Intravesical Sequencing Under a Sustained BCG Shortage
A BCG-naive man with high-grade Ta bladder cancer, caught in an active national BCG shortage. The disagreement isn't about whether BCG works — it's about what to do when the drug's own evidence base and the shortage's actual fix both sit just out of reach.
R.D., a 61-year-old man, has run the hardware store his father opened in 1988 for the last nine years, six days a week on a concrete floor that he blames for a knee that clicks more than it used to, and he has already told the front-desk staff he isn't interested in anything that would keep him off that floor for long. He noticed blood in his urine on a Tuesday closing shift three months ago, assumed it was nothing, and only came in when it happened again two weeks later. Cystoscopy found a 2.5cm papillary lesion; resection pathology returned high-grade Ta urothelial carcinoma, no lamina propria invasion, no carcinoma in situ elsewhere on random biopsies. He has never received intravesical therapy of any kind — this is his first diagnosis, and BCG-naive high-grade Ta is exactly the population BCG's own outcome data was built on, decades of recurrence and progression follow-up behind a single well-defined regimen.
The problem is not his tumor. It is that his urology practice's BCG allocation this quarter is roughly a third of what it was two years ago, a supply constraint that has persisted long enough that AUA's own shortage guidance — one-third-dose induction, prioritized to patients at higher risk of progression — is now the de facto standard many practices apply by default rather than the stopgap it was written as. R.D. sits in an ambiguous priority tier: high-grade disease qualifies him for BCG under any allocation scheme, but he is also exactly the kind of BCG-naive, single-tumor patient whose recurrence-free survival on sequential intravesical gemcitabine and docetaxel has looked comparable to BCG's own numbers in the single-institution retrospective series practices have accumulated since the shortage began — the largest of them, Packiam and colleagues' 312-patient comparison in JAMA Network Open, describing patients treated between 2011 and 2021, a span that begins well before the current shortage rather than inside it — without needing to touch the constrained supply at all. Whichever way this goes, someone else's allocation changes as a direct consequence of his — a one-third dose given to him is a dose unavailable to the next patient on the list.
In clinic, allocating a scarce drug
One-third-dose BCG, full six-week induction, is what I'd start today. This isn't a drug we're choosing to under-treat him with out of convenience — it's the AUA's own published response to a national shortage that has now lasted long enough that most of us are living inside it rather than waiting it out. BCG's recurrence and progression data in exactly his profile, BCG-naive high-grade Ta, is the deepest evidence base intravesical therapy has anywhere. A reduced dose is a real compromise, but reduced BCG exposure is not the same as no BCG exposure, and I'd rather give him some of the best-validated drug we have than none of it.
If supply were normal I wouldn't be having this conversation at all — full-dose induction wouldn't be in question. The argument here is entirely about how to allocate a real shortage, not about whether BCG deserves its first-line status.
I'd rather take him off BCG entirely and start sequential gemcitabine and docetaxel, full dose, six-week induction. Packiam and colleagues' 2023 cohort of 312 treatment-naive high-risk NMIBC patients — 174 on BCG, 138 on gem-doce — found gemcitabine-docetaxel produced better high-grade recurrence-free survival than BCG, with a lower rate of induction-course discontinuation. I'll be honest about its limits before you raise them: it's one institution, it's retrospective, and it spans 2011 to 2021, so I can't tell you what dose the BCG arm was actually getting in any given year. What I can tell you is that a full dose of the available drug held up against BCG as that institution really administered it, which is more than anyone can say for one-third dosing.
The BCG track record you're citing is mostly a full-dose track record. Once we're both talking about a compromised regimen, the fair comparison isn't reduced BCG versus historical full-dose BCG, it's reduced BCG versus full-dose gem-doce — and on that comparison, I don't think one-third dosing wins.
I want to name something neither position is quite reckoning with. SWOG S1602 reported this year — the largest, most current randomized trial actually built around the BCG shortage — and it tested neither of the two regimens on the table. It randomized BCG-naive high-grade patients to full-dose TICE, full-dose Tokyo-172, or intradermal-primed Tokyo-172, and found Tokyo-172 noninferior to TICE for five-year high-grade recurrence-free survival, sixty-four percent versus fifty-eight percent, with comparable CIS response rates. That is a real, rigorous answer to the shortage — but its answer is strain diversification, not dose reduction, and Tokyo-172 isn't commercially available in the United States. So the trial that actually validated a fix for this exact problem validated a fix he can't receive.
That leaves the two regimens actually in front of him unvalidated against each other by anything randomized. One-third-dose induction has never been tested against full-dose BCG in a controlled trial — it's a rationing convention, not a demonstrated noninferior regimen. The gem-doce cohort data is real and reassuring, but it's retrospective, and S1602's own grade 3 toxicity signal with an unfamiliar strain is a reminder that a new regimen adopted under supply pressure can carry costs a shortage-era comparison doesn't always capture cleanly. I don't think this argument resolves to a clean answer today; I think it resolves to picking the imperfectly-evidenced option honestly, rather than borrowing S1602's reassurance for a regimen it never tested.
Agreed: full-dose sequential gemcitabine and docetaxel, six-week induction, with cystoscopic surveillance at three months per standard high-grade Ta follow-up. The uro-oncologist's allocative argument carried the room more than the pharmacologist's critique alone would have — freeing a one-third-dose BCG allocation for a patient whose profile has no comparably-evidenced chemotherapy alternative was a concrete, immediate reason to move, not just a theoretical preference.
Not agreed: whether this choice should generalize to the practice's next BCG-naive high-grade Ta patient, or whether it was specific to R.D.'s single-tumor, no-CIS profile being exactly what the gem-doce cohort data actually describes. The urologist wants BCG allocation decided case-by-case against each patient's own risk features going forward, not defaulted to gem-doce as a new practice-wide first line; the pharmacologist's closing point — that Tokyo-172's validated answer sits just outside reach — was left on the table as a reason to revisit this decision the moment that strain becomes available here.