TMPRSS2–ERG fusion transcripts expression in patients referred for prostate biopsy: combining detection in urine and needle rinse material

Objectives The objective of this study was to combine urine and prostate biopsy rinse material (BRM) assays to increase sensitivity for fusion gene detection. Patients and methods A total of 194 patients with suspicion of prostate cancer were prospectively included. Urine samples were collected before or after prostate biopsy, as well as BRM. RT-qPCR was used for the detection of fusion transcripts. A microfocal cancer on biopsy was defined by a single core involved with less than 3 mm of Gleason score 3 + 3 cancer. The association between RT-qPCR and biopsy results was statistically assessed. Results Seven patients were excluded because of insufficient material. Cancer was detected on biopsy in 100 (53 %) patients. Urine alone, BRM alone and both samples were obtained in 155, 164 and 132 patients, respectively. In patients with evidence of cancer on biopsy, a fusion transcript was detected in 63, 55 and 73 % of the cases on urine alone, BRM alone and paired samples, respectively. Fusion gene detection on BRM was only associated with the amount of cancer on biopsy. Urine fusion score had a larger area under the curve than serum PSA ( p = 0.002) and was significantly higher in patients with high Gleason score and significant cancer on biopsy. Assays of paired samples allowed increasing sensitivity in all subgroups of patients. Conclusions TMPRSS2–ERG fusion gene detection may be performed both in the urine and BRM to increase sensitivity. However, only T–E urine score was associated with adverse pathological features.