Hydrochlorothiazide After NOSTONE: A Hypercalciuric Stone Former
NOSTONE overturned decades of thiazide practice for stone prevention — including, unexpectedly, in the hypercalciuric patients the drug was supposed to help most.
Renee T., a 46-year-old woman, has run competitively since her twenties and is currently training for her fourth marathon, a schedule that has her covering fifty to sixty miles most weeks regardless of weather. Her third calcium oxalate stone in four years passed on its own six weeks ago, and this visit is the follow-up she scheduled herself after her prior urologist mentioned, without much elaboration, that a water pill might help. A 24-hour urine collection done two weeks ago, on a training day representative of her usual routine, shows a calcium excretion of 410mg — well above the 250mg/day ceiling conventionally applied in women, and more than double the >200mg per 24 hours that NOSTONE itself used to define hypercalciuria when it enrolled. Her citrate and oxalate are both within normal range, and her urine volume is generous at 2.8 liters, consistent with the fluid intake her training already demands.
The reflexive next step for a hypercalciuric recurrent stone former has been a thiazide diuretic for decades — reduce distal tubular calcium excretion, reduce stone risk. But that reflex is no longer uncomplicated. The NOSTONE trial, published in 2023, randomized 416 patients with recurrent calcium stones to placebo or one of three hydrochlorothiazide doses and found no dose-response relationship on its composite recurrence outcome; hypercalciuria was not even a requirement for enrollment, but 63% of the trial's patients had it anyway, and the drug still showed no protective effect in that subgroup. Renee's own excretion, at 410mg, sits well beyond the threshold that qualified 63% of NOSTONE's patients as hypercalciuric — which is either a reason her physiology might respond differently than the trial's average patient, or simply a more severe version of the same population the trial already tested and found nothing in, depending on which argument the room finds more persuasive. The trial's secondary finding adds a further wrinkle: radiologic recurrence, tracked separately from symptomatic recurrence, trended lower at the higher doses even though the difference didn't reach statistical significance — a hint the drug-class effect on stone formation itself may not be entirely absent, just too small or too diluted across the trial's broader population to register clearly on the composite outcome.
In clinic, six weeks after her third stone
NOSTONE was built to answer exactly this question. Four hundred and sixteen patients, three doses up to 50mg tested against placebo, and no dose-response signal on the composite outcome — and critically, 63% of the enrolled patients had hypercalciuria at baseline, by a threshold her own number clears twice over, and the drug still didn't separate from placebo in that group. I don't think we should start a drug NOSTONE specifically tested and specifically found didn't work, just because her number is higher than the trial's average.
I agree NOSTONE is the best evidence we have, and I'm not arguing it's flawed.
But her 410mg/day sits well past what a typical NOSTONE enrollee excreted, and the physiologic mechanism — thiazides reducing distal tubular calcium reabsorption — was never the part of this that was in question. A negative population-level trial doesn't mechanically rule out a real effect in a patient whose baseline sits at the tail of the distribution the trial actually studied. I'd try it, with a follow-up 24-hour collection in three months to see whether her own number actually moves, rather than assuming it won't because the average patient's didn't.
Before either of you starts or withholds anything, I want to know what she's actually eating on a training week. Thiazide's calcium-lowering effect is blunted by high dietary sodium through an independent pathway, and NOSTONE's dietary counseling was standardized, not individually enforced — a lot of endurance athletes eat sodium aggressively to manage cramping. If her intake is high, that alone could explain a lot of hypercalciuria that has nothing to do with whether a thiazide would help her specifically.
And I'd want the hyponatremia conversation had out loud before she leaves, not filed under general side effects. She is running fifty to sixty miles a week and already putting out 2.8 liters a day. Thiazides impair free-water excretion, and that is precisely the physiology that turns a hot long run with aggressive fluid replacement into exercise-associated hyponatremia. NOSTONE's population was a median-49-year-old cohort that was four-fifths male and not training for marathons; whatever it does or doesn't tell us about her stones, it tells us nothing about this risk in her.
Agreed: a trial of low-dose hydrochlorothiazide alongside a structured look at her dietary sodium, with a repeat 24-hour urine collection in three months as the actual test of whether either intervention is doing anything in her specific case.
Not agreed: the nephrologist remained skeptical that an individual trial makes sense at all after a trial this size and this well-designed found nothing, and agreed to the plan mainly to settle the question concretely for this one patient rather than because the population-level evidence had changed his own view.