Sialyl Lewis X as a predictor of skip N2 metastasis in clinical stage IA non-small cell lung cancer

Abstract

Background: Radical segmentectomy has been performed for small-sized non-small cell lung cancer (NSCLC). However, underestimation of mediastinal lymph node metastasis in the absence of hilar or interlobar metastasis (skip N2) affects surgical strategy. Our aim was to investigate preoperative and intraoperative predictors of skip N2 in clinical stage (c-stage) IA NSCLC.

Methods: From 1998 to 2011, 279 patients (155 men and 124 women) with c-stage IA NSCLC (230 pN0, 17 pN1, 12 skip N2, 20 non-skip N2) underwent systematic lobectomy (R0 resection) at our institute. We compared preoperative serum concentrations of carcinoembryonic antigen, cytokeratin 19 fragment, sialyl Lewis X (SLX), and pre- and intraoperative clinicopathological features of pN0 and skip N2 patients. Receiver operator characteristic (ROC) curve analysis was performed to distinguish between the two patient groups.

Results: The 5-year survival rate of skip N2 patients was 78.6%, higher than that of non-skip N2 patients (44.9%), and not significantly different than that of pN0 (86.7%) or pN1 patients (82.4%). The mean serum SLX concentration in skip N2 patients (28.0 U/ml) was elevated compared to that in pN0 patients (22.9 U/ml). In ROC analysis of SLX, the area under the curve was 0.710, and the optimal cut-off value was 21.4 U/ml (sensitivity, 91.7%; specificity, 51.7%). In multivariate analysis, SLX was an independent predictor of skip N2 in patients with c-stage IA NSCLC (odds ratio, 9.43; p = 0.006).

Conclusions: Skip N2 metastasis is common in patients with c-stage IA NSCLC with high serum SLX, and lobectomy with complete dissection of hilar and mediastinal lymph nodes should remain the standard surgical procedure for these cases.

Figure 1

Kaplan-Meier curves for postoperative overall survival of 230 pathologic (p)N0 patients, 17 pN1 patients, 12 skip N2 patients, and 20 non-skip N2 patients. The prognosis of skip N2 patients was significantly better than that of non-skip N2 patients (P = 0.008). There was no significant difference between the prognosis of skip N2 and pN0 or pN1 patients (P = 0.801, P = 0.985, respectively). pN0: pathological N0; pN1: pathological N1.

Figure 2

Percentile distribution of serum sialyl Lewis X (SLX) concentrations in pathological (p)N0, pN1, skip N2, and non-skip N2 patients with clinical stage IA non-small cell lung cancer (NSCLC). Each box contains the variable distribution between the 25th and 75th percentiles, with median value indicated with a line in the box. The bars extending above and below the box indicate the 90th and 10th percentiles, respectively. The mean SLX concentration was significantly higher in the skip N2 patients compared to that in the pN0 patients (P = 0.015).

Figure 3

Receiver operating characteristic curve for serum sialyl Lewis X (SLX) in pathological (p)N0 and skip N2 patients. The area under the curve (AUC) = 0.710. The calculated optimal cutoff point for serum SLX concentration was 21.4 U/mL (Youden index = 0.433), and the sensitivity was 91.7% (11 of 12 skip N2 patients), whereas the specificity was 51.7% (109 of 211 pN0 patients).