Benefit-Risk Assessment of Crataegus Extract WS 1442

An Evidence-Based Review

Christian J. F. Holubarsch; Wilson S. Colucci; Jaan Eha


Am J Cardiovasc Drugs. 2018;18(1):25-36. 

In This Article

Clinical Safety and Tolerability

A comprehensive systematic review of the safety profile of Crataegus species, covering all clinical studies and post-marketing surveillance projects with hawthorn monopreparations published in or before January 2005, was presented by Daniele et al..[36] The authors identified 14 studies with WS 1442, of which ten were randomized, controlled trials, two were uncontrolled trials, one was an observational study, and one was a cohort study. In these projects, a total of 1563 subjects (controlled clinical trials: 395; uncontrolled studies: 1168) were exposed to the herbal medicinal product and received daily doses of up to 1800 mg for treatment periods of up to 16 weeks in clinical trials and 2 years in post-marketing surveillance projects. One trial was performed in healthy volunteers, while the participants of all other studies suffered from heart failure corresponding to NYHA classes I–II or II–III.

In the randomized trials included in the review and presenting data on adverse events (AEs), 30 of 356 patients treated with WS 1442 and 34 of 299 receiving placebo or a reference drug reported AEs with any causal relationship. In the trials reviewed, AE incidence in patients treated with WS 1442 was generally on a placebo level. There was no apparent relationship between WS 1442 dose and AE manifestation. In one double-blind study investigating daily doses of 1800 and 900 mg/day WS 1442 compared to placebo during 16 weeks of treatment, the proportions of patients with any AEs were 26.1, 28.6, and 51.4%, respectively, and therefore lower in patients treated with WS 1442 than in placebo-treated patients.[37] The authors of the review interpreted this result to be an indicator of treatment efficacy rather than of safety, since WS 1442 appears to prevent, rather than elicit, symptoms like dizziness that are frequently reported as AEs.[36] In an observational study included in the review, which monitored 1011 patients with heart insufficiency stage NYHA II treated with WS 1442 over a period of 24 weeks, 98.7% of physicians noted a "good" or "very good" tolerance to the treatment, with only 14 AEs.[38] Moreover, in any of the studies reviewed, none of the AEs potentially related to WS 1442 were serious.

One of the merits of the review of Daniele et al.[36] is that their search strategy assured a comprehensive overview of the literature available at the time of its publication regarding potential risks associated with Crataegus extract. Limitations include the lack of a comprehensive overview of the study selection criteria and of an assessment of the methodological quality and validity of the trials (the authors appear to have accepted all identified studies conducted with Crataegus monopreparations irrespective of their methodological quality, unless they were duplicate publications of other work). Since the publication of the review, additional clinical trials involving WS 1442 have been published:

The SPICE (Survival and Prognosis: Investigation of Crataegus Extract WS 1442 in congestive heart failure) trial reported by Holubarsch et al.[39,40] was a placebo-controlled, double-blind, multicenter trial in which 2681 adults with NYHA class II or III congestive heart failure and reduced left ventricular ejection fraction (LVEF ≤ 35%) were randomized to receive 900 mg/day WS 1442 (n = 1338) or placebo (n = 1343) for a total duration of 24 months. At baseline, the average LVEF was 24%, and thus a substantial percentage of the study participants were at a high risk of cardiac mortality. WS 1442 was administered as an add-on treatment to guideline cardiac medication, and thus digoxin or digitoxin, diuretics, β-blockers, and angiotensin-converting enzyme (ACE) inhibitors were permitted as concomitant medication in any clinically appropriate combination or dose. In fact, about 90% of the study participants took at least three concomitant cardioactive drugs: about 85% of the patients in each group were administered diuretics, 83% received ACE inhibitors, 64% were treated with β-blockers, and 56% were administered digitalis or nitrates; concomitant antiarrhythmics were used by about 22% of the study participants, implicating possible proarrhythmic effects.[41,42]

With a sample size of about 1340 patients in each group, the SPICE study had a 95% chance of observing at least one AE whose true incidence in the population is at least 2.2‰ and was thus suited for observing even less frequent complications or drug–drug interactions. During the 2-year follow-up, 67.0% of the patients randomized to WS 1442 and 68.3% of those in the placebo group reported AEs, corresponding to one event in 388 and 370 days of exposure, respectively. Serious AE rates were 39.2 and 41.1%, respectively. The types of AEs were very similar in both treatment groups, and no specific AEs related to WS 1442 were observed.[40]

The Hawthorn Extract Randomized Blinded Chronic Heart Failure (HERB CHF) trial published by Zick et al.[43,44] investigated the effect of 6 months' randomized, double-blind, add-on therapy with WS 1442 or placebo on submaximal exercise capacity in patients with NYHA class II–III chronic heart failure. A total of 120 patients participated, 60 of whom received WS 1442. Any AEs were experienced by 36 patients in the WS 1442 group (60.0%) and by 23 patients (38.3%) in the placebo group. However, 12 patients in each group (20.0%) reported cardiac-related AEs, and their incidences as well as those of common AEs like infections, rash, gastrointestinal complaints, and headache were also comparable. The authors concluded that the difference in the over-all AE rate was unlikely explainable by a pharmacological effect of WS 1442 and may have been attributable to chance. Zick et al.[44] also observed a higher rate of heart failure progression (composite secondary endpoint of death, hospitalization, and increased diuretic use) in the early phase of the trial under WS 1442 as compared to placebo, although the difference disappeared in the course of the 6-month follow-up. The authors interpret this as an indicator that WS 1442 could increase the early risk of heart failure, but also warn that these findings could be due to chance and must be interpreted with caution, since they come from a post hoc analysis of an outcome measure not pre-specified before unblinding, which was performed in a relatively small study (120 patients) with a limited absolute number of heart failure events (28/60 and 26/60 patients for WS 1442 and placebo, respectively). The results of Zick et al.[44] are also not supported by the SPICE trial,[40] in which 2681 patients were observed for up to 2 years. In contrast to Zick et al.,[44] the heart failure hazards in the SPICE trial were proportional over time, and no signals for an increased risk of heart failure under treatment with WS 1442 were observed, neither during the initial months of treatment, nor at any other time during follow-up.

Another randomized, controlled, exploratory trial was performed by Härtel et al. to investigate the effect of endurance exercise training and add-on medication with WS 1442 in patients with HFpEF meeting NYHA class II criteria.[45] A total of 140 patients on guideline cardiac medication received 8 weeks of aerobic endurance training, and 70 of them were randomized to additional, open-label administration of 900 mg/day WS 1442. AEs with any causal relationship were observed in nine patients in the WS 1442 group (12.9%) and in 12 patients in the control group (17.1%). Most events were musculoskeletal complaints and injuries (possibly related to exercise training) and infections. None of the events reported in the WS 1442 group were assessed to be related to the herbal medicinal product.

Signals for possible drug interactions with WS 1442 were not observed in any of the studies reviewed by Daniele et al.,[36] nor in those published after completion of their review,[40,43,45] nor in literature searches that also included spontaneous reports to health authorities about suspected adverse reactions to hawthorn containing products.[9,46] In a randomized, crossover, interaction study in healthy volunteers, concurrent administration of WS 1442 had no effect on the pharmacokinetic properties of digoxin.[47]