A combination of molecular markers accurately detects lymph node metastasis in non-small cell lung cancer patients

Abstract

Occult lymph node metastasis (micrometastasis) is a good prognostic indicator in non-small cell lung cancer (NSCLC) and could be used to direct adjuvant chemotherapy in stage I patients. This study was designed to evaluate molecular markers for detection of occult lymph node metastasis in NSCLC, define the best marker or marker combination to distinguish positive from benign lymph nodes, and evaluate these markers in lymph nodes from pathologically node-negative (pN(0)) NSCLC patients. Potential markers were identified through literature and database searches and all markers were analyzed by quantitative reverse transcription-PCR in a primary screen of six NSCLC specimens and 10 benign nodes. Selected markers were further evaluated on 21 primary NSCLC specimens, 21 positive nodes, and 21 benign nodes, and the best individual markers and combinations were identified. A combination of three markers was further validated on an independent set of 32 benign lymph nodes, 38 histologically positive lymph nodes, and 462 lymph nodes from 68 pN(0) NSCLC patients. Forty-two markers were evaluated in the primary screen and eight promising markers were selected for further analysis. A combination of three markers (SFTPB, TACSTD1, and PVA) was identified that provided perfect classification of benign and positive nodes in all sample sets. PVA and SFTPB are particularly powerful in tumors of squamous and adenocarcinoma histologies, respectively, whereas TACSTD1 is a good general marker for NSCLC metastasis. The combination of these genes identified 32 of 462 (7%) lymph nodes from 20 of 68 (29%) patients as potentially positive for occult metastasis. Long-term follow-up will determine the clinical relevance of these findings.

Fig. 1.

Secondary screen data showing expression profiles of selected markers in 21 primary NSCLC (T), 21 histologically positive lymph nodes (PN), and 21 benign lymph nodes (BN) from the patients without cancer.

Fig. 2.

Fold change of expression value in histologically positive lymph nodes against highest expression value among benign nodes in all markers from secondary screen (A); best combination of three markers from secondary screen (B); and best combination of three markers from validation set (C). Each symbol with different ship and color indicates the different markers. Each histogram (white, benign nodes; green, histologically positive nodes from NSCLCL patient with adenocarcinoma; orange, histologically positive nodes form NSCLC patients with squamous cell carcinoma; blue, histologically positive nodes from NSCLC patients with large cell carcinoma) indicates the highest fold‐change marker among tested markers for this sample.

Fig. 3.

Three‐dimensional view of expression level of PVA, SFTPB, and PVA in combined data from secondary screen data set and validation data set in benign lymph nodes (total 53) and histologically positive nodes (total 59) with different histologic types of primary tumors.

Fig. 4.

Expression of three selected markers in 30 benign lymph nodes and 462 lymph nodes from 68 pN₀ NSCLC patients. The cutoff line is set at the highest observed expression in benign lymph nodes.