Localized prostate cancer death predicted by RNA signature score

January 1, 2012

A composite score calculated from a 46-gene RNA signature independently predicts cancer death among men with clinically localized prostate cancer managed by active surveillance.

Key Points

London-A composite score calculated from a 46-gene RNA signature independently predicts cancer death among men with clinically localized prostate cancer managed by active surveillance, researchers from the Transatlantic Prostate Group reported.

The composite score is determined from RNA levels of 31 genes involved in cell cycle progression (CCP) and 15 housekeeper genes. Gene expression is measured by performing quantitative reverse transcriptase-polymerase chain reaction on mRNA extracted from paraffin-embedded prostate cancer tissue.

Gleason, PSA add to predictive power

In univariate analysis, the hazard ratio (HR) for prostate cancer-specific death was 2.07 for each unit increase in CCP score (p=1.4 x 10-10), Dr. Cuzick reported at the 2011 American Society of Clinical Oncology annual meeting in Chicago. In a multivariate analysis that adjusted for centrally re-reviewed Gleason score, baseline PSA, age, clinical stage, and extent of disease (proportion of positive cores), the HR for death from prostate cancer was 1.67 for each unit increase of CCP (p=1.2 x 10-5). CCP score was a stronger independent predictor of death than Gleason score (p=2.8 x 10-4) or PSA (p= .01). Adding Gleason score and baseline PSA to the CCP score resulted in a model with even greater predictive power.

In analyses stratifying men by baseline PSA and Gleason score, the CCP score independently predicted death from prostate cancer for all PSA subgroups with a level >4.0 ng/mL (4.0-10.0, >10.0-25.0, >25.0-50.0, >40.0-100.0 ng/mL) as well as in men with a Gleason score >7, but not in those with a lower Gleason score.

Dr. Cuzick noted that results of a previously published study also showed the CCP score was a robust predictor of disease outcome in two cohorts of men with prostate cancer (Lancet Oncol 2011; 12:245-55). In the earlier investigation, the RNA expression signature was analyzed in one group of 366 American men using tissue from radical prostatectomy and in a second cohort of 337 men from the United Kingdom conservatively followed after diagnosis of prostate cancer by transurethral resection of the prostate.

The analyses showed the CCP score independently predicted biochemical recurrence after radical prostatectomy and prostate cancer-specific death in the men who underwent active surveillance. As in the current study, the CCP score had stronger predictive accuracy than either PSA or Gleason score in the earlier cohort of men managed conservatively after diagnosis with clinically localized disease.

"The results of the present study provide further evidence of the performance of the CCP score to predict cancer-related death in men with clinically localized prostate cancer and using a very small sample of cancer tissue obtained through needle biopsy to determine the RNA expression signature, which is important because most cases of prostate cancer are currently diagnosed by needle biopsy. Still, further validation in another independent cohort to confirm the robustness of these findings is desirable," Dr. Cuzick said.

The study was done in conjunction with Myriad Genetics, which is developing the test for commercialization. Dr. Cuzick's institution received funds from Myriad to conduct this study, but he does not personally receive funds from the company. Several of the study's co-authors have an employment or leadership position with Myriad.