各位好！今天与大家分享一篇近期发表在JTCVS上的一篇文献。研究纳入了日本三家医疗中心的回顾性数据，针对满足JCOG0802临床试验的肺段切除患者和肺叶切除患者进行了分析。研究提出SUV max<2.0且为非实性结节可能是免于进行纵隔淋巴结清扫的患者。淋巴结转移规律究竟需要什么样的数据/方法才能够进行更进一步的研究？值得思考。
Mediastinal lymph node dissection in segmentectomy for peripheral c-stage IA (≤2 cm) non-small-cell lung cancer
Hiroyuki Adachi, M.D., Hiroyuki Ito, M.D., Takuya Nagashima, M.D., Tetsuya Isaka, M.D., Kotaro Murakami, M.D., Ph.D., Shunsuke Shigefuku, M.D., Ph.D., Noritake Kikunishi, M.D., Naoko Shigeta, M.D., Yujin Kudo, M.D., Ph.D., Yoshihiro Miyata, M.D., Ph.D., Morihito Okada, M.D., Ph.D., Norihiko Ikeda, M.D., Ph.D.
The Journal of Thoracic and Cardiovascular Surgery 2 September 2024
Objective: The necessity of mediastinal lymph node dissection (MLND) in segmentectomy remains uncertain as recent trials on intentional segmentectomy have made MLND mandatory. We conducted a retrospective study to evaluate the necessity of MLND in segmentectomy for patients with peripheral stage IA (≤2 cm) non-small-cell lung cancer (NSCLC).
背景：由于最近关于意向性肺段切除术的试验已规定必须进行MLND，因此在肺段切除术中进行纵隔淋巴结清扫（MLND）的必要性仍然不确定。我们进行了一项回顾性研究，以评估周围IA期（& 2厘米）非小细胞肺癌（SOC）患者在肺段切除术中进行MLND的必要性。
Methods: Of the 5,222 surgical cases for NSCLC from three institutions between 2010 and 2021, 1,457 patients met the JCOG0802 trial eligibility criteria. Initially, we analyzed 574 patients who underwent lobectomy with MLND to identify preoperative risk factors for cN0-pN2 occurrence (Cohort 1). Subsequently, we evaluated the relationship between these factors and the cumulative postoperative recurrence in 390 patients who underwent segmentectomy (Cohort 2).
方法：2010年至2021年间，三家机构共收治了5222例非小细胞肺癌手术病例，其中1457例患者符合JCOG 0802试验资格标准。最初，我们分析了574名接受MLND肺叶切除术的患者，以确定cN 0-pN 2发生的手术前风险因素（队列1）。随后，我们评估了这些因素与390名接受肺段切除术的患者（队列2）累积术后复发之间的关系。
Results: In Cohort 1, risk factors for cN0-pN2 occurrence were consolidation-to-tumor ratio (CTR) =1.0 and maximum standardized uptake value (SUVmax) ≥2.0. When classifying patients into three groups (Group A without any factors, Group B with either factor, and Group C with both factors), the cN0-pN2 occurrence was significantly higher in Group C than in the other groups (0.9%, 3.4%, and 8.4%, respectively, P =0.005). When classifying patients in Cohort 2 using the classification identified in Cohort 1 (117, 131, and 142 were categorized into Group A, Group B, and Group C, respectively), the 5-year cumulative incidence of recurrence rate was significantly higher in Group C than in others (2.0%, 2.0%, and 15.9%, respectively, P <0.001).
结果：在队列1中，cN 0-pN 2发生的风险因素为实变与肿瘤比（TLR）=1.0和最大标准化吸收值（SUVX）= 2.0。当将患者分为三组（A组无任何因素，B组有任何一个因素，C组有两个因素）时，C组的cN 0-pN 2发生率显著高于其他组（分别为0.9%、3.4%和8.4%，P =0.005）。当使用队列1中确定的分类对队列2中的患者进行分类时（分别将117、131和142人分为A组、B组和C组），C组的5年累积复发率明显高于其他组（分别为2.0%、2.0%和15.9%，P <0.001）。
Conclusions: MLND is unlikely to be beneficial in intentional segmentectomy for patients with tumors showing CTR < 1.0 and SUVmax < 2.0.
结论：MLND对CTR < 1.0且SUVs < 2.0的肿瘤患者的意向性肺段切除术可能并无获益。
Central Message: Mediastinal lymph node dissection is unlikely to be beneficial for patients with tumors showing CTR <1.0 and SUVmax <2.0 in segmentectomy for peripheral c-stage IA (≤2 cm) non-small-cell lung cancer.
Perspective Statement: The necessity of mediastinal lymph node dissection (MLND) in segmentectomy remains uncertain. The JCOG0802 trial, which showed the superiority of segmentectomy for c-stage IA (≤2 cm) NSCLC compared to lobectomy in terms of postoperative overall survival, required MLND as mandatory. Our findings identify patients in whom MLND can be omitted, potentially preventing postoperative complications.
1.众所周知，JCOG系列研究为2cm以下早期肺癌的诊疗提供了坚实的诊疗证据。既往研究主要围绕肺叶和亚肺叶切除的对比，而以JCOG1413为牵头的肺叶特异性淋巴结清扫相关研究则不温不火。
2.上细节：
首先本文是基于三家医疗中心的回顾性数据，先针对2cm以下接受肺叶切除术的患者进行了术前高危因素的初步探索，再以所得出的高危因素，在接受肺段切除术的患者中进行验证性预后分析。
其实细看也都有问题。2cm以内磨玻璃为主的淋巴结转移，这个T的范围选择就有些保守。虽然肺段组的淋巴结清扫并非系统性N2，但文中肺段组Group C也仅有4.2%的N2转移率。有限样本量的情况下，纵然是机器学习和深度学习都束手难测。起码得找一个人群的N2转移率稳定在10%左右的，不然真的是鸡蛋里挑骨头了。
第三，早期肺癌JCOG研究方案设计时相对保守。但十余年的数据积累让大家已经认识到磨玻璃成分为主的肺结节的惰性生物学行为。因此在安全范围内说安全范围已经不具有持续性，在安全范围内说不安全的范围才可能具有探索价值。
3. 淋巴结的转移规律很大程度上依赖于现有数据，但目前多数研究很难利用大样本数据把每个患者的每站站淋巴结的转移规律进行数据量化。肿瘤大小/性质在什么范围内进行更严谨的保护机制，仍然需要数据支持。
10.1016/j.jtho.2023.02.010
目录
1. INTRODUCTION
2. Material and Methods (Figure 1)
3. Results (Figure 1)
3.1 Cohort 1 study (Table 1-2)(Supplementary Figure 1)(Table 3)
3.2 Cohort 2 study (Table 4)(Figure 2-4)
4. Discussion
— 图表汇总—
2. Material and Methods
Figure 1. Patient flow chart.
CTR = consolidation-to-tumor ratio, LN = lymph node
3.1 Cohort 1 study
Table 1. Patient characteristics and clinicopathological findings in Cohort 1
* The intrapulmonary and hilar LNs that were resected en-bloc with lung parenchyma were not counted, unless with nodal metastases.
CEA = carcino-embryonic antigen, CTR = consolidation-to-tumor ratio, EGFR = epidermal growth factor receptor, IQR = interquartile range, LCNEC = large cell neuroendocrine carcinoma, LN = lymph node, SCLC = small cell lung cancer, SUVmax = maximum standardized uptake value, VPI = visceral pleural invasion
Table 2. Logistic regression analyses concerning cN0-pN2 occurrence in Cohort 1
CI = confidence interval, CTR = consolidation-to-tumor ratio, OR = odds ratio, SUVmax = maximum standardized uptake value
Supplementary Figure 1. Receiver operating characteristic curve depicting the relation between SUVmax and cN0-pN2 incidence.
Table 3. The cN0-pN2 incidence rate of each group in Cohort 1
Group A: Patients whose primary tumor showed CTR < 1.0 and SUVmax < 2.0. Group B: Patients whose primary tumor showed CTR = 1.0 or
SUVmax ≥ 2.0. Group C: Patients whose primary tumor showed CTR = 1.0 and SUVmax ≥ 2.0.
CTR = consolidation-to-tumor ratio, SUVmax = maximum standardized uptake value
3.2 Cohort 2 study
Table 4. Patient characteristics, clinicopathological findings, and postoperative outcome in Cohort 2
Group A: Patients whose primary tumor showed CTR < 1.0 and SUVmax < 2.0.
Group B: Patients whose primary tumor showed CTR = 1.0 or SUVmax ≥ 2.0.
Group C: Patients whose primary tumor showed CTR = 1.0 and SUVmax ≥ 2.0.
* The intrapulmonary and hilar LNs that were resected en-bloc with lung parenchyma were not counted, unless with nodal metastases.
CEA = carcino-embryonic antigen, CTR = consolidation-to-tumor ratio, EGFR = epidermal growth factor receptor, F/U = follow up, IQR = interquartile range, LCNEC = large cell neuroendocrine carcinoma, LN = lymph node, MLND = mediastinal lymph node dissection, SCLC = small cell lung cancer, SUVmax = maximum standardized uptake value, VPI = visceral pleural invasion
Figure 2. The curves of cumulative incidence of recurrence of each group in Cohort 2.
Group A: Patients whose primary tumor showed CTR < 1.0 and SUVmax < 2.0.
Group B: Patients whose primary tumor showed CTR = 1.0 or SUVmax ≥ 2.0.
Group C: Patients whose primary tumor showed CTR = 1.0 and SUVmax ≥ 2.0.
CI = confidence interval, CTR = consolidation-to-tumor ratio, HR = hazard ratio, SUVmax = maximum standardized uptake value
Supplemental Table 1: Comparison of the patient characteristics and clinicopathological findings between the cohorts
* The intrapulmonary and hilar LNs that were resected en-bloc with lung parenchyma were not counted, unless with nodal metastases.
CEA = carcino-embryonic antigen, CTR = consolidation-to-tumor ratio, EGFR = epidermal growth factor receptor, F/U = follow-up, IQR = interquartile range, LCNEC = large cell neuroendocrine carcinoma, LN = lymph node, MLND = mediastinal lymph node dissection, SCLC = small cell lung cancer, SUVmax = maximum standardized uptake value, VPI = visceral pleural invasion
Figure 3. The Kaplan–Meier curves of recurrence-free survival stratified according to the extent and degree of lymph node dissection in each subset.
Group A: Patients whose primary tumor showed CTR < 1.0 and SUVmax < 2.0.
Group B: Patients whose primary tumor showed CTR = 1.0 or SUVmax ≥ 2.0.
Group C: Patients whose primary tumor showed CTR = 1.0 and SUVmax ≥ 2.0.
CI = confidence interval, CTR = consolidation-to-tumor ratio, HR = hazard ratio, MLND = mediastinal lymph node dissection, NA = not available, SUVmax = maximum standardized uptake value
Figure 4. Graphical abstract of the study.
