Intermittent or Continuous Androgen Deprivation in Biochemical Recurrence
A man with a slow-rising PSA after prostatectomy, no detectable metastases, weighs intermittent androgen deprivation against continuous therapy — testing how far a landmark trial's inconclusive result in sicker patients can honestly reach.
D.S., a 66-year-old man, drives long-haul routes for a regional freight company, work he's done for over thirty years and one that keeps him away from home four or five nights a week — a rhythm his wife has learned to plan around and one he isn't eager to disrupt with a treatment schedule that leaves him fatigued or hot-flashing behind the wheel. He underwent radical prostatectomy four years ago for intermediate-risk disease with negative margins and undetectable PSA at his first postoperative check, a result that had let him stop thinking about the cancer entirely for the better part of two years. Eighteen months ago his PSA became detectable and has been rising slowly since, now at 0.6 ng/mL with a doubling time of roughly fourteen months — biochemical recurrence, without any evidence of metastatic disease on conventional or PSMA-PET imaging, and no local recurrence identified on repeat imaging of the prostate bed. That combination, a slow doubling time and a clean scan, is exactly the profile that makes the timing and scheduling of any hormonal therapy a genuinely open question rather than an urgent one.
The decision in front of him is whether to start androgen deprivation therapy intermittently, cycling on and off based on PSA thresholds, or continuously. The largest trial addressing this exact drug-scheduling question, SWOG 9346, could not establish that intermittent therapy was noninferior to continuous therapy for overall survival — its confidence interval exceeded the prespecified boundary, meaning a twenty percent greater risk of death with intermittent therapy could not be ruled out, though too few events accrued to confirm that risk was real either. But SWOG 9346 enrolled men with metastatic hormone-sensitive prostate cancer, a population with a fundamentally different disease burden than D.S.'s slow-doubling biochemical recurrence with no detectable metastatic disease at all. Whether that trial's own inconclusive survival signal, generated in a sicker population, has anything meaningful to say about a man like him is the actual question underneath the scheduling decision.
In clinic, scheduling therapy for a recurrence that may take years
I'd recommend continuous therapy. SWOG 9346 is the largest trial we have on this exact scheduling question, and it could not show intermittent therapy was noninferior for survival — the confidence interval left open the possibility of up to a twenty percent greater risk of death with the intermittent approach. Until there's a trial that clears that bar, I don't think we should be recommending a schedule the best available evidence couldn't confirm is safe.
I'd offer him intermittent therapy as a reasonable, evidence-consistent option, specifically because SWOG 9346 wasn't testing his situation. That trial enrolled men with metastatic hormone-sensitive disease — a meaningfully sicker population than a man with a slow-doubling biochemical recurrence and no detectable metastases at all. The same trial found real quality-of-life benefits with intermittent dosing, and a man who drives for a living, four or five nights a week away from home, has a legitimate stake in not being hot-flashing and fatigued on a schedule that never lets up, especially treating a recurrence that may take years to become clinically meaningful.
Citing SWOG 9346's survival caution as if it applies directly to him treats a trial about metastatic disease as though it studied biochemical recurrence. It didn't. Applying its inconclusive result to a population it never enrolled understates how much the disease context differs.
I think both of you are reading more into SWOG 9346 than the trial itself can honestly support, just in opposite directions. The trial's confidence interval didn't confirm intermittent therapy is unsafe — it simply couldn't rule out a difference, in a population of men with actual metastatic disease. Treating that inconclusive result as a reason to insist on continuous therapy in a man with biochemical recurrence and no metastases overstates the caution's reach. But treating the population mismatch as license to set the survival question aside entirely also overstates what we actually know — we don't have a dedicated noninferiority trial in biochemical recurrence either, we have an absence of the specific caution that applies to sicker patients, which is a different and weaker kind of reassurance.
What I'd actually say to him: the evidence doesn't forbid intermittent therapy, and it doesn't confirm it's equivalent either, particularly outside the population SWOG 9346 actually studied. Given his slow PSA doubling time, the absence of any detectable disease on modern PSMA imaging, and a clear, stated preference tied to his work, I think intermittent therapy is a defensible choice here — but it should be offered as a preference-sensitive decision under real uncertainty, not as a confidently evidence-backed equivalent to continuous therapy.
D.S. chose intermittent therapy, starting a twelve-month induction course with PSA-threshold-triggered cycling thereafter, informed explicitly that this choice rests on the absence of a specific caution rather than on a dedicated trial confirming equivalence in his exact clinical situation.
Not agreed: how the team should counsel the next patient whose biochemical recurrence looks less favorable — a shorter PSA doubling time, or a positive PSMA-PET showing oligometastatic disease closer to SWOG 9346's own population. The urologic oncologist believes the survival caution should carry more weight as a patient's disease features drift closer to that trial's enrolled population; the radiation oncologist and pharmacologist agree in principle but haven't set a specific threshold for where that shift should happen.