Literature - Armstrong PW, Zheng Y, Troughton RW, et al. - JACC Heart Fail. 2022 Sep;10(9):677-688. doi: 10.1016/j.jchf.2022.04.015

Introduction and methods

Background

Recently, the VICTORIA (Vericiguat Global Study in Subjects With Heart Failure With Reduced Ejection Fraction) trial showed a reduction in the primary composite outcome of CV death or HF hospitalization with vericiguat—an oral soluble guanylate cyclase stimulator —in patients with recent worsening chronic HF compared with placebo [1]. A prespecified subgroup analysis suggested less benefit for patients with an NT-proBNP level in the highest quartile at randomization (>5314 pg/mL) and more benefit for those with NT-proBNP <4000 pg/mL [2]. For the high-risk population of HFrEF patients with a recent worsening event, it is important to determine whether there is a relationship between sequential changes in NT-proBNP levels, treatment with vericiguat, and clinical outcomes.

Aim of the study

The objectives of this post-hoc analysis of the VICTORIA trial were: (1) assessing the relationship between sequential changes in plasma NT-proBNP level and the primary composite outcome of CV death or HF hospitalization; (2) evaluating the effect of vericiguat compared with placebo on changes in NT-proBNP level; and (3) exploring the association between the efficacy of vericiguat and NT-proBNP changes.

Methods

In the VICTORIA trial, 5050 patients with recent worsening chronic HFrEF and elevated plasma NT-proBNP levels (>1000 pg/mL in sinus rhythm and >1600 pg/mL in AF) were randomized to vericiguat or placebo. For 4805 patients (95%), NT-proBNP measurements were available at randomization: 2414 in the vericiguat group and 2391 in the placebo group. Additional samples were taken at 16, 32, 48, and 96 weeks.

Patients were divided into 4 groups based on their NT-proBNP level at randomization (≤2816 pg/mL or >2816 pg/mL, which was the median value) and the subsequent relative NT-proBNP change between randomization and week 16 (≥20% reduction or <20% reduction, including increases).

Outcomes

The association of the primary composite outcome with the NT-proBNP change between randomization and week 16 was assessed. The authors chose week 16 because it was the first post-randomization NT-proBNP sample collected and it provided the optimal point at which the extent of change was greatest and loss due to missing samples, death, or other reasons was smallest. Joint modeling was used to examine how vericiguat treatment was related to (event-free) survival via NT-proBNP levels. The relationship between vericiguat, NT-proBNP levels, and clinical outcomes was assessed with mediation analysis.

Main results

Association between NT-proBNP changes at week 16 and clinical outcomes

• Irrespective of treatment allocation, patients with an NT-proBNP level ≤2816 pg/mL at randomization and an NT-proBNP reduction ≥20% at week 16 had the lowest cumulative incidence of the primary composite outcome during follow-up, whereas patients with NT-proBNP >2816 pg/mL at randomization and reduction <20% had the highest primary outcome event rate (P<0.001 across all 4 groups).
• When patients with NT-proBNP ≤2816 pg/mL at randomization and NT-proBNP reduction ≥20% at week 16 were used as the reference, the highest primary outcome event rate at 2 years was seen in patients with NT-proBNP >2816 pg/mL and <20% reduction (adjusted hazard ratio (HR): 3.66; 95%CI: 2.97–4.50; P<0.001), followed by patients with NT-proBNP ≤2816 pg/mL and <20% reduction (adjusted HR: 2.35; 95%CI: 1.92–2.89; P<0.001).
• A similar pattern of increasing hazard was seen for the individual components of the primary composite outcome (i.e., CV death or HF hospitalization).

Evolution of NT-proBNP changes according to treatment group

• In both the vericiguat and placebo groups, a significant decline in median NT-proBNP level was observed between randomization and week 16, which continued through week 96.
• However, the median reduction at week 16 was greater in the vericiguat-treated patients than in the placebo-treated patients (450 pg/mL (23%) vs. 200 pg/mL (11%); P<0.001) and remained less overall through the follow-up period .
• The odds ratio (OR) for any NT-proBNP reduction for the vericiguat versus placebo group was 1.45 (95%CI: 1.28–1.65; P<0.001), and it was 1.27 (95%CI: 1.10–1.47; P=0.001) for ≥50% reduction.
• Compared with patients receiving placebo, vericiguat-treated patients were less likely to have an increase in NT-proBNP levels (OR for ≥20% increase: 0.68; 95%CI: 0.59–0.78; P<0.001; OR for ≥50% increase: 0.70; 95%CI: 0.59–0.82; P<0.001).

Relationship between evolution of NT-proBNP changes and efficacy of vericiguat

• At week 16, there was a modest treatment effect of vericiguat related to serial NT-proBNP levels on the primary composite outcome compared with placebo (HR: 0.96; 95%CI: 0.95–0.99).
• This treatment effect increased to a ≥10% relative decline in the primary outcome event rate at week 48 (HR: 0.90; 95%CI: 0.85–0.96).
• The average extent of mediation of the composite outcome related to NT-proBNP level was 45%, which further supported NT-proBNP as a potential mediator of vericiguat’s treatment effect.

Conclusion

References

Find this article online at JACC Heart Fail.