Richard E. Clark Memorial Paper for General Thoracic Surgery: Lobectomy vs. Sublobar Resection Among Lung Cancer Patients Understudied in Recent Clinical Trials

G. A.. Woodard¹, M. Grau-Sepulveda², M. Onaitis³, B. V. Udelsman⁴, E. A. David⁵, J. Jacobs⁶, A. Kosinski², J. Blasberg⁷, D. Boffa^8
1Yale School of Medicine, New Haven, Connecticut ²Duke Clinical Research Institute, Durham, North Carolina ³UC San Diego Health, La Jolla, California ⁴Keck School of Medicine, University of Southern California, Los Angeles, California ⁵University of Colorado, Aurora, Colorado ⁶University of Florida, Gainesville, Florida ⁷Yale University School of Medicine, Westport, Connecticut ⁸Yale University School of Medicine, New Haven, Connecticut

Purpose: The prospective randomized CALGB 140503 trial demonstrated equivalent survival between sublobar resection and lobectomy in healthy non-small cell lung cancer (NSCLC) patients with tumors 2 cm or less. We examined survival across patient groups that were poorly represented in the trial to evaluate widespread applicability of the trial findings.

Methods: Patients with 2 cm or less, node negative NSCLC (cT1N0) who underwent elective lobectomy, segmentectomy, or wedge resection in the Society of Thoracic Surgeons prospective database from years 2012-2000 were linked to Medicare survival data using a well-documented probabilistic matching algorithm. Patients with previous chemotherapy, radiation, or operation to the same thoracic site, those with Zubrod performance status 4 or 5 were excluded. Patient groups underrepresented in randomized trials included age greater than 75 years, decreased pulmonary function (FEV1=10-80%), performance status greater than 1, never smokers, and incomplete or positive nodal evaluations. An incomplete nodal evaluation was defined as 2 or fewer lymph nodes sampled. Propensity scores overlapping weights were computed for each patient to ensure balance of known cofounders. Survival was assessed by propensity score weighted Kaplan Meier analysis.

Results: Overall, 20,031 patients were identified including 11,976 lobectomy, 2,586 segmentectomy, and 5,469 wedge resection patients. Sublobar patients tended to have worse performance status (Zubrod 2-3, 5% vs 3%), worse pulmonary function (17% vs 7.6% FEV1 less than 60%), fewer never smokers (13.9% vs 15.7%), fewer lymph nodes sampled (5.5 vs 12.8), and less pathologic upstaging (7.1% vs 14.2%). Among subgroups underrepresented in the randomized trial, sublobar and lobar resection resulted in similar overall survival in patients over age 75 years (p=0.0663), poor FEV1=10-59% (p=0.1391), and Zubrod performance status 2-3 (p=0.2314). Similarly, never smokers undergoing sublobar and lobar resections also had no survival differences (p=0.3413).

However, failure to adequately evaluate lymph nodes (2 or fewer nodes) during a sublobar resection was associated with worse survival, and less clearly similar to the overall lobectomy cohort. Unanticipated nodal disease found on pathologic upstaging was associated with worse survival after lobectomy and sublobar resection without significant differences in outcomes based on the extent of pulmonary resection (p=0.4163).

Conclusion: Recent trial findings of similar survival following sublobar and lobar resection in node-negative NSCLC tumors less than 2 cm appear to extend to several patient groups understudied in the CALGB trial, including less healthy, older patients with worse pulmonary function, and never smokers. Similarly, nodal upstaging was associated with worse survival regardless of extent of resection; lobectomy did not improved survival over sublobar resection. Further study is warranted to clarify the optimal extent of resection in less healthy patients and those with micrometastatic nodal disease.

Identify the source of the funding for this research project: Supported by STS GTSD Access and Publications Research Project