ATLANTA, GA — Despite promise for improving heart failure (HF) in pilot studies, an interarterial shunt (IAS) device did not meet the primary composite efficacy endpoint in a pivotal trial, a result stemming from apparent harm among the subgroup with preserved (> 40%) left ventricular ejection fraction (LVEF).
The results from a prespecified analysis support the idea that IAS “is beneficial in patients with reduced LVEF and harmful in patients with preserved LVEF,” reported Gregg W. Stone, MD, professor of medicine, Icahn School of Medicine at Mount Sinai, New York City.
However, given the failure to reach the primary endpoint, the conclusion of a benefit in one type of HF and harm in another is, by definition, an exploratory hypothesis, said Stone, who presented the data April 6 here at the American College of Cardiology (ACC) Scientific Sessions 2024.
Mostly NYHA Class III Patients Randomized
In this double-blind multicenter trial called RELIEVE-HF, 508 patients in ischemic or non-ischemic ambulatory HF (> 95% with NYHA class III) were randomized to IAS implantation with a Ventura device (V-Wave Ltd) or a placebo procedure. Patients and all post-catheterization laboratory personnel were blinded to assignment.
Documented HF for at least 6 months and being on guideline-directed medical therapy for HF were inclusion criteria. Severe valve lesions, severe pulmonary hypertension, and significant right ventricular dysfunction were among exclusion criteria. Anatomical anomalies incompatible with IAS and hemodynamic instability were also exclusion criteria.
After up to 2 years of follow-up, there were no significant differences overall or for any of the composite hierarchical endpoints of all-cause death, cardiac transplantation or implantation of a left ventricular assist device (LVAD), HF hospitalization, worsening outpatient HF events, and an adverse 5-point change in the Kansas City Cardiomyopathy Questionnaire Overall Summary (KCCQ-OS) score.
There was also no harm. The incidence of the primary safety endpoint of device- or procedure-related death, stroke, embolism, need for open-heart surgery, or need for endovascular repair (MACNE) over the 2-year follow-up was 0%. Relative to prespecified definition of these risks, the safety of IAS was reached at a high level of statistical significance (P < .0001).
Win Ratio Analyses Showed No Differences
When assessed as a win ratio, an increasingly used method of outcomes analysis that compares wins, ties, and losses between randomized groups, there was no difference overall (P = .20) or for any individual outcome. When confined to cardiovascular (CV) events (death, HF hospitalizations, or worsening signs of HF), the relative risk ratio was exactly 1.0, meaning no difference (P = .96).
However, the differences became remarkable when HF patients with reduced LVEF (HFrEF) were compared to those with preserved LVEF (HFpEF). In both groups, CV event curves separated within a few months of randomization. In HFrEF, the separation was in favor of IAS, reaching a 45% relative rate ratio (RR) reduction in the composite of CV events (RR, 0.55; P < .0001) by the end of follow-up.
In those with HFpEF, the separation was in favor of placebo. As a result, the curve reached a 68% RR of harm for IAS (RR, 1.68; P = .0001) by end of follow-up.
When evaluated by specific events, there was no significant difference in the rate of outpatient worsening of HF or LVAD or CV transplantation, but the RR of HF hospitalizations was doubled (RR, 2.05; P = .0008) and death was more than tripled (RR, 3.24; P = .004) among those with HFpEF that received the IAS device. Stone said the vast majority of the deaths were related to CV causes.
A closer analysis of event rates in relation to LVEF showed the transition from benefit to harm occurred at an LVEF of approximately 40%, Stone noted. He said several statistical analyses support the conclusion that the differences in outcomes between those with HFrEF or HFpEF were “not a play of chance.”
Right Chamber Overload Might Explain Harm
Speculating about the difference, Stone suggested IAS benefit in those with HFrEF and harm in those with HFpEF is related to hemodynamics. While the purported benefit of IAS is to reduce left atrial pressure by shunting blood from the left to right chambers, the potential harm is right-side overload. This trial appears to confirm the benefit for those with HFrEF but the harm for those with HFpEF.
The reason, according to Stone, is that blood shunted to the right chamber can be accommodated in the larger compliant hearts of the HFrEF population but not in the smaller and relatively stiff hearts of those with HFpEF. He believes the relative differences in the resilience of the two types of HF to change in blood flow explains the results. Ultimately, despite the design of the trial, the outcomes show that the HFrEF and HFpEF populations turned out “not to be poolable,” Stone said.
Asked if the trial should be repeated with HFrEF patients only, Stone predicted more trials with this and other IAS devices are coming. Differences in the design of IAS devices, particularly the diameter of the shunt, will be important in assessing results.
These trials will be informed by the lessons of RELIEVE-HF, but Stone also said the findings he presented “cannot be ignored.” He said the benefits for improved outcomes, such as those involving worsening HF, are “beyond any statistical room for chance.” As a result, they are relevant for considering how to lower risk in HFrEF patients such as those enrolled in RELIEVE-HF.
Patrick T. O’Gara, MD, chair of cardiology at Brigham and Women’s Hospital, Boston, Massachusetts, agreed that “there is something there,” referring to a benefit from the IAS device in HFrEF patients in this trial. However, referring to the negative outcome of the primary hypothesis, he indicated a routine role in patient management will require further studies that confirm a benefit in the HFrEF subgroup.
To address the question of whether IAS devices improve outcomes in HF, Mary N. Walsh, MD, medical director of heart failure and cardiac transplantation, Ascension St. Vincent Hospital, Indianapolis, Indiana, praised researchers for the rigor of the blinded study design. However, based on the data so far, she expressed concern that it is unclear when IAS devices should be introduced even if future trials confirm the benefit.
Speaking specifically about patients with declining LVEF as they transition from HFpEF to HFrEF, Walsh asked: “How long would we wait?” According to Walsh, this should be a question specifically answered in the next set of trials testing the role of IAS for improving HF outcomes.
Stone reports financial relationships with more than 35 pharmaceutical companies, including V-Wave, the sponsor of the RELIEVE-HF trial. O’Gara reports no conflicts of interest. Walsh has financial relationships with several pharmaceutical companies but not V-Wave or another device manufacturer.
Ted Bosworth is a medical journalist based in New York City.