A TMEFF2-regulated cell cycle derived gene signature is prognostic of recurrence risk in prostate cancer

Authors

Constantin Georgescu, Oklahoma Medical Research Foundation
Joshua M. Corbin, Oklahoma University Health Sciences Center
Sandra Thibivilliers, Stephenson Cancer CenterFollow
Zachary D. Webb, Oklahoma University Health Sciences Center
Yan D. Zhao, Oklahoma University Health Sciences Center
Jan Koster, University of Amsterdam
Kar-Ming Fung, Oklahoma University Health Sciences Center
Adam S. Asch, Stephenson Cancer Center
Jonathan D. Wren, Oklahoma Medical Research Foundation
Maria J. Ruiz-Echevarria, Oklahoma University Health Sciences CenterFollow

ORCID IDs

0000-0001-7133-9982

Document Type

Article

Date of this Version

2019

Citation

BMC Cancer (2019) 19:423

Comments

© The Author(s).

Open access

https://doi.org/10.1186/s12885-019-5592-6

Abstract

Background: The clinical behavior of prostate cancer (PCa) is variable, and while the majority of cases remain indolent, 10% of patients progress to deadly forms of the disease. Current clinical predictors used at the time of diagnosis have limitations to accurately establish progression risk. Here we describe the development of a tumor suppressor regulated, cell-cycle gene expression based prognostic signature for PCa, and validate its independent contribution to risk stratification in several radical prostatectomy (RP) patient cohorts.

Methods: We used RNA interference experiments in PCa cell lines to identify a gene expression based gene signature associated with Tmeff2, an androgen regulated, tumor suppressor gene whose expression shows remarkable heterogeneity in PCa. Gene expression was confirmed by qRT-PCR. Correlation of the signature with disease outcome (time to recurrence) was retrospectively evaluated in four geographically different cohorts of patients that underwent RP (834 samples), using multivariate logistical regression analysis. Multivariate analyses were adjusted for standard clinicopathological variables. Performance of the signature was compared to previously described gene expression based signatures using the SigCheck software.

Results: Low levels of TMEFF2 mRNA significantly (p < 0.0001) correlated with reduced disease-free survival (DFS) in patients from the Memorial Sloan Kettering Cancer Center (MSKCC) dataset. We identified a panel of 11 TMEFF2 regulated cell cycle related genes (TMCC11), with strong prognostic value. TMCC11 expression was significantly associated with time to recurrence after prostatectomy in four geographically different patient cohorts (2.9 ≤ HR ≥ 4.1; p ≤ 0.002), served as an independent indicator of poor prognosis in the four RP cohorts (1.96 ≤ HR ≥ 4.28; p ≤ 0.032) and improved the prognostic value of standard clinicopathological markers. The prognostic ability of TMCC11 panel exceeded previously published oncogenic gene signatures (p = 0.00017).

Conclusions: This study provides evidence that the TMCC11 gene signature is a robust independent prognostic marker for PCa, reveals the value of using highly heterogeneously expressed genes, like Tmeff2, as guides to discover prognostic indicators, and suggests the possibility that low Tmeff2 expression marks a distinct subclass of PCa.