Segmentectomy vs Lobectomy for Patients With 2- to 3-cm Non-Small Cell Lung Cancer

Two landmark studies recently demonstrated the noninferiority of sublobar resection to lobectomy in patients with T1a-bN0M0 non-small cell lung cancer (NSCLC). However, whether segmentectomy is associated with similar oncologic outcomes to lobectomy for patients with node-negative T1c tumors (ie, those 2-3 cm) remains unknown.

What is the survival of patients undergoing segmentectomy vs those undergoing lobectomy for T1cN0M0 NSCLC (defined as NSCLC presenting with tumors of 2 to 3 cm not involving the pleura or main bronchi that has not spread to lymph nodes or distant sites)?

Patients in the National Cancer Database (2010-2020) who underwent segmentectomy or lobectomy for clinical T1cN0M0 NSCLC were identified for analysis. Only patients with no comorbidities were included in an effort to minimize selection bias. Overall survival of patients undergoing segmentectomy vs those undergoing lobectomy was evaluated using multivariable-adjusted Cox proportional hazards modeling and propensity score-matched analysis. These analyses then were repeated among a subset of patients with pathologic T1cN0M0 NSCLC.

Of the 12,814 patients with cT1cN0M0 NSCLC identified, 526 patients (4.1%) underwent segmentectomy and 12,288 patients (95.9%) underwent lobectomy. Among patients with clinical T1c disease, no difference was found in overall survival among patients undergoing segmentectomy vs those undergoing lobectomy in multivariable-adjusted analysis (adjusted hazard ratio [aHR], 1.13; 95% CI, 0.92-1.39; p = 0.23) or propensity score-matched analysis (5-year survival: 70.1% [95% CI, 63.8%-75.6%] vs 72.0% [95% CI, 65.6%-77.3%]; P = .85). Similarly, among a subset of 8,502 patients with pathologic T1cN0M0 NSCLC, no difference was found in overall survival between patients undergoing segmentectomy and those undergoing lobectomy in multivariable-adjusted analysis (aHR, 1.23; 95% CI, 0.96-1.59; P = .11) or propensity score-matched analysis (5-year survival: 74.1% [95% CI, 66.7%-80.2%] vs 74.0% [95% CI, 66.4%-80.2%]; P = .73).

In this national analysis, segmentectomy was noninferior to lobectomy for patients with clinical or pathologic T1c NSCLC. These findings suggest that future prospective studies comparing lobectomy and sublobar resection should consider including patients with larger tumors up to 3 cm because these patients also may benefit from parenchymal-sparing approaches.