Abstract 3559
Background
We have previously shown that a 17-gene, biopsy-based RT-PCR assay (Oncotype Dx® Genomic Prostate Score™, GPSTM) is a strong independent predictor of 1) adverse pathology (defined as surgical GS > 4 + 3 and/or non-organ confined disease) at radical prostatectomy (RP), and 2) biochemical recurrence after RP in men with biopsy GS 6 and 7 prostate cancer. We performed a more refined analysis of the association between GPS and upgrading from biopsy to surgery.
Methods
Patients in a prior clinical validation study of the GPS assay were assigned a biopsy Gleason sum (BxG) and pathologic ISUP 2014 surgical Gleason grade (pGS) by a single uropathologist (IS) in a blinded fashion. We performed logistic regression analyses to model the association of GPS with Gleason upgrading (GU) defined as (i) Any downgrade, (ii) Any increase in GS, (iii) BxG 6 to pGS 3 + 4, (iv) BxG 6 to pGS 4 + 3, and (v) BxG 7 to pGS 8-10, all versus no change in GS. GPS values, as determined from diagnostic biopsy tissue, are reported on a scale of 0-100, and predictions of GU were modeled in 20-unit GPS increments.
Results
Among 395 patients, median patient age at RP and GPS score was 62 years and 30.3, respectively. There were 249 patients (63%) who had no change in GS from biopsy to surgery, 45 patients (11%) were down-graded (BxG 7 to pGS 6) and 101 patients (31%) were upgraded (85 BxG 6 to pGS3 + 4, 10 BxG 6 to pGS 4 + 3, and 6 BxG 7 to pGS 8-9). GPS was a significant predictor of any GS upgrade, BxG 6 to pGS 3 + 4, and BxG 7 to pGS 8-9 (see Table).Table: 837P
| BxG vs pGS | OR/ per 20 units in GPS | 95% CI of OR | P value | |
|---|---|---|---|---|
| Any GS Downgrade | 1.00 | 0.58 | 1.74 | 0.978 |
| Any GS upgrade | 1.77 | 1.20 | 2.60 | 0.004 |
| bGS6 to pGS3 + 4 | 1.61 | 1.06 | 2.41 | 0.027 |
| bGS6 to pGS 4 + 3 | 2.46 | 0.90 | 6.73 | 0.079 |
| bGS 7 to pGS 8-9 | 4.49 | 1.32 | 15.533 | 0.017 |
Referent in all models is no change in GS.
Conclusions
GPS is a significant predictor of the likelihood of Gleason score upgrading from biopsy to RP. Incorporation of GPS results into treatment decision making can improve risk assessment for newly diagnosed patients with clinically low-risk prostate cancer.
Clinical trial identification
Legal entity responsible for the study
GHI & CPDR.
Funding
Genomic Health, Inc.
Editorial Acknowledgement
Disclosure
H.J. Lawrence: Consultant, stock owner: GHI. All other authors have declared no conflicts of interest.