By Dr Jonathan Shurlock
Edited by Dr Saadia Aslam
PARAGLIDE-HF was presented at Heart Failure 2023, with peer reviewed publication awaited. The rationale, design and baseline characteristics have been published. Patients with HF and ejection fraction (EF) >40%, NT-proBNP levels ≥ 500 pg/mL, and within 30 days of an episode of HF decompensation were eligible and randomised to treatment with sacubitril/valsartan versus valsartan alone.
467 participants were randomised (52% women, age 70 ± 12 years). Median (IQR) EF was 55% (50%–60%). 23% of participants had ‘mildy’ reduced EF (LVEF 41%–49%). 24% had an EF > 60%. 33% of participants presented with de novo HFpEF. At enrolment median NT-proBNP was 2009 (1291–3813) pg/mL.
The primary endpoint was defined as a change in NT-proBNP, in which a significant time-averaged reduction from baseline to 4 and 8 week follow up was observed in the sacubitril/valsartan group compared with valsartan only group (ratio of change 0.85; 95% CI 0.73-0.999; p=0.049) was reported.
There was no significant difference observed in secondary outcomes, defined as cardiovascular death, hospitalisation for heart failure and hospitalisation for any reason, between the two groups (unmatched win ratio 1.19; 95% CI 0.93-1.52; p=0.16).
Safety endpoints were defined as, worsening renal function (composite of a decline in eGFR >50% and reaching ESRD) and symptomatic hypotension (systolic blood pressure <100mmHg with attributable symptoms).
The authors surmise that their data demonstrates the potential role of sacubitril/valsartan as a treatment option for patients with HFpEF who have a worsening heart failure event. This conclusion is reached based on finding a significant difference in their primary outcome measure, and a ‘trend towards’ favouring sacubitril/valsartan for secondary outcome measures.
While the full peer reviewed study is not yet available some critiques include the short duration of study and the categorisation of HF with ‘mildly’ reduced ejection fraction. Most notable is the variable use of decimal places in the reported confidence interval ranges. For the primary outcome measure the authors report the upper limit of the confidence interval to 3 decimal places (0.999), compared with 2 decimal places for all other limits. It will be useful to see the full data set on publication to identify the impact of this on overall findings.
Read more:
