各位好！今日与大家分享一篇发表在Chest杂志上针对肺癌第九版TNM分期更新的概览和阙疑。没错，自从第九版分期发布后，最让人耳熟能详的就是N2分成了N2a和N2b，以及R分期的一些证据汇总。但在具体临床应用的细节上，回望历史，究竟第九版TNM分期解决了什么？没解决什么？哪些是短期可能能解决的？哪些可能问题是遥遥无期的？一起来看看耶鲁团队的观点。
The Proposed 9th Edition TNM Classification of Lung Cancer
Frank C. Detterbeck, MD, Gavitt A. Woodard, MD, Anna S. Bader, MD, MS, Assistant Professor, Sanja Dacic, MD PhD, Professor of Pathology, Michael J. Grant, MD, Assistant Professor, Henry S. Park, MD, MPH, Associiate Professor, Lynn T. Tanoue, MD, Professor of Medicine
Chest June 15, 2024
A universal nomenclature of the anatomic extent of lung cancer has been critical for individual patient care as well as research advances. As progress occurs, new details emerge that need to be included in a refined system that aligns with contemporary clinical management issues. The 9th edition TNM classification of lung cancer, which is scheduled to take effect in January of 2025, addresses this need. It is based on a large international database, multidisciplinary input and extensive statistical analyses. Key features of the 9th edition include validation of the significant changes in the T component introduced in the 8th edition, subdivision of N2 after exploration of fundamentally different ways of categorizing the N component, and further subdivision of the M component. This has led to reordering of the TNM combinations included in stage groups, primarily involving stage groups IIA, IIB, IIIA and IIIB. This paper summarizes the analyses and revisions for the TNM classification of lung cancer to familiarize the broader medical community and facilitate implementation of the 9th edition system.肺癌解剖范围的通用命名方式对于患者的个体化诊疗与研究至关重要。随着研究的进展，新的细节出现，需要使之包括在一个与当代临床管理问题保持一致的简洁系统中。计划于2025年1月生效的第9版肺癌TNM分类满足了这一需求。该分期以一个大型国际数据库、多学科投入和广泛的统计分析为基础。第9版的主要特点包括验证第8版中引入的T组分的重大变化，在探索了根本不同的N组分分类方法后对N2进行细分，以及对M组分进行进一步细分。这导致了Stage组中包括不同TNM组合的重新排序，主要涉及Stage组IIA、IIB、IIIA和IIIB。本文总结了对肺癌TNM分类的分析和修订，以使医学界更加熟悉和促进第9版系统的实施。
Take Home Points
Study question: How is the 9th edition TNM classification of lung cancer different from the 8th edition system?
Results: N2 is subdivided into single- and multi-station N2, M1c is subdivided into single- and multi-organ system M1c, resulting in a re-arrangement of T and N categories included in the stage groups IIA, IIB, IIIA and IIIB.
Interpretation: A consistent nomenclature about anatomic extent of disease is fundamental to clear communication about clinical trial results and applicability to individual patients.
1. 首先开篇介绍TNM分期并不是一个预后模型，只作为一种肿瘤侵犯范围的解剖性语义词典（预后除了受到TNM分期影响，还有患者、临床背景、治疗相关的因素）。颇为有趣的是作者说了字典也可能会“错”！临床诊疗请不要教条主义（Finally, treatment is not determined by the nomenclature of stage; this merely enables communication.）。
2.看细节：
①推荐记录分期时，记录下来评估方式（E1/2/3/4），这样有助于沟通。
②针对T、N、M组分，分期委员会澄清了在分析数据时做了什么探索和没办法做某些分析的原因（例如为什么没有采用淋巴结个数/站数/区组作为纳入N分期的参数之一）？甚至也谈到了未来哪些可能可以做，哪些根本没前途。也是体现了耶鲁团队在为第九版TNM分期做注过程中的科研品味。
③针砭时弊，不藏着掖着。对于分期进行了澄清就告诉大家哪几类人群的对比很有意思，例如Higer T/lower N和Lower T/higher N的预后对比（虽然目前两类人群OS无显著性差异）。也对临床上常见的肿瘤侵犯区域淋巴结、侵犯局部肋骨/肝脏等（不分为M1）的分期大原则进行了澄清。
④针对肺内多灶病变患者，在四大分类之外，特殊提及了lung cancer with air lucency (LCAL)的研究必要性。
⑤颇为有趣的是R分类居然也可以用成这个样子，R0 (thorax) or R0 (adrenal)。哥，怎么还能这么用！就意思是为了寡转移的深入定义和临床干预方式做铺垫。颇有主考官暗示-外加点题的意思。也提了下除了日本外其他地区几乎不送胸水灌洗液的脱落细胞学。这也导致这个参数的应用范围受限。
⑥Minimal disease Manifestations模块甚至还尝试把乳腺癌中判宏转移和微转移的原则提出到肺癌中进行潜在应用。
3. 究竟非解剖性参数该如何纳入分期，一直是一个颇具争议的话题。可能也不应该拘泥于结构化的分期，未来可能需要按找更复杂的临床状态构建诊疗证据图（前后的诊疗模式识别，而不是诊疗状态识别）。但你回头想下，NCCN指南严格按照分期罗列其实是有问题的，很多潜在的临床场景被忽略了。
我们也更加期待围绕9th分期产出的临床证据~！
目录
1. INTRODUCTION (e-Figure 1)
2. METHODS
2.1 TNM Structure (Table 1-2)(see e-Appendix.)
2.2 Database
2.3 Analysis
3. Results
3.1 TNM Categories and Descriptors (Table 3)
3.2 Stage Groups (Figure 2)(e-Figures 2-4)(Figure 3-4)(e-Tables 1,2)
3.3 General Rules Regarding TNM Classification
3.4 Multiple Pulmonary Sites (Table 4, e-Figure 5)(e-Tables 3-6)
3.5 Other Staging-Related Definitions
4. Discussion
5. Conclusion
— 图表汇总—
1. INTRODUCTION
e-Figure 1. Scope of Data Analyzed for Editions of Lung Cancer TNM Classification
Number of cases, contributing centers and countries for analyses underlying various editions of the lung cancer TNM classification system.
2. METHODS
2.1 TNM Structure
Table 1. Context of TNM Classification
Table 2. Type of Evaluation Used to Identify the Stage of a Tumor in a Patient
CT, computed tomography; EBUS-TBNA, endobronchial ultrasound-transbronchial needle aspiration); MRI magnetic resonance imaging; PET, positron emission tomography.
Supplemental Appendix
2.2 Database
2.3 Analysis
3. Results
3.1 TNM Categories and Descriptors
Table 3. Definitions for T,N,M Descriptors
TX, NX: T or N status unable to be assessed. TX includes tumors proven by the presence of malignant cells in sputum or bronchial washings but not visualized by imaging or bronchoscopy.
a Solitary adenocarcinoma (≤3 cm), predominantly lepidic, and ≤5 mm invasion in any one focus;
b A superficial spreading tumor of any size with its invasive component limited to the bronchial wall, which may extend proximal to the main bronchus, is classified as T1a;
c atelectasis/obstructive pneumonitis may involve part of or the entire lung;
d Pleural effusions are excluded that are cytologically negative and clinically judged not to be due to cancer (e.g. transudative, non-bloody);
e This includes involvement of a single non-regional node;
f A diffuse organ system, such as the skeleton, is considered one organ – i.e. metastases limited to several bones are classified as M1c1.7
Reproduced with permission from Rami-Porta et al J Thorac Oncol 2024.7
The International Association for the Study of Lung Cancer Node Map.
IASLC, International Association for the Study of Lung Cancer; L, left; R, right;
Key boundaries include: The apex of the lung or top of the manubrium distinguishes supraclavicular (#1) from #2R/L nodes, the left border of the trachea to distinguish right from left mediastinal nodes, the lower border of the left innominate vein to distinguish #2R from #4R nodes, the upper border of the aortic arch to distinguish #2L from #4L nodes, the lower border of the azygous vein to distinguish #4R from #10R nodes and the upper border of the left pulmonary artery to distinguish #4L from #10L. The ligamentum arteriosum (in a sagittal plane) is the boundary between #4L (medial) and #5 (lateral to this). The subcarinal station (#7) extends inferiorly to the upper edge of the lower lobe bronchus on the left and the upper edge of the middle lobe/superior segment bronchus on the right.
3.2 Stage Groups
Figure 2. Comparison of 8th and 9th edition Stage Groups
Comparison of 8th and 9th edition stage groups for lung cancer. New N and M categories indicated in Bold font; Red outlines highlight the TNM combinations that are reassigned.
Contr Nod, contralateral separate tumor nodule; Inv, invasion; Ipsi Nod, ipsilateral separate tumor nodule; Les, lesion (extrathoracic metastatic lesion); Mult, multiple; Pl Dissem, pleural or pericardial involvement; Same Lobe Nod, same lobe separate tumor nodule.
Reproduced with permission from: Rami-Porta et al. J Thorac Oncol 20247
e-Figures 2. Overall Survival of 9th Edition Stage Groups
Overall survival by 9th ed clinical and pathologic stage groups. Data involves all evaluable patients with non-small cell lung cancer in the 9th ed database: pre-treatment (c-stage) and M0, surgical patients only, R-any, excluding neoadjuvant therapy (p-stage). Reference: Rami-Porta et al J Thorac Oncol 2024.1
The tables in the lower panel show hazard ratios comparing adjacent stage subgroups calculated by multivariable cox regression, adjusting for covariates of age, sex, region, cell type, and stratified by data source. This analysis demonstrates ordered, stepwise and statistically significant discrimination between each stage subgroup, with progressively worse survival (HR>1).
CI, confidence interval; Ed, edition; HR, hazard ratio; NR, not reached; OS, overall survival; R2, percent of variance explained statistic.
Reproduced with permission from: Rami-Porta et al. J Thorac Oncol 20241
e-Figures 3. Internal Validation – Comparing Training and Validation Datasets
Assessing consistency of discrimination between stage groups by comparing performance of the 9th edition stage groups in the training and validation datasets; top row, c-stage, bottom row, p-stage. Datasets were created by randomly splitting into 2/3rd and 1/3rd cohorts (balanced by year of diagnosis, and data source: electronic data capture system or batch datasets). Reference: Rami-Porta et al J Thorac Oncol 2024.1
NR, not reached; OS, overall survival.
Reproduced with permission from: Rami-Porta et al. J Thorac Oncol 20241
e-Figures 4. Assessment of Generalizability
Assessment of generalizability in domains of time period, geographic region, and aspects of disease- patient- and treatment-spectrum. Additional tests of generalizability are not shown. Green shading = statistically significant, yellow shading = not statistically significant. Numerical values are p values for pairwise (unadjusted) comparisons of Kaplan-Meier overall survival between the stage groups indicated at the beginning of the row in the cohort specified by the table column. Reference: Rami-Porta et al J Thorac Oncol 2024.1
Analyses generally found consistent discrimination and ordering, with rare non-significant exceptions among comparisons with limited sample sizes (indicated by a red cell border). Red font indicates cases with severely limited sample size (<50 cases per cohort).
8th ed DB, 8th edition database; ≤2017, cases diagnosed 2011-2017; ≥2018, cases diagnosed after 2017; N Am, North America; Neoadj, neoadjuvant therapy; Non-sq, non-squamous carcinoma; path, pathologic; PS, performance status; R0, complete resection; ROW, rest of world; Squam, squamous carcinoma; Surg, surgical resection.
Figure 3. Stage Groups in the 9th edition TNM Classification of Lung Cancer
Grid of TNM categories included in stage groups in the 9th edition TNM classification of lung cancer
Figure 4. Specific TNM Categories Included in the 9th edition Stage Groups
Centr, central; Inv, invasion; Mult, multiple; Satel, same lobe separate tumor nodule; Visc Pl, visceral pleural invasion
e-Tables 1. 5-year Overall Survival (%) by Stage Group and TNM Edition
Overall survival in the 9th and 8th ed analyses. Data involves the IASLC global databases, for 9th ed, patients diagnosed 2011-19,1 for 8th ed, patients diagnosed 1999-2010,6 for 7th ed, patients diagnosed 1990-99.7 The proportion of cases from Asia increased (51% for 9th ed, 44% for the 8th ed, 12% for 7th ed).
DB, database; ed, edition; IASLC, International Association for the Study of Lung Cancer
e-Tables 2A: 5-year Overall Survival (%) by Stage Group and Region, 9th edition Data
e-Table 2B: 5-year Overall Survival (%) by Stage Group and Time Period or PS, 9th edition Data
5-year overall survival in the 9th ed global database, NSCLC: A) by region. B) by year of diagnosis and PS.1
Limited sample size (≤20 cases) indicated by parentheses. Rest of world includes Australia, Central and South
America, Africa, and Middle East. a 4-year survival
ed, edition; NSCLC, non-small cell lung cancer; PS, performance status (Zubrod).
3.3 General Rules Regarding TNM Classification
3.4 Multiple Pulmonary Sites
Table 4. Patterns of Disease of Patients with Multiple Pulmonary Sites of Lung Cancer
AIS, adenocarcinoma in situ; GG/L, ground glass/lepidic; LPA, lepidic predominant adenocarcinoma; MIA, minimally invasive adenocarcinoma
Reproduced with permission from: Detterbeck et al. J Thorac Oncol 2016;11(5):639-650.17
e-Figure 5. Examples of Four Patterns of Disease that Present as Multiple Pulmonary Sites of Lung Cancer
Representative examples of 4 patterns of disease which manifest multiple pulmonary sites of lung cancer.
A. Multifocal GG/L lung cancer. A patient with multifocal GG/L tumors in the right upper lobe (who had other GG/L tumors in other lobes). Arrows point to 2 GG/L tumors on CT in the left 2 panels; the next 2 panels show corresponding microscopic images (both were adenocarcinoma with a prominent lepidic component, although with different other adenocarcinoma subtypes). These tumors are classified together as GG/L tumors regardless of such secondary differences.
B. Second primary cancers. A patient with 2 primary lung cancers in the RUL. CT images of each in the left 2 panels, corresponding microscopic images showing an adenocarcinoma and a squamous carcinoma in the next 2 panels. Note that most 2nd primary cancers are of the same (not a different) histologic type.
C. Separate tumor nodules. A patient with a separate tumor nodule of the same histotype as the index tumor. The left panels show CT images of each lesion; the right panels the corresponding microscopic images.
D. Pneumonic-type of lung cancer. A patient with pneumonic-type of lung cancer (this patient also had focal sites of disease in the RLL). The left panels show CT images of the RUL and RML with the typical regional areas with a ground glass and consolidative appearance; the next panels show the corresponding microscopic images.
Adeno, adenocarcinoma; CT, computed tomography; GG/L tumors, tumors with prominent ground glass (imaging) or lepidic (histologic) features; MIA, minimally invasive adenocarcinoma; RLL, right lower lobe; RML, right middle lobe; RUL, right upper lobe; Squam, squamous cell carcinoma.
Reproduced with slight modification with permission from: Detterbeck et al J Thorac Oncol 2016.2
e-Tables 3. Criteria to Distinguish Second Primary vs. Related Tumors
e-Tables 4. Criteria to Categorize a Lesion as a Separate Tumor Nodule
e-Tables 5. Criteria to Categorize a Tumor as Multifocal GG/L Adenocarcinoma
e-Tables 6. Criteria to Categorize a Tumor as a Pneumonic-Type of Adenocarcinoma
3.5 Other Staging-Related Definitions
a recommended assessment is ≥6 node stations (including subcarinal and two other mediastinal stations);
b applies to any site of tumor resection (i.e. primary tumor, involved nodes, resected pleural implants, resected extrathoracic metastasis);
c applies when identified microscopically, regardless of how the nodes are resected (individually, in fragments, en-bloc packet of an entire node station) – provided there is no gross tumor remaining;
d this classification (R1) applies if a resection has been accomplished that meets criteria for R0 in a patient with a malignant pleural (or pericardial) effusion or resected nodules
Reproduced with permission from Detterbeck et al J Thorac Oncol.24
Classification of completeness of surgical resection. Reproduced with permission from: Detterbeck et al J Thorac Oncol. 3
