Acetazolamide for Periodic Paralysis Before the Genotype Is Known
A single patient with a clear family history of episodic weakness and a genetic panel still weeks from returning. The disagreement is whether to treat the likely diagnosis now or wait for confirmation of a gene that could make the same drug worse, not better.
S.P., a 26-year-old graphic designer, has learned to recognize the pattern by now — a heavy pasta dinner, a weekend spent resting after a hard week at the gym, and by the next morning his legs feel like they belong to someone else, weak enough some episodes that he's had to call in sick rather than risk a fall on the stairs. The pattern started in adolescence and has recurred several times a year since, and his father and paternal uncle both describe strikingly similar episodes, neither of them ever formally diagnosed. A witnessed attack last month caught his serum potassium at 2.9 during the weakness — low enough, and timed to the episode closely enough, to support hypokalemic periodic paralysis as the working diagnosis. Genetic testing for the two genes most commonly implicated, CACNA1S and SCN4A, was sent at that visit and won't return for another three to four weeks.
The treatment question can't wait comfortably for that result, and it can't be resolved by exam or history alone either. Acetazolamide is first-line prophylaxis for CACNA1S-related hypokalemic periodic paralysis, the more common of the two genotypes and the one S.P.'s presentation most resembles. But a real, described subset of patients with SCN4A-related disease paradoxically worsens on acetazolamide rather than improving — and nothing in S.P.'s clinical picture, or his family's undiagnosed history, distinguishes which gene is actually responsible before the panel comes back. Dichlorphenamide, a related carbonic anhydrase inhibitor, carries randomized placebo-controlled evidence of its own — the HYP HOP trials, reported by Sansone and colleagues in 2016, which lowered attack rates in hypokalemic periodic paralysis and were run across both the hypokalemic and hyperkalemic forms rather than being stratified by gene — and sits as a real alternative to consider while the answer is still pending.
Treating a genotype the panel hasn't confirmed yet
I'd start acetazolamide now rather than wait three to four weeks for genetic confirmation. Most hypokalemic periodic paralysis is CACNA1S-related, which is exactly the genotype acetazolamide works best for, and his presentation — carbohydrate-triggered, post-exertional rest attacks, documented hypokalemia during an episode — fits that picture closely. Untreated attacks carry real risk, falls and potential respiratory involvement in severe episodes, and I don't think making him wait a month for a test result before offering any prophylaxis serves him well.
The base-rate argument is reasonable, but I want to name the specific risk it's trading against: a described subset of SCN4A-related hypokalemic periodic paralysis patients paradoxically worsens on acetazolamide rather than improving, and nothing in his clinical presentation or his family's own undiagnosed history rules that genotype out. We genuinely can't tell from the bedside which gene is responsible.
I'd hold off on acetazolamide specifically and consider dichlorphenamide instead if prophylaxis feels urgent — the HYP HOP trials, Sansone and colleagues in 2016, showed a real reduction in attack rate in hypokalemic periodic paralysis, and enrolled by clinical phenotype rather than by genotype, so prescribing it doesn't require us to bet on which gene he carries before the panel tells us.
I'd start acetazolamide now, not blindly, but under an explicit plan: his family history of two affected relatives with similar, undiagnosed episodic weakness meaningfully shifts the odds toward a heritable channelopathy consistent with the more common CACNA1S pattern, even without a confirmed genotype yet.
I'd set a clear stop-rule with S.P. directly — any new or worsening symptoms after starting, not just an absence of improvement, prompts immediate discontinuation and a switch to dichlorphenamide while we wait for the panel, rather than assuming an initial trial is automatically safe to continue.
Resolved: acetazolamide started under a monitored, stop-rule trial, with S.P. explicitly instructed on which symptoms mean call immediately rather than wait for his next appointment. Genetic testing continues in parallel.