Significance of Lymphadenectomy in Part-Solid Lung Adenocarcinoma: Propensity Score Matched Analysis.

Abstract

It is evident that the frequency of nodal metastasis in part-solid lung adenocarcinoma is significantly low. For those cases, the prognostic impact of the lymph node dissection (LND) is highly controversial. We reviewed 462 clinical stage I radiologic part-solid lung adenocarcinoma patients who underwent lobectomy or segmentectomy with lymphadenectomy. Consolidation tumor ratio was evaluated for all patients. Extent of LND was classified into systematic/lobe-specific mediastinal LND (m-LND) and hilar LND only (h-LND). Prognostic significance of LND was assessed by a multivariable analysis using propensity score matching. Survival was calculated by the Kaplan-Meier method using the log rank test. The m-LND was performed in 314 patients (68%), and h-LND in 148 (32%). Overall survival (OS) was not significantly different between m-LND and h-LND (5-year OS, 94.2% versus 92.8%, p = 0.585), as shown in 92 matched pairs (5-year OS, 94.0% versus 93.2%, p = 0.845). Cox proportional hazards model revealed maximum standardized uptake value as an independently significant oncologic prognosticator (p = 0.003), but extent of LND was not (p = 0.997). Nodal involvement was found in 16 patients, and was exclusive in solid-dominant lesions (ie, 0.5 less than consolidation tumor ratio of 1.0 or less; 4.9%). Multivariable analysis revealed carcinoembryonic antigen and maximum standardized uptake value as significant predictors of nodal metastasis among solid-dominant lesions (0.001, 0.002). For 329 solid-dominant lesions (71%), however, survival was similar between m-LND and h-LND (5-year OS, 93.2% versus 87.1%, p = 0.097), and was proven in 58 matched pairs (5-year OS, 82.9% versus 82.9%, p = 0.822). Extent of LND was not associated with survival outcome of part-solid lung adenocarcinoma. Furthermore, it had no influence on the prognosis even for solid-dominant lesions by propensity score analysis, which could be selectively omitted based on both preoperative variables including carcinoembryonic antigen or maximum standardized uptake value.