LBA22 - A prospectively-designed study to determine the association of a 17-gene genomic prostate score with recurrence following surgery for localised pro...

Date 28 September 2014
Event ESMO 2014
Session Genitourinary tumours, prostate
Topics Prostate Cancer
Translational Research
Surgical Oncology
Basic Principles in the Management and Treatment (of cancer)
Radiation Oncology
Presenter Jennifer Cullen
Citation Annals of Oncology (2014) 25 (5): 1-41. 10.1093/annonc/mdu438
Authors J. Cullen1, I. Rosner2, T. Brand3, A. Ali1, Y. Chen1, N. Zhang4, A. Tsiatis4, D. Knezevic5, T. Maddala4, H.J. Lawrence4, P. Febbo5, S. Srivastava1, I. Sesterhenn6, D. McLeod2
  • 1Epidemiologic Research Program, Center for Prostate Disease Research, 20852 - Rockville/US
  • 2Urology, Walter Reed National Military Medical Center, Bethesda/US
  • 3Urology, Madigan Army Medical Center, Tacoma/US
  • 4Oncology Development, Genomic Health, Inc., Redwood City/US
  • 5Oncology Development, Genomic Health, Inc., 94063 - Redwood City/US
  • 6Pathology, Armed Forces Institute of Pathology, Washington DC/US



Molecular markers of tumor aggressiveness can improve risk assessment for men with PCa. A large, racially diverse cohort of men (21% African-American -AA) with PCa in the Center for Prostate Disease Research multi-center national database was used to determine the association of the Oncotype DX® Genomic Prostate Score (GPS) with recurrence following radical prostatectomy (RP) for localized PCa.


Archival biopsy specimens from 431 eligible men treated with RP for NCCN very low, low or intermediate risk PCa from 1990-2011 at 2 U.S. military medical centers were tested to validate the association between GPS and recurrence following RP, and adverse pathology (AP) at RP. Biochemical recurrence (BCR) was defined as 2 successive PSA levels > 0.2 ng/mL or initiation of salvage therapy for a rising PSA. AP was defined as high-grade (primary Gleason pattern 4 or any pattern 5) and/or pT3 disease. RPs were centrally reviewed by a uropathologist (IAS). Cox proportional hazards and logistic regression models were used.


GPS (scale 0-100) was successfully obtained in 402 cases (93%). 62 men (15%) experienced BCR, 5 developed metastases (MR) and 163 had AP at RP. In univariate analysis, PSA, biopsy GS and NCCN risk group were associated with BCR, but race was not. GPS was predictive of time to BCR by univariate analysis (HR/20 GPS units = 2.9; 95% CI: 2.0-4.2; p < 0.001), and after adjusting for age, NCCN risk group, and race ((HR/20 units = 2.7; p < 0.001). The gene groups in GPS most strongly associated with BCR included androgen signaling and stromal response genes. GPS was predictive of time to MR (HR/20 units = 3.8; p = 0.032). GPS was strongly associated with AP - OR/20 units = 3.3 (95% CI: 2.1–5.1; p < .001) and high-grade disease - OR/20 units = 2.5 (95% CI: 1.6–4.0; p < .001), adjusted for NCCN risk group. GPS values were similar between AA and Caucasians, and similarly predictive of disease outcome.


GPS is a significant independent predictor of BCR, MR and AP in men treated with RP for localized PCa. Tumor aggressiveness, as measured by GPS, and clinical outcomes were similar in AA and Caucasians in this patient population with equal healthcare access.


J. Cullen, A. Ali and Y. Chen: I have received research funding from Genomic Health, Inc.; N. Zhang, A. Tsiatis, D. Knezevic, T. Maddala, H.J. Lawrence and P. Febbo: I am a full-time employee and stockholder for Genomic Health, Inc. All other authors have declared no conflicts of interest.