XXB750 Phase 2 Trial in Heart Failure Shows Harm Signal

Key Takeaways

• XXB750 was associated with more death or worsening heart failure events than sacubitril/valsartan or placebo, and the trial was stopped early after safety review.
• At 16 weeks, pooled XXB750 was associated with higher NT-proBNP and lower cGMP, while sacubitril/valsartan showed the opposite biomarker direction.
• The authors interpreted the pattern as consistent with possible functional antagonism of endogenous natriuretic peptides, although early stopping limited certainty.

An unexpected harm signal emerged in a Nature Medicine phase 2 randomized trial of XXB750 in symptomatic heart failure. Death or worsening heart failure events occurred in 25% of pooled XXB750 recipients, versus 8% with sacubitril/valsartan and 0% with placebo. Over 16 weeks, NT-proBNP increased and cGMP decreased with XXB750, opposite to the biomarker direction seen with sacubitril/valsartan. Together, those findings ran counter to the intended NPR1 agonist strategy in heart failure.

Investigators conducted an international, multicenter, parallel-group phase 2 trial in adults aged 18 years or older with symptomatic heart failure and NYHA class II to III. Eligibility also required left ventricular ejection fraction below 50% and NT-proBNP of at least 600 ng/L in sinus rhythm or 900 ng/L with atrial fibrillation or flutter. Participants had to be on stable background therapy with an ACE inhibitor, ARB, or sacubitril/valsartan before randomization.

Among 136 randomized participants, those on background ACEi or ARB were assigned to blinded subcutaneous XXB750 60 mg, blinded XXB750 titrated to a maximum of 120 mg, matching placebo, or open-label sacubitril/valsartan; those already on sacubitril/valsartan were assigned to blinded XXB750 60 mg, blinded XXB750 titrated to a maximum of 120 mg, or matching placebo. XXB750 and matching placebo were given every four weeks for four doses across 16 weeks, whereas open-label sacubitril/valsartan was titrated, where possible, over the same treatment period.

The primary endpoint was change in NT-proBNP from baseline to week 16 after treatment initiation. At week 16, pooled XXB750 showed an NT-proBNP ratio of 1.34 (95% CI 1.07-1.66) and a cGMP ratio of 0.77 (95% CI 0.65-0.91). Sacubitril/valsartan moved in the opposite direction, with an NT-proBNP ratio of 0.70 (95% CI 0.45-1.10) and a cGMP ratio of 1.38 (95% CI 1.13-1.69). The investigators noted a dose-response pattern for both biomarkers, and the overall findings ran counter to the intended agonist effect.

Overall adverse events, serious adverse events, and treatment-emergent adverse events were more frequent in the XXB750 groups. Four deaths occurred with XXB750, one occurred with sacubitril/valsartan, and none occurred with placebo during follow-up. Heart failure leading to hospitalization occurred in 0% with placebo, 8% with sacubitril/valsartan, 8% with XXB750 60 mg, and 13% with XXB750 120 mg. Dyspnea and hypervolemia were more common with XXB750, while no classical immune-mediated hypersensitivity reactions were reported. After review of excess heart failure events on 6 August 2024, the data monitoring committee recommended stopping the trial prematurely, while follow-up continued per protocol.

The authors interpreted the rise in NT-proBNP, the fall in cGMP, and the excess worsening heart failure events as consistent with possible functional antagonism of endogenous natriuretic peptides. They also noted that worsening heart failure appeared most pronounced among participants receiving background sacubitril/valsartan, although those subgroup patterns were not definitive. Early trial stopping left the study underpowered for its primary biomarker hypothesis, and subgroup interaction analyses warrant cautious interpretation. That proposed mechanism remains an author interpretation rather than an established causal explanation for the observed findings.