BLAST-AHF: insights into biased AT1 ligands and heart failure. Beginning of the end or end of the beginning?

This editorial refers to ‘Biased ligand of the angiotensin II type 1 receptor in patients with acute heart failure: a randomized, double-blind, placebo-controlled, phase IIB, dose-ranging trial (BLAST-AHF)’^†, by P.S. Pang et al., on page 2364.

Based upon the latest statistics, heart failure (HF) will continue to grow as a major public health problem worldwide, with increasing hospitalization rates, post-discharge death, and readmission for HF remaining a healthcare challenge.¹ While the recently approved drug Entresto has demonstrated favourable outcomes in patients with chronic HF,² new and innovative therapies for acute HF (AHF) have repeatedly failed in pivotal trials. Recently phase III studies investigating the therapeutic peptides serelaxin (RELAX-AHF-2)³ and ularatide (TRUE-AHF)⁴ failed to demonstrate reduction in cardiovascular death post-discharge and rehospitalization. Thus, there remains a high and unmet need for innovative therapeutics for AHF.

A hallmark of HF is the activation of the renin–angiotensin system (RAS), resulting in the increase of circulating angiotensin II (ANG II). Indeed, anti-RAS drugs have proven to improve HF outcomes, and include drugs such as angiotensin-converting enzyme inhibitors (ACEis), angiotensin type I receptor blockers (ARBs), and now angiotensin receptor neprilysin inhibitors (ARNis) such as Entresto. What we have learned from studies of ANG II is that its pathological actions such as vasoconstriction, impairment of renal function, and induction of myocardial remodelling are initiated by an interaction with the angiotensin type 1 receptor (AT1R) and activation of an AT1R-dependent G-protein pathway.

TRV027 represents a novel first-in-class AT1R antagonist which targets the traditional G-protein-dependent signalling pathway to block ANG II effects such as vasoconstriction and retention of sodium and water which may lead to vasodilation and a reduction in congestion. However, TRV027 also possesses very novel properties which also involve targeting a G-protein-independent pathway, with activation of the effector molecules β-arrestins resulting in increased cardiac contractility and suppression of cardiac apoptosis, as supported by elegant laboratory-based research (Figure 1).⁵ Like ANG II, TRV027 is an octapeptide, and in fact the two peptides are identical in six amino acids. Importantly, elegant studies established that key amino acid changes resulted in the ability to inhibit G-protein signalling, activate β-arrestin signalling, and stabilize the complex between β-arrestins and AT1R in endosomes.⁶ Thus, TRV027 is a novel biased ligand of AT1R, antagonizing ANG II-stimulated G-protein activation while stimulating β-arrestin. Such actions of TRV027 are therapeutically attractive for HF.

Figure 1.

ANG II activates G-protein-dependent AT1R. TRV027 inhibits AT1R and activates the β-arrestin pathway independent of G-protein stimulation, which induces vasodilatation and preserved renal function. In addition, TRV027 enhances cardiomyocyte contractility and reduces apoptosis, which may induce improvement in myocardial function (see Violin et al.¹⁵).

Supporting this concept of TR027 in HF were key laboratory studies. In a model of HF in rodents, TRV027 compared with ARBs resulted in a lowering of blood pressure (BP) with each drug. In contrast, only TRV027 increased myocardial performance and preserved cardiac stroke volume.⁷ In a large animal model of HF in canines, TRV027 was a potent vasodilator which reduced pulmonary capillary wedge pressure and systemic and renal vascular resistance, increased cardiac output, and preserved sodium excretion and glomerular filtration rate despite a decrease in BP.⁸ In addition, when added to furosemide, TRV027 preserved furosemide-mediated natriuresis and diuresis, while reducing cardiac preload and afterload.⁹

Following robust pre-clinical testing, a first-in-human study was reported in 2013 with an ascending dose design to explore tolerability, pharmacokinetics, and pharmacodynamics in healthy volunteers.¹⁰ Sodium intake was controlled to activate the RAS, and TRV027 proved to be safe and well tolerated with a short half-life. Importantly, TRV027 reduced BP to a greater degree in subjects with RAS activation as determined by an elevated plasma renin activity (PRA) as compared with those with a normal PRA. In 2013, a phase IIa study in patients with NYHA class III and IV HF with elevated filling pressures was also reported. This study in chronic HF, with i.v. TRV027, observed dose-dependent decreases in BP, in which the actions of TRV027 in reducing pulmonary capillary wedge pressure were noted primarily in patients with RAS activation.¹¹

Based on these early clinical studies of TRV027, a key phase IIb clinical trial BLAST-AHF (Biased Ligand of the Angiotensin Receptor Study in Acute Heart Failure) was conducted and the results of this trial are reported in the current issue by Pang and co-workers.¹² BLAST-AHF was designed as an international prospective, randomized, double-blind, placebo-controlled, dose-ranging study that randomized AHF patients to one of three doses of i.v. TRV027 (1, 5, or 25 mg/h) or matching placebo (1:1:1:1) for at least 48 h and up to 96 h. It is important to note that after 254 patients were completed, a pre-specified interim analysis resulted in several protocol changes consistent with an adaptive design. Due to an unexpected lack of significant BP lowering in the trial, the BP entry criterion was lowered while also lowering the BP threshold for terminating the study drug infusion. These changes were made to increase the potential activation of the RAS, which was thought to be important in promoting the AT1–TRV027 interactions. A key change as well was the overweighting of both placebo and the 5 mg/h dose due to greater efficacy of the 1 and 5 mg/h TRV027 doses. The reason for this redesign of study was also motivated by the observed findings that high dose (25 mg/h) TRV027 suggested worsening of HF. In addition, the main secondary endpoints were also modified, in which the change in NT-proBNP was replaced with a new main secondary endpoint: a modified primary endpoint incorporating changes in high-sensitivity troponin T (hsTnT) and cystatin-C through 48 h to assess for end-organ protection or prevention of injury.

At the completion of the trial, a total of 621 patients were randomized in BLAST-AHF. Of the 618 treated patients, 594 survived to day 30, 540 patients completed day 180 follow-up, and 78 died prior to day 180 with a mean follow-up of 170 days. The seminal results of the trial were that that the three tested doses of TRV027 showed no efficacy signals in any of the five components of the primary composite endpoint which included (i) death through day 30; (ii) HF re-hospitalization through day 30; (iii) worsening HF through day 5; (iv) dyspnoea through day 5, visual analogue scale (VAS) area under the curve; or (v) length of hospital stay. A critical observation was that the three components, worsening HF through day 5, change in dyspnoea VAS area under the curve through day 5, and length of initial hospital stay through day 30, tended to increase in the group receiving TRV027 at a dose of 25 mg/h. Moreover as pointed out by Pang et al,, when changes in cystatin-C and hsTnT at 48 h were added to the primary composite endpoint, no significant differences were observed between any of the treatment groups and placebo. In addition, NT-proBNP decreased from baseline to 48 h in the placebo group, which was greater than in the active TRV027 treatment groups. Surprisingly, TRV027 did not decrease BP compared with placebo. Thus, TRV027 in BLAST-AHF was without benefit over placebo at any dose with regards to the primary composite or any of the individual components.

What mechanisms might explain the unexpected lack of action of TRV027 in patients with AHF? One mechanistic hypothesis could be the lack or reduced level of RAS activation that could have resulted in the reduction in ANG II, the hormonal substrate needed for the actions of TRV027 at the level of AT1. Clearly, activation of the RAS is one of the fundamental aspects of HF pathophysiology and it is assumed that there is greater RAS activation in patients during HF decompensation such as AHF. Thus, as speculated by the investigators, the population of AHF patients in BLAST-AHF may have had inadequate activation of the RAS. Relevant to this speculation is the recent report of Basu et al. that circulating ANG II levels were not elevated in patients with AHF compared with normal controls using a sensitive equilibrium assay for ANG II.¹³ A second mechanism may be suggested by the trend for worsening of HF parameters with high dose TRV027 (25 mg/h). Could it be that TRV027, which shares structural similarities to the ANG II analogue saralasin, could actually act as an agonist to AT1 signalling with loss of β-arrestin signalling in human HF?

Another potential explanation for the failed trial could be study design and duration of treatment. One of the primary composites of BLAST-AHF was dyspnoea VAS area under the curve through day 5. Several recently completed clinical trials for AHF (ASCEND, nesiritide; VERITAS, tezosentan; PROTECT, rolofylline) have failed to show favourable effects for symptom resolution including dyspnoea in the acute phase of HF. Only the EVEREST study with arginine vasopressin (AVP) receptor antagonism showed improvement in patient-assessed dyspnoea, body weight, and oedema.¹⁴ These results suggest that improvement of dyspnoea requires robust decongestion. Further, it should also be noted that the low 30-day events rate in the placebo arm suggests the need for longer term outcomes to be followed in order to ensure a robust event rate. Like recent AHF trials with seralaxin and ularitide, BLAST-AHF treated for up to 96 h. Is this too short to achieve more sustained improvements in outcome? Has the time come to design AHF trials that both treat the AHF hospitalization phase and also continue for a longer defined time such as 30 days? Perhaps the time has come really to address this critical period of post-AHF and a transition from short 3-day AHF trials to perhaps an integration of AHF and post-AHF phases of HF.

What are the next steps moving forward with TRV027 if any? Despite the disappointing results of BLAST-AHF, TRV027 demonstrated beneficial effects in patients with chronic HF in early phase IIa. Perhaps this positive outcome could be related to greater activation of the RAS in chronic HF compared with AHF as some studies may suggest. Thus, more in-depth studies in chronic HF and defining its relationship to RAS activity may be worthwhile. Further, we may not yet fully understand the natural history of RAS activation throughout the stages and phases of HF including at the AT1 receptor level. Thus, clinical studies are needed which define the evolution of RAS activation throughout the spectrum of HF.

What we know for sure is that AHF is a challenging therapeutic target, and more in-depth mechanistic studies are required addressing both pathophysiology and clinical pharmacology. Pang and co-workers are to be congratulated for successfully completing and reporting the findings of BLAST-AHF even if negative. Indeed there is much to be learned from negative as well as from positive clinical trials. With further studies of TRV027 or other drugs in this class of biased ligands for ATI, we may answer Churchill’s question of ‘is it the beginning of the end or the end of the beginning’.

Funding

The authors acknowledge support from the National Institutes of Health and RO1 HL36634-29 (awarded to J.C.B.).

Conflict of interest: none declared.