Risk Factors for Locoregional Relapse After Segmentectomy: Supplementary Analysis of the JCOG0802/WJOG4607L Trial

INTRODUCTION: The JCOG0802/WJOG4607L trial revealed superior overall survival in segmentectomy compared with lobectomy for small-peripheral NSCLC. Nevertheless, locoregional relapse (LR) is a major issue for segmentectomy. An ad hoc supplementary analysis aimed to determine the risk factors for LR and the degree of advantages of segmentectomy on the basis of primary tumor sites. METHODS: Participants in multi-institutional and intergroup, open-label, phase 3 randomized controlled trial in Japan were enrolled from August 10, 2009, to October 21, 2014. Risk factors for LR after segmentectomy and clinical features following the primary tumor site were investigated. RESULTS: Of 1105 patients, 576 and 529 underwent lobectomy and segmentectomy, respectively. The primary tumor site for segmentectomy was the left upper division, left lingular segment, left S6, left basal segment, right upper lobe, right S6, or right basal segment. Multivariable analysis in the segmentectomy group revealed that pure-solid appearance on thin-section computed tomography (OR = 3.230; 95% confidence interval [CI]: 1.559-6.690; p = 0.0016), margin distance less than the tumor size (OR = 2.682; 95% CI: 1.350-5.331; p = 0.0049), and male sex (OR = 2.089; 95% CI: 1.047-4.169; p = 0.0366) were significantly associated with LR. Patients with left lingular segment tumors (OR = 4.815; 95% CI: 1.580-14.672) tended to experience LR more frequently than those with left upper division tumors, although primary tumor sites were not statistically significant. CONCLUSIONS: Thin-section computed tomography findings and margin distance are important factors to avoid LR in segmentectomy.