Identification of a gene-expression signature for predicting lymph node metastasis in patients with early stage cervical carcinoma.

Abstract

Pelvic lymph node metastasis (PLNM) is an important prognostic factor for patients with cervical carcinoma. The objective of this study was to identify a gene-expression signature that could predict PLNM in cervical carcinoma. Eighty-eight women with cervical carcinoma with PLNM (n = 23) and without PLNM (n = 65) were divided randomly into a training group and a test group. An oligonucleotide microarray that contained probes for 1440 human cancer-related genes was fabricated in-house and was used to detect the gene expression profile of cervical carcinoma. The gene expression levels detected in the microarray were verified by quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR). A gene-expression signature for predicting PLNM was developed in patients from the training group, including 11 genes: ribosomal protein L35 (RPL35); thymosin β 10 (TMSB10); tyrosine 3-mono-oxytenase/tryptophan 5-mono-oxygenase activation protein, ζ polypeptide (YWHAZ); biotinidase (BTD); lactate dehydrogenase A (LDHA); glucuronidase β (GUSB); superoxide dismutase 2 (SOD2); nuclear receptor subfamily 3, group C, member 2 (NR3C2); fructosamine 3 kinase (FN3K); x-ray repair cross-complementing 4 (XRCC4); and wingless-type mouse mammary tumor virus integration site family member 2 (WNT2). In the test group, the signature's accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 91%, 90.9%, 93.9%, 83.3%, and 96.9%, respectively, for predicting PLNM. The expression levels of 5 genes in the signature were confirmed by qRT-PCR. A multivariate analysis demonstrated that patients with 11-gene high-risk scores were had a 33-fold increased risk for PLNM compared with patients who had low-risk scores. The 5-year overall and disease-free survival rates for patients who had 11-gene high-risk scores were marginally significantly lower than the rates for patients who had 11-gene low-risk scores (P = .087 and P = .174, respectively). In this study, 11-gene signature for predicting PLNM in cervical carcinoma was identified that may help clinicians in planning therapy for patients with cervical carcinoma.