RTOG 9601 NCT00002874
ForpN0 pT2 R1/pT3 post-RP, detectable PSA 0.2-4.0 ng/ml, salvage RT
HR 0.77
95% CI 0.59-0.99, P=0.04; 12-yr OS 76.3% vs 71.3%
TL;DR12-yr OS 76.3% vs 71.3%, HR 0.77 (0.59-0.99), adding 24mo bicalutamide 150mg to salvage RT in pN0 post-RP pts with PSA 0.2-4.0.
The benefit follows pre-SRT PSA: 12-yr OS HR 0.45 (0.25-0.81) above 1.5 ng/ml and 0.61 at 0.7-1.5, but HR 1.13 (0.77-1.65) below 0.7, where the median entry PSA of 0.6 sat. RT was prostate bed only, 64.8 Gy/36 fx, so this informs hormone addition separately from the pelvic nodal coverage decision.
In a pN0 man after RP heading to salvage RT with PSA above 0.7 ng/ml, this supports adding hormone therapy to bed RT; it does not extend to PSA below 0.7 ng/ml, where the post hoc OS HR was 1.13.
Entry PSA gates the hormone add-on to prostate-bed salvage RT: OS HR 0.45 above 1.5 ng/ml, 0.61 at 0.7-1.5, 1.13 below 0.7. RT was 64.8 Gy/36 fx with nodes omitted in pN0 pts, so the trial leaves the elective pelvic nodal question open.
The drug tested, bicalutamide 150 mg for 24 months, is no longer standard, and gynecomastia reached 69.7% without prophylaxis. Whether GnRH-agonist ADT or a shorter course reproduces the OS HR 0.77 was not tested; the benefit concentrates at entry PSA ≥0.7 ng/ml.
11 details 1 trial watching
Randomized, double-blind, placebo-controlled NRG/RTOG trial; 840 randomized 1998-2003, 760 eligible analyzed (384 vs 376). Median follow-up 13 years among survivors.
Post-RP with lymphadenectomy, pT2 with positive margin or pT3, pN0, detectable PSA 0.2-4.0 ng/ml at least 8 weeks after surgery. No prior RT or chemo; prior hormones limited to short preoperative courses (6.4%). Staging by CT and bone scan.
Bicalutamide 150 mg daily for 24 months vs placebo, starting with RT. Adherence ≥18 months was 69.8% vs 74.7%.
64.8 Gy in 36 fractions of 1.8 Gy to the prostatic bed and membranous urethra, 6-10 MV photons, 2D or 3D planning. Pelvic nodes omitted because all dissections were negative. QA: 64.8% per protocol, 4.4% unacceptable variation.
Primary: overall survival. Secondary: disease-specific death, distant metastases, local progression, non-disease-specific death, second biochemical recurrence, AEs.
Primary met; every disease-control endpoint favored bicalutamide (table). The OS gain is concentrated at entry PSA ≥0.7 ng/ml in post hoc subgroups.
| Endpoint | RT + bicalutamide | RT + placebo | HR (95% CI), P |
|---|---|---|---|
| OS (1° EP) | 76.3% | 71.3% | 0.77 (0.59-0.99), P=0.04 |
| Death from PCa | 5.8% | 13.4% | 0.49 (0.32-0.74), P<0.001 |
| Metastatic PCa | 14.5% | 23.0% | 0.63 (0.46-0.87), P=0.005 |
| 2nd biochemical recurrence | 44.0% | 67.9% | 0.48 (0.40-0.58), P<0.001 |
| PSA at entry | n (%) | Bicalutamide | Placebo | HR (95% CI), P |
|---|---|---|---|---|
| <0.7 ng/ml | 405 (53.3) | 76.8% | 80.7% | 1.13 (0.77-1.65), 0.53 |
| 0.7-1.5 ng/ml | 237 (31.2) | 77.0% | 67.5% | 0.61 (0.39-0.95), 0.03 |
| >1.5 ng/ml | 118 (15.5) | 73.5% | 48.9% | 0.45 (0.25-0.81), 0.007 |
Gynecomastia 69.7% vs 10.9%. Late GU G3 7.0% vs 6.0%; no excess early bladder, bowel or hematologic toxicity, and no excess non-disease-specific death.
SPPORT and RADICALS-HD tested GnRH-based ADT with salvage RT, but neither used a placebo-controlled OS design. The Early Prostate Cancer Program trials of the same drug and dose reported a matching gynecomastia signal (68% vs 8%).
Post-randomization exclusion of 79 ineligible pts departs from strict ITT. Cardiac events were captured only as reported AEs, which risks ascertainment bias. Pre-PSMA staging leaves occult extra-field disease in the cohort.
OS benefit is real but risk-adapted: below PSA 0.7 the subgroup trended the other way. The open questions are whether a modern GnRH agonist reproduces the effect and which duration is enough.
CONSORT flow
Double-blind placebo-controlled RCT, OS primary met at 13y median f/u. Benefit concentrated at PSA ≥0.7 (post hoc); bicalutamide 150mg since superseded by GnRH agonists.
- Does GnRH-agonist ADT reproduce the bicalutamide OS benefit?
- Optimal hormone duration with salvage RT at low PSA active Comparison of HT Concomitant With RT vs RT Alone in Patients With a Detectable PSA After Prostatectomy Phase 2n=120 · primary completion 2022-03 · short degarelix + SRT vs SRT alone, pN0, PSA ≥0.2
- Does PSMA PET staging change who benefits from added hormones?
📚 Sources · 📄 1 paper
The longer read
This is the trial that justified hormone therapy alongside salvage RT, and its weight comes from the endpoint. A double-blind, placebo-controlled design with 13 years of median follow-up and overall survival as the primary is rare in the postoperative setting, and OS HR 0.77 met that bar. The disease-specific endpoints move the same way with larger relative effects (death from prostate cancer HR 0.49, metastases HR 0.63), with no excess non-disease-specific death, so the survival signal is hard to dismiss as chance.
The harder question is whom the average effect describes. Median entry PSA was 0.6 ng/ml, and the post hoc forest plot shows no OS benefit below 0.7 ng/ml (HR 1.13, 0.77-1.65). The effect sits at 0.7-1.5 ng/ml (HR 0.61) and above 1.5 ng/ml (HR 0.45), and PSA was the only factor with a significant interaction. When the 53.3% of pts who entered below 0.7 trend toward no benefit, the headline is telling the reader that hormone therapy at salvage depends on risk and is not a uniform add-on. Modern referral patterns make this sharper: pts now reach salvage RT at lower PSA than in this era, which moves the typical candidate toward the stratum that did not benefit.
Two features of the regimen limit direct transfer. Bicalutamide 150 mg is not what most clinics now give, and the authors acknowledge GnRH agonists replaced it. With no prophylaxis, the gynecomastia rate of 69.7% is a toxicity profile few would accept routinely. Whether two years of GnRH-agonist ADT reproduces the survival effect, or a shorter course is enough, is outside what this trial tested. GETUG-AFU 16 and SPPORT are the natural comparators for short-term GnRH-based therapy with salvage RT, and RADICALS-HD addressed duration. None of them repeats the placebo-controlled OS design, which is why this trial is still the anchor even though its drug is obsolete.
For an RT reader, the radiotherapy arm belongs to its era: 64.8 Gy in 1.8 Gy fractions to the bed alone, 2D or 3D planning, staging by CT and bone scan, pelvic nodes deliberately untreated after negative dissection. The placebo arm's 67.9% second biochemical recurrence at 12 years shows how often bed-only RT at this dose failed. A PSMA-staged population would reveal some of that failure before RT, so part of the bicalutamide effect may reflect treatment of occult disease outside the field. A modern nodal field or PET selection could absorb some of it. That would shrink the absolute benefit without necessarily reversing its direction.
What could make the headline wrong? Excluding 79 ineligible pts after randomization departs from strict intention-to-treat, though the arms stayed balanced. Cardiac events were collected only as adverse events. Neither is large enough to overturn a placebo-controlled survival result. Both argue for reading the 5.0-point absolute OS gain and the NNT of 20 as upper-end estimates for today's lower-risk, better-staged salvage population, not as a number that transfers unchanged.