Their work appears in the online edition of Lancet Oncology. The group studied cell cycle progression (CCP) genes that may reflect tumor progression. The expression of the genes came from tumor-derived RNA expression signatures which may fluctuate as cells progress through the cell cycle.
They selected 126 CCP genes from which to study the RNA in 96 commercially available samples and developed a final signature of 21 CCP genes. The signature was robust and reproducible for measuring cell proliferation. Tumor-derived RNA was obtained from radical prostatectomy (RP) and TURP patient samples and underwent reverse-transcription to cDNA for amplification. There was adequate RNA quality for amplification of 90% of RP and 85% of TURP specimens. The CCP score was calculated from 3 replicates of the 31 CCP genes after normalization.
A total of 410 patients who underwent RP were analyzed. They had a median follow-up of 9.4 years and 148 of them (36%) had biochemical recurrence by 10 years after RP. In the final analysis, 366 (89%) of samples were adequate for calculation. A high CCP score was predictive of biochemical recurrence; a 1-unit change in CCP score had a hazard ratio of 1.89. The CCP score weakly correlated with other variables but was higher in patients with positive lymph nodes. The CCP score was prognostic for different Gleason scores, PSA levels and pathological stages. The proportion of patients who had biochemical recurrence over time increased as the CCP score increased. The CCP score was also predictive for death from CaP with an HR=2.99. In multivariate analysis, CCP score and PSA level were most predictive of biochemical recurrence. A multivariate model was developed as a combined risk score to predict biochemical recurrence and included CCP score, T stage, Gleason score, margin status, and PSA. Among TURP patients, the CCP score was the most prognostic in univariate analysis. In this patient cohort, the CCP score correlated more with Gleason score and remained dominant in the multivariate analysis. A combined risk score model was also developed for these patients. The 10-year death rate from CaP increased as a function of the combined risk score. They found the combined score was able to identify disease with a low fatality risk, but did find some Gleason score 6 tumors that were predicted not to have indolent behavior.
Cuzick J, Swanson GP, Fisher G, Brothman AR, Berney DM, Reid JE, Mesher D, Speights V, Stankiewicz E, Foster CS, Møller H, Scardino P, Warren JD, Park J, Younus A, Flake DD 2nd, Wagner S, Gutin A, Lanchbury JS, Stone S
Lancet Oncol. 2011 Mar;12(3):245-255