Can non-performance of radical systematic mediastinal lymphadenectomy be justified in elderly lung cancer patients? An evaluation using propensity-based survival analysis☆

Abstract

The increasing age of the population has raised the importance of determining the minimally required surgical treatment for elderly lung cancer patients. Despite a number of previous studies, the therapeutic impact of a radical mediastinal lymphadenectomy (RLA) associated with a pulmonary resection for lung cancer remains controversial. Herein, we investigated the impact of lymph node dissection on the overall survival for elderly lung cancer patients and assessed whether the non-performance of an RLA could be justified in the surgical treatment for these elderly patients. We analysed the data for 160 patients aged 70 years and older (113 males, 47 females) who underwent curative-intent surgery for non-small-cell lung cancer. They were divided into two groups, according to the method used for the intra-operative mediastinal lymph node dissection, the radical systematic lymphadenectomy (RLA, n=76) and the non-radical lymphadenectomy (NLA, n=94) groups. A Cox proportional hazards model and the Kaplan-Meier method were used for the survival analyses. Propensity-based analyses were also used to reduce the effect of non-randomisation and possible bias in indication of treatment between the two groups. RLAs had no protective effect on mortality; the hazard ratio for the RLA group in comparison to the NLA group was 0.97 (95% confidence interval (CI): 0.32-2.89) in the multivariate analysis and 1.43 (95% CI: 0.42-4.91) in the propensity-based stratifying analysis. The 3-year survival probability was 81.3% (95% CI: 67.1-89.8) for the NLA group, which was marginally better than that of the RLA group (77.5% (95% CI: 63.3-86.8)). There was no significant difference in the overall survival between the two groups (p=0.26). The 3-year survival probability of the NLA group at each quartile of the propensity score also tended to be better than that of the RLA group, which did not show any significant difference. There was no survival benefit shown for RLA associated with pulmonary resections in the present cohort, even in the propensity-based analyses. Although some reports recommend a systematic mediastinal lymphadenectomy for proper staging and better survival, a pulmonary resection with non-performance of radical lymphadenectomy could be an acceptable surgical treatment for the increasing number of elderly lung cancer patients.