Antiarrhythmic Choice in Structural Heart Disease: Two Shocks in Six Weeks
Two ICD shocks in six weeks for sustained monomorphic VT, on top of already-optimized heart failure therapy — the harder question isn't whether he needs an antiarrhythmic added, it's which one his QTc and his kidneys can actually tolerate.
R.K., a 67-year-old man, moved into his daughter's converted basement apartment eighteen months ago, not quite a year after his wife died and the house they'd shared for three decades became more than he wanted to manage alone. He was still recovering from that loss more than he let on, his daughter told the admitting resident, when the first shock came — mid-afternoon, no warning, just the device firing and R.K. on the kitchen floor before he understood what had happened.
His cardiac history goes back four years, to an anterior STEMI treated with primary PCI that left him with an ejection fraction of 32% despite maximally tolerated GDMT, in place for over a year. His dual-chamber ICD, placed as primary prevention eighteen months after the infarct, had never fired until six weeks ago. Since then it has fired twice for sustained monomorphic VT, most recently yesterday, both times with antitachycardia pacing failing to terminate the rhythm before the shock delivered. He also carries type 2 diabetes, managed on metformin, and stage 3b chronic kidney disease — creatinine clearance 44 mL/min on his most recent labs — neither of which has been the center of any conversation about him until this admission, when both suddenly matter to a decision nobody has made yet.
Two shocks in six weeks is a real signal, not statistical noise, and the panel convened this admission isn't debating whether he needs an antiarrhythmic added to his regimen — GDMT and the device alone have already been tried and have already failed to prevent this. What's contested is which drug, and the reason it's contested is that his own history closes off the option that would otherwise be simplest. A Class Ic agent would be the more mechanistically direct choice against a well-defined re-entrant scar, but the CAST trial settled that argument for anyone with structural heart disease three decades ago, and nothing since has meaningfully reopened it. What's left are the Class III agents, and here the case gets harder rather than easier: his admission ECG already shows a QTc of 462 ms — not dramatically abnormal, but high enough that adding a drug whose entire mechanism is further potassium-channel blockade is not a decision anyone in the room wants to make casually, and his creatinine clearance means whichever Class III drug is chosen will accumulate faster than it would in a patient with normal renal function, raising torsadogenic risk further still.
At the bedside, after the second shock
Load him on amiodarone. Two appropriate shocks in six weeks with ATP failing both times means his current regimen has already been tested and has already failed — the only real question is which drug does the best job of not letting there be a third one, and on that specific question amiodarone has the strongest evidence behind it in exactly this population: reduced ICD shocks more than sotalol or beta-blockade alone in trials built for this comparison.
I know how that sounds against a QTc of 462, but amiodarone's QT effect isn't the same kind of risk as sotalol's. It blocks multiple channels, not just IKr in isolation, and that broader, more homogeneous effect on repolarization is exactly why amiodarone's real-world torsades rate stays low even at QT intervals that would worry me on almost any other drug. The QTc number alone doesn't tell you which drug it came from, and that distinction matters here.
I'm not disputing the acute torsades comparison — it's real, and it's why I'm not arguing against starting amiodarone tonight. What I want on the table before anyone treats this as settled is that amiodarone's risk doesn't end when the QTc stabilizes. Thyroid dysfunction, pulmonary fibrosis, hepatotoxicity, corneal deposits — those show up on a years-long clock, not a this-admission clock, and R.K. is 67 with no other life-limiting diagnosis. If he's on this drug for a decade, that's a decade of cumulative-dose risk we're accepting for a problem sotalol might control adequately with a cleaner long-term profile.
I'll say the counterargument against my own position plainly, because it's real: at this CrCl and this QTc, sotalol isn't a casual substitution either — renal clearance means it accumulates here faster than in a patient with normal kidneys, and starting it would mean stopping carvedilol to avoid stacking two beta-blocking drugs, which gives up a mortality-benefit agent to make room for one that doesn't carry the same evidence in HFrEF. That's not a free trade, and I'm not pretending it is.
Whichever of you turns out to be right in a year, neither drug gets started tonight on a fixed, empiric dose. CrCl 44 means renal-adjusted dosing, telemetry, and serial QTc and electrolyte checks are not a preference, they're a requirement, for amiodarone's oral loading phase as much as for sotalol if it ever gets started later. And there's one piece of this that doesn't require either of you to concede anything: his potassium and magnesium are already low-normal. Repleting them to the middle of the normal range costs nothing, carries no tradeoff, and lowers his torsades risk regardless of which drug wins this argument. That part we can just do.
Agreed within the hour: admit to a telemetry bed, replete potassium and magnesium to the middle of the normal range before anything else, and begin an amiodarone load — the option all three could accept immediately, since it doesn't require stopping carvedilol and doesn't carry sotalol's degree of acute torsadogenic risk at this creatinine clearance. Class Ic agents were never seriously on the table; that part of the reasoning was settled before anyone spoke.
Not agreed, and left for the team following him after discharge:
Amiodarone continues at maintenance dose, with thyroid and liver function checked at six months per routine surveillance — the toxicity conversation moves to a slower clock than the one that mattered this week.
Amiodarone is reduced or reassessed, and referral for VT ablation — raised but not pursued this admission — becomes the primary plan rather than a backup one.
The electrophysiologist and the pharmacologist left holding genuinely different defaults for what happens at the three-month follow-up; the nephrologist's contribution was making sure neither default got acted on before the numbers that should decide it were actually in hand.