Since its inception, the tumor, node, metastasis (TNM) classification of non-small cell lung cancer (NSCLC) has sought to estimate the prognosis of the disease and guide patient care. Over the past five years, largely though efforts by the International Association for the Study of Lung Cancer (IASLC), this classification has evolved from empiric taxonomy based on anatomic extent of spread to a more evidence-based categorization through studying the impact of each clinical variable on patient outcome.
Analysis of registry databases by the IASLC's multidisciplinary team provided “real world” estimates of the individual contribution of clinical variables to prognosis, which informed the 7th edition of the TNM classification of NSCLC, which was adopted in 2010.
With advances in treatment of lung cancer, better supportive care, and changes in clinical practice, it became necessary to revisit the staging, and with this iteration, examine a larger body of patient data, with adequate representation of the global community and using various statistical algorithms to further refine the staging system.
The IASLC International Staging Committee used an international database of 94,708 patients, with source data from various consortia, registries, surgical and institutional series and registries to inform the proposals for the 8th edition.
The majority of data continue to be retrospective, though prospective data are also being curated. Surgically treated patients continue to form the largest chunk of this database, but information regarding radiotherapy, systemic therapy, and various permutations of these options has also been collected.
The most extensive changes proposed by IASLC for the 8th edition pertain to the T classification, with the importance of tumor size being highlighted, which has been shown to predict outcome. A tumor size descriptor has been added to each T category (T1 to T4). Each centimeter of tumor size has its own T descriptor up to 5 cm of size (T1a, T1b, T1c, T2a, T2b). Five to seven cm is categorized together as T3 according to prognosis. Beyond the 7 cm size cutoff, tumor size alone does not reliably sort patients into prognostic categories and larger tumors are grouped together as T4.
Atelectasis of part of the lung (previously T2) and complete atelectasis (previously T3) were both found to have a similar prognosis, and therefore both have been reclassified into the T2 category. Diaphragm invasion (previously T3) was upgraded to T4 due to the worse prognosis associated with it. Endobronchial tumors were previously staged based on distance from carina (previously T2 and T3).
This round of analysis showed similar prognosis irrespective of the distance from carina, due to which more proximal (previously T3) tumors were downgraded to T2 tumors. Mediastinal pleural invasion (previously T3) was described so infrequently that the decision was made to remove it from staging altogether.
The nodal staging for the 8th edition remains unchanged, though new N descriptors have been proposed for prospective testing and validation. Proposals include dividing pathologic (p) N1 into two subcategories based on single (pN1a) and multiple (pN1b) nodal station involvement.
The proposal further subdivided pN2 involvement into three categories:
- pN2a1, which described single pN2 nodal station involvement without pN1 disease, also termed “skip metastasis” which carried a better prognosis than traditional pN2 disease;
- pN2a2 with single station pN2 and pN1 involvement; and
- pN2b with involvement of multiple pN2 nodal stations.
The M staging has also undergone revision, with the addition of a third category. M1a remains as it is, including patients with pleural metastasis, malignant pleural or pericardial effusions or metastasis to the contralateral lung. The M staging for the first time recognizes oligometastatic disease as a separate category, such as isolated brain metastasis where surgery or gamma knife therapy may be considered, or isolated adrenal metastasis where adrenalectomy may be offered.
Patients with a single metastasis in a single organ have been termed M1b, while patients with multiple metastases either in a single organ, or in multiple organs, were reclassified as M1c.
The overall stage groupings have also changed as a result:
- Stage IA continues to include tumors that are node negative and 3 cm or less in size, though it has been subdivided into three groups—IA1, IA2, and IA3—with a category for each centimeter increase in size;
- Stage IB, which previously included tumors greater than 3 cm but less than 5 cm, now includes only tumors up to 4 cm in size, while IIA includes tumors up to 5 cm in size;
- Tumors will be categorized as IIB if they are node negative and greater than 5 cm and less than 7 cm in size, or are smaller tumors with N1 nodal involvement;
- Stage IIIA will include node-negative tumors larger than 7 cm, in addition to those that invade the diaphragm, mediastinum, heart or great vessels, trachea or carina, esophagus, or vertebral body. It will also include T3-T4 N1 tumors, and T1a-T2b N2 tumors;
- Stage IIIB will include T3-4 N2 tumors and T1a-T2b N3 tumors.
- A new category, Stage IIIC will include T3-T4 N3 involvement; and
- Stage IV has also been divided into IVA and IVB: Stage IVA will include tumors of any size or nodal involvement with M1a and M1b metastases, while IVB will include patients with more than one metastasis in one or more distant organs.
It is heartening to see the incorporation of evidence-based approaches to stratify NSCLC prognosis. Future clinical trials should incorporate the proposed staging changes. Studies in patients with locally advanced NSCLC should stratify patients into IIIA, IIIB, and IIIC, while trials in advanced stage disease should group patients by the presence of intrathoracic metastasis only, oligometastatic disease, and more than one site of distant disease (M1a, M1b, and M1c respectively).
The TNM staging still does not incorporate data regarding tumor genomics, including the presence of driver alterations in EGFR, ALK, and ROS1, which have been shown to influence outcome. It is possible that future versions may encompass these molecular alterations, in addition to immunologic markers. We will stay tuned.