Refractory Hypomagnesemia and the Proton Pump Inhibitor Question
Two failed rounds of oral magnesium and a new QTc prolongation force a real decision about a drug she's taken for eight years without incident, against a bleed in her history that was genuinely serious.
Carol M., a 67-year-old widow, has lived alone since her husband passed four years ago, and still makes it to her Tuesday bridge club most weeks — though lately the muscle cramps in her legs have made the walk from the parking lot longer than it used to feel. She's been on omeprazole for eight years, started after a peptic ulcer bleed serious enough that she remembers the hospital stay in detail — four units of blood and three days on a monitor before anyone let her go home, a memory that surfaces every time a doctor so much as mentions changing the medication that's kept it from happening again — and has taken it every morning since without a second thought until this month's labs.
Her serum magnesium is 1.1 mg/dL, low, and has stayed low through two separate rounds of oral repletion over the past two months — a pattern, not a single abnormal draw. Proton pump inhibitors are thought to cause this by impairing TRPM6, a transporter the intestine relies on to actively absorb magnesium, which is also why simply supplying more magnesium by mouth often fails: the absorption pathway itself is the part that isn't working. A new ECG obtained for her fatigue picked up mild QTc prolongation, a finding low magnesium is a recognized, direct cause of, and one that raises the stakes on getting her magnesium corrected beyond just resolving her cramps. The FDA's 2011 safety communication on proton pump inhibitors specifically identified long-term use, generally over a year, as a cause of hypomagnesemia that in a meaningful share of reported cases did not respond to oral magnesium supplementation and resolved only after the PPI itself was discontinued — the exact pattern Carol's two failed repletion attempts now reproduce. Her own history complicates a clean answer, though not quite symmetrically. The bleed that put her on omeprazole was real and serious — but it was a single event, eight years back, with nothing since. Her hypomagnesemia is current, has now reproduced itself through two separate repletion attempts, and is the only one of the two still generating findings on her ECG. What the group is weighing is not two live risks of comparable standing; it is one remote and historical against one present and recurring.
At the outpatient follow-up, reviewing the second failed repletion
I'd stop the PPI. The FDA's 2011 safety communication describes her situation almost exactly — long-term use, magnesium that doesn't respond to oral repletion, resolution only after the drug is discontinued. She's already failed two rounds of oral repletion, and now she has a QTc finding that low magnesium directly causes. I don't think continuing the likely cause while treating the effect is going to work any better a third time.
Her ulcer bleed eight years ago was serious — she was hospitalized, and it's the reason she's been on a PPI this long in the first place. Stopping acid suppression entirely isn't a neutral choice for her. I'd rather switch her to famotidine, an H2 blocker, than stop outright — different mechanism, no comparable magnesium-wasting signal in the literature, and it still gives her some acid suppression rather than none while we see whether that alone fixes the magnesium.
You're right that stopping entirely isn't necessary if famotidine is a reasonable substitute — I was reaching for the more drastic fix. I'd support the switch instead of a full stop.
One thing that shouldn't get folded into the drug-switch decision: given her QTc finding, she needs magnesium repleted now, by IV, not another round of oral tablets. Oral absorption is capped and slow, and that's exactly why two outpatient rounds already failed to move the number — this is a separate, more urgent question from which acid-suppressing drug she ends up on.
Agreed: IV magnesium repletion now given the QTc finding, switch from omeprazole to famotidine, and recheck both magnesium and ECG in one week.
Not agreed: whether she needs any acid suppression at all long-term, now that eight years have passed since her bleed, or whether famotidine itself could eventually be tapered. The group left this as a separate future question rather than deciding it today.