IPM Take
CARDIO-TTRansform is more interesting than a simple failed trial because the biology appeared to work while the clinical endpoint did not. Eplontersen, an antisense oligonucleotide designed to reduce hepatic production of transthyretin, produced sustained suppression of circulating TTR, yet the overall trial showed no statistically significant reduction in cardiovascular mortality and recurrent cardiovascular events compared with placebo.
The key may lie in treatment context. Fifty-seven percent of participants were already receiving a TTR stabiliser at baseline, predominantly therapies designed to prevent transthyretin from misfolding, and another 24% initiated stabiliser treatment during the study. In a prespecified analysis, patients who were not receiving a stabiliser at baseline appeared to benefit from eplontersen, while those already receiving one did not gain additional benefit from adding the silencer. A subsequent secondary analysis reported a statistically significant interaction between baseline stabiliser use and treatment effect.
That makes the policy lesson uncomfortable but useful. Precision medicine cannot be reduced to matching a molecular mechanism with a disease. It also has to determine when that mechanism adds enough incremental value on top of existing treatment to justify another lifelong therapy, another cost and another layer of clinical complexity.
Executive Summary
CARDIO-TTRansform was the largest ATTR-CM trial conducted to date, enrolling 1,432 patients across 130 centres in 20 countries with either wild-type or hereditary transthyretin amyloid cardiomyopathy. Participants received eplontersen 45 mg or placebo every four weeks for 140 weeks in addition to contemporary standard care.
The primary endpoint, a cumulative composite of cardiovascular death and recurrent cardiovascular clinical events, was not met. There were 381 primary endpoint events among 210 patients receiving eplontersen and 392 events among 231 receiving placebo, producing a rate ratio of 0.89, 95% CI 0.73 to 1.09, P=0.28. Cardiovascular deaths numbered 74 with eplontersen and 70 with placebo. Serious adverse events occurred in 57.8% and 59.4% respectively.
However, baseline stabiliser therapy appeared to modify the result. Among patients not receiving a stabiliser at enrolment, the rate ratio for the primary endpoint was 0.71, 95% CI 0.54 to 0.93. Among those already receiving a stabiliser, it was 1.14, 95% CI 0.85 to 1.53, with an interaction P value of 0.017 in the published secondary analysis.
The findings do not establish eplontersen monotherapy as a new standard for ATTR-CM, particularly because the overall Phase III trial was negative. They do, however, weaken the assumption that combining a TTR silencer with a stabiliser will necessarily produce additive clinical benefit.
Why it matters
- HTA bodies: A molecularly targeted therapy should not receive additional value simply because it acts through a different mechanism. If eplontersen advances in ATTR-CM, assessment will need to distinguish its potential value as an alternative to stabilisers from its value as an add-on, because CARDIO-TTRansform suggests those may be very different propositions.
- Payers: ATTR-CM therapies are long-term and potentially cumulative. Evidence that two mechanistically complementary treatments do not necessarily produce additional outcome benefit makes treatment sequencing and combination reimbursement a central policy issue rather than a purely clinical choice.
- Industry / innovation partners: The trial raises the development bar for the next generation of TTR silencers. Demonstrating profound target suppression may no longer be enough in a disease where patients increasingly receive effective background therapy. Future programmes will need to identify the phenotype, treatment history and disease stage in which a new mechanism provides incremental benefit.
Transthyretin amyloid cardiomyopathy has become one of the clearest examples of precision medicine entering mainstream cardiology. The disease is driven by misfolded transthyretin protein accumulating as amyloid in the myocardium, and modern treatments increasingly intervene directly in that biological pathway rather than treating heart failure symptoms alone.
There are, however, different ways to attack the same pathway. TTR stabilisers such as tafamidis are designed to keep the transthyretin tetramer from dissociating and misfolding, while silencers such as eplontersen reduce production of transthyretin at its hepatic source. The biological logic of combining the two is attractive: make less protein and stabilise what remains.
CARDIO-TTRansform tested whether that logic translated into better outcomes.
The answer, at least in the overall population, was no.
Eplontersen produced the expected pharmacodynamic effect and substantially suppressed circulating TTR, yet cardiovascular death and recurrent cardiovascular events were not significantly reduced over 140 weeks. The primary rate ratio of 0.89 was directionally favourable but statistically compatible with no benefit.
That distinction is important for personalised medicine because target engagement is not the same as patient benefit. A therapy can successfully alter the molecular process it was designed to modify without producing enough additional clinical effect to matter once the patient is already receiving effective treatment.
The background therapy may matter as much as the new drug
CARDIO-TTRansform was conducted in a rapidly changing treatment environment. At baseline, 57% of participants were already taking a TTR stabiliser, and another 24% initiated one during follow-up.
The prespecified subgroup analysis therefore becomes central to interpreting the result. Among people who were not taking stabilisers when they entered the trial, eplontersen was associated with fewer cardiovascular deaths and recurrent cardiovascular events, with a rate ratio of 0.71. In participants already receiving stabiliser treatment, there was no apparent additional benefit, with a rate ratio of 1.14. The interaction between baseline stabiliser status and treatment effect was statistically significant in the subsequently published analysis.
This should not be rewritten as proof that eplontersen works as monotherapy and fails in combination. The overall trial missed its primary endpoint, and subgroup findings require appropriate caution even when prespecified. They do, however, provide evidence that treatment effect may be heterogeneous rather than uniform.
That is where the personalised-medicine story becomes more consequential than the corporate pipeline story highlighted by the initial reporting.
ATTR-CM is no longer a therapeutic vacuum in which every biologically active drug is tested against minimal background care. Tafamidis has already demonstrated reductions in mortality and cardiovascular hospitalisation compared with placebo, meaning new therapies increasingly have to prove what they add in a population whose baseline prognosis is being changed by treatment.
For clinicians, this may eventually shift the question from “Which mechanism works?” toward “Which mechanism should this patient receive first, and when does adding another mechanism produce enough incremental benefit to justify it?”
For regulators, payers and HTA bodies, that is an even harder question because combination precision medicine can quickly become expensive without necessarily becoming more effective.
The trial also illustrates why biomarkers need to remain subordinate to outcomes. Eplontersen suppressed the protein responsible for ATTR amyloid formation, yet that biological success was not sufficient to deliver a significant improvement in the trial’s primary clinical endpoint.
That is not evidence against gene silencing as a therapeutic strategy. It is evidence against assuming that deeper intervention in a disease pathway automatically produces greater clinical benefit in every treatment context.
The future of eplontersen in ATTR-CM is therefore uncertain, and AstraZeneca and Ionis have been evaluating the full dataset and next steps following the Phase III result. More broadly, CARDIO-TTRansform may force the field to become more precise about precision medicine itself.
The next generation of ATTR-CM treatment may not be defined by how many targeted drugs can be combined. It may be defined by knowing which patient needs which mechanism, at which stage of disease, and what another therapy genuinely adds to the care they already receive.
