AstraZeneca & Ionis Stumble: Eplontersen Misses Primary Endpoint in Landmark ATTR-CM Trial

The largest transthyretin amyloid cardiomyopathy (ATTR-CM) trial ever conducted just reported its top-line results — and the headline reads "miss." But the clinical story underneath that headline is considerably more complicated, and arguably more important, than the binary outcome suggests.

On July 9, 2026, Ionis Pharmaceuticals and AstraZeneca announced that the CARDIO-TTRansform Phase 3 trial of eplontersen (Wainua) did not meet its primary composite endpoint of cardiovascular mortality and recurrent cardiovascular clinical events through Week 140. In a field where every major readout reshapes the competitive map, this is a significant stumble for one of the most anticipated RNA silencer programs in cardiometabolic medicine.

What the trial was designed to prove — and what changed while it ran

CARDIO-TTRansform enrolled 1,432 patients across 130 sites in 20 countries, randomized 1:1 to eplontersen 45mg subcutaneous injection every four weeks versus placebo, on top of available standard of care. The primary endpoint: a composite of CV mortality and recurrent CV clinical events through Week 140. When enrollment began, the ATTR-CM treatment landscape was relatively sparse. By the time the trial read out, it had been fundamentally transformed.

TTR stabilizers — tafamidis and acoramidis — went from emerging therapies to standard of care during the trial's run. The result: 57% of patients in both arms were already on a stabilizer at baseline, and a further 24% in each arm initiated one during the trial. Eplontersen was effectively being tested as an add-on to a class of drugs that had already demonstrated meaningful mortality and hospitalization benefits on their own.

The signal buried in the subgroup

This is where it gets clinically interesting. In a pre-specified subgroup analysis of patients who received eplontersen without a background stabilizer — monotherapy only — a nominally significant hazard ratio of 0.71 was observed on the composite primary endpoint. A 29% relative risk reduction is not a trivial number in ATTR-CM. In the stabilizer subgroup, no treatment effect was observed at all.

The mechanistic explanation is biologically plausible: TTR stabilizers prevent misfolding of the transthyretin protein. TTR gene silencers like eplontersen reduce TTR production entirely via RNA interference at the liver. In theory, combining a stabilizer with a silencer could be complementary. In practice, this trial suggests that in patients already adequately managed with stabilizers — whose TTR fibrillogenesis is already suppressed — adding a silencer provides no measurable incremental benefit on hard endpoints.

Importantly, the drug still worked mechanistically. Large and sustained reductions in circulating TTR protein were observed, consistent with the silencer class. Multiple secondary, imaging, and biomarker endpoints also favored eplontersen. The biology is sound. The trial design collided with an evolving standard of care.

What this means for the ATTR-CM competitive map

Eplontersen already holds approval for hereditary ATTR polyneuropathy in over 20 countries as Wainua (US) and Wainzua (EU) — that business is intact. The cardiomyopathy indication was the growth opportunity, and it remains elusive.

The competitive beneficiaries are clear: Pfizer's tafamidis franchise and BridgeBio's acoramidis now look even more entrenched as the backbone of ATTR-CM management. Acoramidis showed a statistically significant 34% reduction in CV hospitalizations versus tafamidis in head-to-head data at Heart Failure 2026 — making it the emerging SoC anchor. Alnylam's vutrisiran is the other RNA silencer watching this readout carefully, with its own ATTR-CM data from HELIOS-B already filed.

For Ionis, this is a setback but not an existential one: TRYNGOLZA is in commercial launch, the pipeline is intact, and cash flow breakeven is targeted for 2028. For AstraZeneca, the CM indication expansion narrative is now in limbo pending the ESC full data readout.

The ESC question

Full data will be presented at the European Society of Cardiology Congress in August 2026. That readout will determine whether the monotherapy subgroup signal — HR 0.71, nominally significant, pre-specified — is robust enough to support a label carve-out, a revised regulatory strategy, or simply hypothesis-generating context for next-generation combination trials in stabilizer-naive patients.

The FDA has historically shown willingness to consider pre-specified subgroup analyses in rare and serious disease settings. Whether the agency would entertain a narrower ATTR-CM label for stabilizer-naive patients is the central regulatory question now sitting on the table.

The broader lesson for trial design in fast-moving fields

CARDIO-TTRansform joins a growing list of trials that were designed in one treatment era and read out in another. When background standard of care shifts substantially during a multi-year trial — particularly when the new standard addresses the same molecular target — the trial's comparator arm is no longer what it was at enrollment. This is the structural challenge of running long-duration Phase 3 trials in active therapeutic areas, and it demands the field think more carefully about adaptive designs, biomarker-stratified enrollment, and combination studies run concurrently with monotherapy pivotals.

The drug did not fail the biology. The biology passed. The trial ran into the future.

---

Scientific Validity: 7/10 — Mechanistic data robust, TTR reduction confirmed, secondary and biomarker endpoints favored eplontersen. Primary miss driven by trial design collision with evolving SoC, not drug failure. Monotherapy HR 0.71 is a real signal requiring ESC contextualization.

Clinical Maturity: 8/10 — Approved in hATTR-PN; ATTR-CM indication unresolved pending full ESC data. Regulatory pathway for a narrower label remains theoretically possible.

Quad Label: Binary Gamble / Regulatory Risk — High-maturity asset, but the ATTR-CM label outcome now hinges on subgroup interpretation and ESC readout.