Early vasodilator infusion does not affect CV mortality in acute HF, despite exerting expected physiological effects
ESC HF 2017 Enrollment of ineligible patients in the TRUE-AHF study may have undermined the ability of the trial to demonstrate a benefit of ularitide in patients with acute heart failure.
BackgroundNews - Apr. 29, 2017
Paris, France | The Effects of Ularitide on the Clinical Course and Cardiovascular mortality in acutely decompensated heart failure depend on the eligibility of patients for the trial: deep-dive results of the TRUE-AHF Trial
Presented at ESC Heart Failure 2017 by Milton PACKER (Dallas, United States of America)
Increased intravascular volume due to sodium retention, vasoconstriction and/or transcapillary plasma shifts can lead to acute ventricular distension. This is accompanied by increases in NT-proBNP and haemodilution, and may cause worsening heart failure (WHF) events. Myocardial microinjury (and an increase of troponins) confers an accelerated risk of hospitalisation, and an increased long-term risk of CV death. It is an intriguing question whether the fluid retention pathway is causally linked to the pathway of myocardial injury leading to clinical events.
The TRUE-AHF study aimed to test the hypothesis that if the fluid retention cascade is disrupted, it can affect the natural course of heart failure favourably. 2157 AHF patients were screened and included in 156 centres in 23 countries. These patients were randomised to i.v. infusion of the vasodilator ularitide (15 ng/kg/min) or placebo, which was administered after a median of 6.1 hours.
Main results
- Ularitide exerted the expected effects on cardiac distension, and showed a larger drop in systolic blood pressure during infusion than placebo, and a 47% greater decrease in Nt-proBNP at 48 hours.
- As compared with placebo, at 48 hours, ularitide led to significant increases in haemoglobin and serum creatinine, and decreases in hepatic transaminases, indicative of vascular decongestion.
- A significant benefit of ularitide over placebo was seen on in-hospital HF events during infusion, but this effect waned when the infusion was stopped.
- No effect of ularitide on CV mortality was seen up to 34 months (HR: 1.03, 95%CI: 0.85-1.25, P=0.75).
- Patients on ularitide did not experience dyspnoea relief significantly more often than patients receiving placebo.
To understand the puzzling finding that, despite reduction in cardiac wall stress and intravascular decongestion, no dyspnoea relief was seen with ularitide treatment, the eligibility of enrolled patients was reviewed, before the trial blink was broken. To detect a dyspnoea benefit, the study mandated a background of 2 hour medication stability, no concomitant confounding medication, minimal levels of NTproBNP and safe blood pressure. The eligibility review yielded 358 (17%) patients who were prospectively identified as not having met one or more of the protocol’s prespecified conditions.
- Subsequent analyses in eligible (83%, n=1799)) and ineligible (17%, n=358)) patients revealed that ularitide was superior to placebo for the hierarchical clinical composite (dyspnoea relief) in eligible patients, and inferior to placebo in ineligible patients.
- The lack of effect on CV mortality was not influenced by eligibility of patients.
An analogy was made to the TOPCAT trial, in which patients in Russia and Georgia, had much lower risk of events than patients enrolled in the Americas, in both the placebo and spironolactone treatment groups. Based on the event rates, patients in Russia and Georgia clearly suffered from another condition. In TRUE-AHF, ‘sites of concern’ were identified based on having enrolled at least three ineligible patients.
- In Czech Republic, Estonia, Poland and Serbia, 18 out of the 19 top recruiting sites were sites of concern, which recruited 772 patients.
Conclusion
The TRUE-AHF study results were neutral and did not show a difference in risk of CV death up to 34 months, although the drug exerted the expected physiological effects. Analyses of eligibility of patients who were enrolled showed that 17% should not have been enrolled, which may have affected the results. A different effect on the clinical composite was seen based on eligibility of patients, but not on CV mortality. Dr. Packer concluded that these analyses do not alter the primary findings of the TRUE-AHF trial. The results underscore that eligibility criteria for a clinical trial matter and must be respected. The findings highlight the importance of ensuring that a small minority of sites who enrol ineligible patients do not undermine the ability of trials to demonstrate a drug effect.
Dr Milton Packer, trial PI and Distinguished Scholar in Cardiovascular Science at Baylor University Medical Center, Dallas, Texas, US, said in a press release: “What is absolutely amazing is that the drug was statistically significantly better in eligible patients and statistically significantly worse than placebo in ineligible patients. So not only did the two groups respond differently, they responded in opposite directions. The eligibility criteria not only identified the patients who would respond but also identified the patients who would not be harmed. When you put the eligible and ineligible patients together you get a neutral result. Ineligible patients were being treated with confounding medications which made the interpretation of the data very difficult and took away from the safety of the drug. If the trial’s eligibility criteria had been followed it is likely that we would have found a benefit of ularitide on shortness of breath.”
Dr Packer concluded: “This shows how it is possible for a drug to actually work but a trial not to find it.”
As a discussant, Dr. John Cleland (London, Great Britain) noted that a benefit was seen during the infusion, which does not seem to result in long-term gain. He thinks the findings might be the result of a failure of design of the study. When looking at the patients on placebo, low numbers died, had WHF or had persisting problems. He suggested it would be hard to improve these numbers with a short-term intervention. Also, he said that not many patients have dyspnoea at rest, only when walking for a few meters. So, patients hospitalised for AHF may need to be divided in subsets better. Moreover, many patients with acute breathlessness feel better within 3-6 hours with conventional therapy, and hence they do not need treatment for AHF. The remaining patients who fail to respond to initial measures thus form an unmet need. Another unmet need is longer term treatment to prevent rehospitalisation and death.
Disclosures
Our coverage of ESC HF 2017 is based on the information provided during the congress.