OS rate of 81.8% vs. 71.8% (P=0.535), respectively. No impact of lateral lymphadenectomy on DFS and OS rates was found, but prospective studies and large sample data are expected. Comparing (C)RT + LPLN boost + TME vs. (C)RT + TME, it showed a 3-year DFS rate of 78.6% vs. 83.3% (P=0.762), and a 3-year OS rate of 84.4% vs. 90.3% (P=0.803), using the propensity score matching (PSM) method. (C)RT + LPLN boost was initially confirmed to be safe and effective, and further prospective research and expanded sample size are needed.

【Conclusions】LPLN metastasis observed by high-resolution MRI needs to be paid attention to in patients with locally advanced rectal cancer, and it is a poor prognostic factor. (C)RT+TME is effective and avoids lateral dissection with the resulting damage. Different treatment methods, such as TME + lateral dissection, (C)RT + TME + lateral dissection, or (C)RT + LPLN boost + TME, need to be evaluated through prospective, large sample studies.