Several speakers took the stage during the 2026 World Conference on Lung Cancer (WCLC) to discuss the recently released 9th Edition of the Tumor, Nodes, Metastasis (TNM) Classification.

The session focused on the evolution of the TNM Classification system for lung cancer, as well as future directions for its 10th Edition.

Impact of Smoking Status on Staging

Lawson Eng, MD, SM, FRCPC, a medical oncologist at the Princess Margaret Cancer Center and an assistant professor in the Department of Medicine at the University of Toronto, Canada, addressed lessons learned from the 9th Edition, specifically concerning smoking and prognostication.

Dr. Eng highlighted that, while staging has traditionally been anatomical in many cancer types, growing evidence suggests that non-anatomical factors can also affect outcomes and may help define prognostic or staging groups.

“This led us to ask: Does smoking affect prognosis in lung cancer, and how much does it affect prognosis compared with anatomical factors?” Dr. Eng said.

He presented data evaluating smoking status in the IASLC 9th Edition staging cohort, which consisted of 48,531 patients with non-small cell lung cancer (NSCLC). The results showed that having a history of tobacco use effectively worsens prognosis by one stage group.

Within each stage, patients who currently or formerly smoked had a poorer prognosis than patients who never smoked. Additionally, across stages, patients who smoked had a prognosis similar to those who never smoked in the next highest stage group.

As for next steps, Dr. Eng noted that the 10th Edition aims to improve data collection on tobacco elements, specifically smoking intensity and quit dates.

“We’re also looking at information about other tobacco-related exposures, such as vaping and cannabis use,” Dr. Eng said. “We plan to hopefully understand the effects of these behaviors alongside other non-anatomical factors in the 10th edition.”

Refining Anatomic Descriptors

Masaya Yotsukura, MD, attending surgeon in the Department of Thoracic Surgery at the National Cancer Center Hospital in Japan, discussed the potential to further refine anatomical risk factors within the TNM Classification.

He reviewed key changes made in the 9th Edition classification. These include dividing the N2 category into N2a and N2b, depending on the extent of mediastinal metastases. Another important change is the division of the M1c category into M1c1 and M1c2, depending on the number of extrathoracic metastases.

There were no changes made to the T descriptors in the 9th Edition. Since the T category is based primarily on anatomical extent, the survival curves are well separated, leaving limited scope for further refinement based solely on anatomical factors, Dr. Yotsukura noted.

Four areas he said may warrant further consideration include chest-wall invasion, interlobar invasion, tumors with ground-glass opacity or a lepidic component, and histological descriptors. Dr. Yotsukura also noted other possible areas for refinement, including separate staging for ypStage and pStage, as well as incorporating biomarkers into future TNM classifications.

“Biomarkers might improve the accuracy of lung cancer staging, but there are important challenges: Measurement methods need to be standardized, and the system must be globally applicable,” Dr. Yotsukura said. “If biomarkers are included in the TNM, they must be usable around the world.”

Adding Non-Anatomic Factors to the TNM Classification

Frank C. Detterbeck, MD, FCCP, professor of thoracic surgery at Yale University School of Medicine, discussed the current landscape of tumor classification and potential future directions.

Walking attendees through how tumor classification is used, Dr. Detterbeck described the TNM system as a clinically oriented tool designed to communicate the anatomic extent of tumors and streamline discussions in patient care. However, he noted that while the TNM system summarizes anatomic extent, no tool summarizes biomarker and predictive factors.

Additionally, he noted that non-anatomical factors are becoming increasingly important, citing that anatomic tumor extent accounts for only 67% of prognosis and that non-anatomic tumor characteristics increasingly inform clinical decision-making regarding treatment approaches.

While he acknowledged the need to incorporate prognostic factors, he emphasized the importance of understanding why the TNM system has been successful. He also cautioned against losing sight of previous efforts and urged consideration of how an enhanced tumor classification system can improve care.

“Before we focus solely on prognosis, we need to decide what we actually want and how it will function in practice. That should guide the modeling used for the 10th edition,” Dr. Detterbeck said.

AI Tools for Staging Diagnosis

Kwun Fong, MBBS, PhD, FRACP, clinical manager of the Pulmonary Malignancy Unit at the Prince Charles Hospital University of Queensland Thoracic Research Center, Australia, explored the potential of artificial intelligence (AI) tools—such as deep learning models, convolutional neural networks, and vision transformers—to enhance staging.

He explained how such tools could assist pathologists at each stage, from pre-procedural planning and real-time navigation to lesion localization, biopsy, and on-site tissue evaluation.

“These tools are coming, and we need to understand how we might use them,” Dr. Fong said.

He said AI could potentially reduce the need for certain procedures and prioritize high-risk lesions for biopsy by identifying features on the CT or PET scans that aren’t easily visible to the human eye. He also discussed how AI tools can help provide procedural guidance during bronchoscopy. However, he noted that implementation presents challenges, such as obtaining regulatory approval and integrating AI into workflows.

“AI is promising. It may help risk-stratify lesions and lymph nodes, guide procedures, and improve staging. However, the evidence is still limited and largely retrospective, and real-world barriers remain,” Dr. Fong said. “We need appropriate performance metrics and evidence of clinical utility. Multimodal approaches combining radiomics and pathology may become particularly useful. For now, AI is more likely to augment our practice than replace it.”

Mitigating Global Disparities in the TNM Staging System

Francisco Suarez, MD, associate professor and thoracic surgeon at Clínica Santa María, Chile, explained that while the TNM system is universal, staging is not.

“TNM does one thing very well: it describes the extent of cancer. It provides a common language for planning treatment, estimating prognosis, and comparing outcomes across centers and countries,” Dr. Suarez said. “However, what we can describe depends on how well we look for disease.”

He explained that where a patient is staged can influence how much disease is detected, highlighting differences in staging capabilities between high- and low-resource settings.

Dr. Suarez said that while the TNM system cannot eliminate global inequity, it should not amplify it. He highlighted several distinct problems: missing staging information, unequal diagnostic intensity, an uneven evidence base, and new resource requirements.

He advocated for a single TNM system, rather than separate systems for different regions, that uses consistent definitions and language across diverse contexts.

“We should represent the world better, report stage within context, validate TNM in different settings, and consider equity when designing future editions,” Dr. Suarez said. “The goal is not only greater precision but precision that remains useful worldwide.”