D. Amico, A. V. Whittington, R. Malkowicz, and S. B. , Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer, JAMA, vol.280, issue.11, pp.969-974, 1998.

M. R. Cooperberg, D. J. Pasta, and E. P. Elkin, The University of California, San Francisco Cancer of the Prostate Risk Assessment score: a straightforward and reliable preoperative predictor of disease recurrence after radical prostatectomy, J. Urol, vol.173, issue.6, pp.1938-1942, 2005.

D. F. Gleason and G. T. Mellinger, Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging, J. Urol, vol.111, issue.1, pp.58-64, 1974.

J. Cuzick, G. P. Swanson, and G. Fisher, Prognostic value of an RNA expression signature derived from cell cycle proliferation genes in patients with prostate cancer: a retrospective study, Lancet Oncol, vol.12, issue.3, pp.245-255, 2011.

, Cancer Genome Atlas Research Network. The Molecular Taxonomy of Primary Prostate Cancer, Cell, vol.163, issue.4, pp.1011-1025, 2015.

E. Lalonde, A. S. Ishkanian, and J. Sykes, Tumour genomic and microenvironmental heterogeneity for integrated prediction of 5-year biochemical recurrence of prostate cancer: a retrospective cohort study, Lancet Oncol, vol.15, issue.13, pp.1521-1532, 2014.

S. R. Setlur, K. D. Mertz, and Y. Hoshida, Estrogen-dependent signaling in a molecularly distinct subclass of aggressive prostate cancer, J. Natl. Cancer Inst, vol.100, issue.11, pp.815-825, 2008.

C. E. Barbieri, S. C. Baca, and M. S. Lawrence, Exome sequencing identifies recurrent SPOP, FOXA1 and MED12 mutations in prostate cancer, Nat. Genet, vol.44, issue.6, pp.685-689, 2012.

B. S. Taylor, N. Schultz, and H. Hieronymus, Integrative genomic profiling of human prostate cancer, Cancer Cell, vol.18, issue.1, pp.11-22, 2010.

H. Ross-adams, A. D. Lamb, and M. J. Dunning, Integration of copy number and transcriptomics provides risk stratification in prostate cancer: A discovery and validation cohort study, EBioMedicine, vol.2, issue.9, pp.1133-1144, 2015.

M. Fraser, V. Y. Sabelnykova, and T. N. Yamaguchi, Genomic hallmarks of localized, nonindolent prostate cancer, Nature, vol.541, issue.7637, pp.359-364, 2017.

D. Knezevic, A. D. Goddard, and N. Natraj, Analytical validation of the Oncotype DX prostate cancer assay-a clinical RT-PCR assay optimized for prostate needle biopsies, BMC Genomics, vol.14, p.690, 2013.

N. Erho, A. Crisan, and I. A. Vergara, Discovery and validation of a prostate cancer genomic classifier that predicts early metastasis following radical prostatectomy, PloS One, vol.8, issue.6, p.66855, 2013.

Z. Jia, F. B. Rahmatpanah, and X. Chen, Expression Changes in the Stroma of Prostate Cancer Predict Subsequent Relapse, PLOS ONE, vol.7, issue.8, p.41371, 2012.

S. Irshad, M. Bansal, and M. Castillo-martin, A molecular signature predictive of indolent prostate cancer, Sci. Transl. Med, vol.5, issue.202, pp.202-122, 2013.

A. Chaux, R. Albadine, and A. Toubaji, Immunohistochemistry for ERG Expression as a Surrogate for TMPRSS2-ERG Fusion Detection in Prostatic Adenocarcinomas, Am. J. Surg. Pathol, vol.35, issue.7, pp.1014-1020, 2011.

K. M. Koo, E. Wee, P. N. Mainwaring, and M. Trau, A simple, rapid, low-cost technique for naked-eye detection of urine-isolated TMPRSS2:ERG gene fusion RNA, Sci. Rep, vol.6, p.30722, 2016.

F. Demichelis, S. R. Setlur, and R. Beroukhim, Distinct genomic aberrations associated with ERG rearranged prostate cancer, Genes. Chromosomes Cancer, vol.48, issue.4, pp.366-380, 2009.

M. S. Geybels, J. J. Alumkal, and M. Luedeke, Epigenomic profiling of prostate cancer identifies differentially methylated genes in TMPRSS2:ERG fusion-positive versus fusion-negative tumors, Clin. Epigenetics, 2015.

H. Takeda, Z. Wei, and H. Koso, Transposon mutagenesis identifies genes and evolutionary forces driving gastrointestinal tract tumor progression, Nat. Genet, vol.47, issue.2, pp.142-150, 2015.

K. Choo, T. Chuang, and W. Lin, Evolutionary expansion of SPOP and associated TD/POZ gene family: impact of evolutionary route on gene expression pattern, Gene, vol.460, issue.1-2, pp.39-47, 2010.

S. A. Tomlins, B. Laxman, and S. Varambally, Role of the TMPRSS2-ERG gene fusion in prostate cancer, Neoplasia N. Y. N, vol.10, issue.2, pp.177-188, 2008.

F. C. Hamdy, J. L. Donovan, and J. A. Lane, 10-Year Outcomes after Monitoring, Surgery, or Radiotherapy for Localized Prostate Cancer, N. Engl. J. Med, vol.375, issue.15, pp.1415-1424, 2016.

D. Gasi-tandefelt, J. L. Boormans, and H. A. Van-der-korput, A 36-gene signature predicts clinical progression in a subgroup of ERG-positive prostate cancers, Eur. Urol, vol.64, issue.6, pp.941-950, 2013.

, OncotypeDX-like, and Decipher-like scores, and to low-risk subtypes identified through the reproduction of Irshad et al or Jia et al (Proveri) computational methods (Supplementary Methods). Prolaris, OncotypeDx and Decipher low risk groups were defined by identifying cut-off values that best discriminated patients who had a biochemical recurrence from patients who did not. Resulting cut-offs used for discretization were 3, 31, and 1.7 for Prolaris-like, OncotypeDX-like and Decipher-like scores. Hazard ratios refer to the relative risk of BCR in a multivariate Cox model including ISUP group (1-2 vs 3-4), tumour stage (T2 vs T3-T4), and PSA level (below or above 4ng/mL) from 729 tumour samples. D) Association of CIT36 predictions with BCR-free survival in ERG fusion positive tumours. The 36 genes listed in Supplementary Table S4 were used to build a predictor (CIT36) of S2 subtype among ERG-positive tumours as described in Supplementary Methods. The Kaplan-Meier, S2 subtyping relatively to other molecular approaches. We compared the prognostic significance of S2 subtype to low-risk groups defined by discretization of Prolaris-like