Abstract
OBJECTIVES: The ratio of positive and resected lymph nodes (LN ratio) has been shown to be prognostic in non-small cell lung cancer (NSCLC). Contrary to the LN ratio, calculating the LN log-odds ratio (LN-LOR) additionally considers the total number of resected lymph nodes. We aim to evaluate LN-LOR between positive and resected lymph nodes as a prognostic factor in operable NSCLC. METHODS: Patients with NSCLC who underwent curative intent lobectomy treated at two high-volume centers were retrospectively studied. LN-LOR was dichotomized according to impact on OS and further combined with N descriptors and correlated with clinical variables and survival. RESULTS: 944 patients were included. Cut-off analysis revealed that an LN-LOR of -0.34 significantly discriminated patients according to OS (p < 0.001, chi-squared test 41.26). When combined with N1 and N2 descriptors, LN-LOR low risk (median OS not reached and 83 months) and LN-LOR high-risk patients (median OS 50 and 59 months) had similar survival irrespective of the anatomical location of the positive lymph nodes. Multivariable Cox regression analysis revealed that age (HR 1.02, 95% CI 1.001-1.032), sex (male, HR 1.65, 95% CI 1.25-2.19), histological subtype (HR 2.11, 95% CI 1.35-3.29), pathological stage (HR 1.23, 95% CI 1.01-1.45) and LN-LOR risk groups (low risk, HR 0.48, 95% CI 0.32-0.72) were independent prognostic factors for OS. CONCLUSIONS: This retrospective two-center analysis shows that LN-LOR is significantly associated with OS in resectable NSCLC and might better reflect the biological behavior of the disease, regardless of anatomical lymph node locations. This finding may additionally support the value of extensive LN dissection.