Sentinel Lymph Node Genes to Predict Prognosis in Node-Positive Melanoma Patients
- 443 Downloads
Melanoma patients with a single microscopically-positive sentinel lymph node (SLN) are classified as stage III and are often advised to undergo expensive and substantially toxic adjuvant therapy. However, the 5-year survival rate for these patients, with or without adjuvant therapy, varies from 14 to 85 %, representing a heterogeneous biological population with a variable prognosis. We aimed to identify an SLN gene signature to aid in risk stratification of patients with tumor-positive SLNs.
Microarray experiments were performed to screen SLN genes in recurrence (N = 39) versus non-recurrence (N = 58) groups in the training dataset. Quantitative reverse-transcriptase polymerase chain reaction (RT-PCR) assay was applied to confirm the expression of selected SLN genes, which were further verified using an independent validation cohort (N = 30). Area under the receiver operating characteristic curve (AUC) was calculated to evaluate prognostic accuracy of the selected SLN gene panel, and the prognostic value of our SLN gene signature was also compared with the current American Joint Committee on Cancer (AJCC) staging system.
We identified two SLN genes (PIGR and TFAP2A) that provided high prognostic accuracy in SLN-positive melanoma patients (AUC = 0.864). These two SLN genes, along with clinicopathological features, can differentiate the high- and low-risk groups in node-positive melanoma patients in this cohort.
The two SLN genes, when combined with clinicopathological features, may offer a new tool for personalized patient risk assessment.
KeywordsMelanoma Overall Survival Sentinel Lymph Node Sentinel Lymph Node Biopsy Training Dataset
This work was supported by the University of Louisville School of Medicine Collaborative Matching Grant (H. Hao), University of Louisville Clinical and Translational Science Pilot Grant Program Innovative Award (K.M. McMasters), and Melanoma Research Foundation Established Investigator Award (K.M. McMasters). The authors thank Mrs Margaret Abby for her expert assistance with manuscript preparation, and the University of Louisville genomics facility for their expert support of this work. We are grateful to Ms Sherri Matthews at the Department of Surgery, and Dr Andrei Smolenkov at the James Graham Brown Cancer Center Bio-Repository (University of Louisville), for their coordination with the clinical samples.
Hongying Hao, Deyi Xiao, Jianmin Pan, Jifu Qu, Michael Egger, Sabine Waigel, Mary Ann G. Sanders, Wolfgang Zacharias, Shesh N. Rai, and Kelly M. McMasters disclose no potential conflicts of interest.
- 11.McMasters KM, Egger ME, Edwards MJ, et al. Final results of the Sunbelt Melanoma Trial: a multi-institutional prospective randomized phase III study evaluating the role of adjuvant high-dose interferon alfa-2b and completion lymph node dissection for patients staged by sentinel lymph node biopsy. J Clin Oncol. 2016;34:1079–86.CrossRefPubMedGoogle Scholar
- 15.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Statis Soc B. 1995;57:289–300.Google Scholar
- 16.Cary NC. The SAS system V9. Cary, NC: SAS Institute Inc.; 2003.Google Scholar
- 17.Gonen M. Receiver operating charateristics (ROC) curves. In: Proceedings of the thirty-first annual SAS users group international conference; 2006. p. 210–31.Google Scholar
- 19.Cox DR. Regression and life-tables. J R Stat Soc B. 1972;34:187–220.Google Scholar