Introduction
The article “The effect of Crataegus oxycantha Special Extract WS 1442 on clinical progression in patients with mild to moderate symptoms of heart failure” is a 2008 publication in the Eur J Heart Fail, volume 10, number 6 from page 587 to 593. This article evaluates how WSS 1442 extract from C. oxyacantha works in patients who have mild to moderate symptoms of heart failure. This retrospective study assessed the time it took to develop heart failure. The study identified that administration of C. oxyacantha extract leads to faster progression to heart failure. Hence, the extract was found to be ineffective in reducing heart failure progression and instead it seemed to exacerbate progression.
Methods
The study by Zick and colleagues [1] was a retrospective analysis of 120 heart failure patients who had participated in the HERB CHF trial. In this trial, the Crataegus Special Extract WS 1442 was tested among patients who had class II to IV symptoms of the New York Heart Association. The NYHA heart failure classification depends on how severe the symptoms are as well as functionality. Class I HF does not limit physical activity and normal physical activity rarely causes fatigue. Class II HF is characterized by some impaired physical activity but the patient feels comfortable while resting. Normal activities however lead to fatigue or dyspnoea. In Class III HF, physical activity is limited to a large extent and though the patient may feel comfortable when resting, even less than ordinary activities may lead to fatigue. Class IV HF patients perform all activities with discomfort and symptoms are exhibited even at rest [2].
Zick et al [1] studied patients with a positive diagnosis of HF (NYHA classes II-IV) and the diagnosis had to have lasted for more than three months with the patient having left ventricular ejection fraction (LVEF). Eligible participants were at least 18 years old and were receiving a “diuretic, beta blocker, angiotensin receptor blocker or an ACE inhibitor” [1, p. 588]. NYHA class III patients received a spironolactone as a qualification to participate in the study. It is noted that administering spironolactone in low doses for NYHA class III patients improves HF symptoms [3]. Only the patients who had shown stability for at least three months after treatment were included in the study. However, those who had shown stability of at least one month after being treated with diuretics were also included.
Patients who displayed various heart conditions less than three months prior to randomization were excluded. Such conditions included hypertrophic cardiomyopathy, valvular surgery, unstable angina, active myocarditis, myocardial infarction and angioplasty among others. If a patient exhibited sustained ventricular tachycardia or symptomatic tachycardia and was not receiving any treatment, they were excluded from the study. Pregnant and nursing women were also exempted from the study as was the case with conditions that had potential to interfere with exercise other than HF. Such conditions included orthopaedic complications and pulmonary disease among others. Griffin, Topol and Nair [4] mention that such conditions limit exercise.
This study involved human subjects and therefore ethical guidelines had to be followed. The researchers followed the guidelines provided under the Declaration of Helsinki [5] in addition to approval by the University of Michigan Medical School Institutional Review Board. Institutional review boards help in advancing respect for human rights in research [6]. To ensure total adherence to the ethical guidelines, a Data Safety Monitoring Board was available as an independent overseer. All participants gave written informed consent before participating in the study. As such, this study can be considered to be ethical.
Study procedures
It is important to remember that this was a retrospective study that used secondary data. To screen for eligible participants, patients were required to walk for six minutes and those who managed a distance of 150 to 450 meters within the six minutes attended a baseline visit. The ethicality of the study was maintained since participants gave written consent prior to the baseline visit. Stable patients during the baseline study did another six-minute walk and those who managed a distance of 150 and 450 meters were recruited and randomized with all the others being classified as ineligible. Randomization was done into two groups with one group receiving 450mg of the extract two times in a day. The second group received a placebo of the same. To ensure consistency of results, the researchers used one batch of C. oxyacantha extract. The placebo was made of glucose and this was assessed using HPLC thus ensuring standardization [7].
Statistical analysis
Zick et al [1] ensured that baseline characteristics were taken by recording their means and standard deviations (for continuous variables). Counts were taken for some categorical variables while others were determined as percentages. It is evident that the baseline characteristics for both the treatment and placebo group did not differ considerably thus enabling competent comparison of results. The Cochran-Mantel-Haenszel analyses were performed to determine primary outcomes in terms of RR at 95% CI. The Kaplan-Meier method was used to determine time for HF progression. Aalen and colleagues [8] mention that the Kaplan-Meier estimator is the most appropriate non-parametric method for estimating survival function. Estimates were determined using log-rank test while Cox modelling was used to determine how various factors affected prognosis of HF. Such factors included NYHA functional class, age, HF cause, sex and LVEF among others. A backward stepwise regression assessed the most parsimonious approach with an alpha value of ≤.05 being classified as significant. Using the above statistical procedures and parameters, it was possible to obtain a power of 84% in detecting a hazard of at least 2.0. For the LVEF subgroup, there existed an 81% power of detecting a hazard ration of at least 2.6.
Results
It was observed that the demographic and clinical characteristics of the subjects were not different thereby guaranteeing uniformity and consistency of results. During the study, it was observed that 38 participants exhibited an LVEF greater than 40% and these were supposed to be excluded from the study since eligibility required an LVEF ≤ 40%. This may have affected the results since the baseline characteristics differed in 38 patients.
Participation and completion of the study can be rated high among both the placebo and treatment groups. Among the total 120 participants who were recruited, only 9 patients failed to complete the six months study for various reasons. Among the 60 placebo subjects, only 3 patients dropped from the study due to death, pneumonia and heart transplant. For the C. oxycantha extract group, a larger number (6) of patients failed to complete the study with three deaths being reported (2 due to HF and 1 due to aplastic anemia). The other three patients dropped due to complications such as Grave’s disease, thyomas and heart transplant.
HF progression was recorded in 46.6 percent of patients in the treatment group and 43.3 percent of the placebo group. Since the p value was greater than.05 (i.e. p =.86), this difference was termed as non-significant. There was a detection hazard of 1.14 as far as progression of HF was concerned as portrayed by the odds ratio. The confidence interval was not wide as both values were positive (95% CI = 0.56 – 2.35). There were more deaths and hospitalizations among Crataegus Special Extract WS 1442 group (3.3% and 18.3% respectively) compared to the placebo group (1.6 and 10.0% respectively). The number of those who needed increased diuretic was however higher among placebo group that the treatment group (19 versus 15 patients respectively).
According to the Kaplan-Meier curve for both treatment groups, it was evident that hazards changed over time. This was demonstrated by a crossing pattern that was observed at different time periods of therapy. Although hazard for Crataegus Special Extract WS 1442 was slightly lower than that of placebo, there was crossing at around 180 and 220 days of therapy showing that the hazard was equal for both groups of treatments at the crossing points. A Cox regression was performed to find risk of heart failure progression over time. There was a significant change in risk (p =.047) in both groups over time. It was demonstrated that the risk of HF progression among Crataegus Special Extract WS 1442 was 3.9 times higher than that of the placebo. This difference was significant and a fairly large number of population had the risk (95% CI = 1.0 – 13.7; p =.035). There was a decrease in hazard ratio for WS 1443 group with time and at 6 months, there was no increase in HF progression due to Crataegus Special Extract WS 1442. However, this change was not significant (HR = 0.63, 95% CI = 0.29 – 1.47; p =.72).
It was also observed that there was an increased risk of HF progression at baseline for the treatment group and this risk was significant HR = 6.4 (95% CI = 1.5 – 26.5; p =.011) compared to the placebo group. Progression of HF at six months due to Crataegus Special Extract WS 1442 was non-significant relative to the placebo group (HR = 0.41, 95% CI = 0.14 – 1.08: p =.64). The increase in time taken to reach primary endpoint for a subgroup of 68 subjects who had a LVEF ≤ 35% was non-significant (p =.48) with an odds ratio of 1.42 (95% CI 0.55 – 3.69) for the treatment group compared to the placebo group. Risk of HF progression over time was also shown to change over time according to Kaplan-Meier analysis with the interaction being non-significant (p =.13). Adjustment in various factors that affect HF progression led to a significant increase in hazard risk among the treatment group relative to the placebo (HR = 3.2;95% CI = 1.3 – 8.3: p =.02).
Discussion and Conclusion
It was observed that C. oxycantha does not reduce the risk of HF progression. Instead, HF progression counts seemed to increase at baseline for the C. oxycantha group. A subgroup analysis of patients with LVEF ≤ 35% showed increased HF progression risk throughout the study. Increased HF progression risk was thought to be due to low power since the study used secondary data and the sample size was small. Even the subgroup analysis that had sicker patients was more limited in power since the sample size was very small (n = 68).
Paucity of studies examining effect of C. oxycantha on HF progression made it impossible to make comparisons in the results. The authors of this study cited the “Survival and Prognosis: Investigation of Crataegus Extract WS 1442 in Congestive Heart Failure (SPICE)” [1, p. 591] trial as a comparable study. However, the study had not been published by the time Zick and colleagues [1] conducted this study. The SPICE clinical trial involved almost 3,000 HF subjects whose LVEF was less than or equal to 35% and progression was monitored over a period of 24 months. The same extract (Crataegus Special Extract WS 1442) was used at the same dosage and baseline characteristics same as this current study [9]. The unpublished SPICE trial results indicated a non-significant HF progression difference for treatment and placebo groups but the progression in Zick and colleagues’ [1] study was higher. It was impossible to compare relative risk over time using the SPICE trial since Kaplan-Meier curves were not available. Possible drug interactions may have resulted among patients who were using digoxin, which interacted with flavinoids from the C. oxycantha extracts [10].
Some of the limitations in this study included a small sample size which led to reduced power of detecting hazard ratios. The study was a secondary data analysis, and hence monitoring HF progression may have been limited. With these limitations, Zick et al [1] caution the use of C. oxycantha extracts in treatment of HF progression where close monitoring should be done. Crataegus Special Extract WS 1442 was identified as ineffective in improving HF progression. Instead, there is likelihood that the treatment exacerbates the risk during early phase of HF progression.
Reference List
- Zick SM, Gillespie B, Aaronson KD. The effect of Crataegus oxycantha Special Extract WSS 1442 on clinical progression in patients with mild to moderate symptoms of heart failure, Eur J Heart Fail. 2008;10(6): 587-593.
- Baliga RR , Pitt B, Givertz M. Management of heart failure, Volume 1. Berlin: Springer; 2008.
- Chatburn RL. Handbook for health care research. 2nd edition. Sudbury, MA: Jones & Bartlett Learning; 2009.
- Amdur RJ, Bankert EA. Institutional review board: management and function. Sudbury; MA: Jones & Bartlett Learning; 2002.
- Griffin BP, Topol EJ, Nair D. Manual of cardiovascular medicine. 3rd edition. Lippincott Williams & Wilkins;2008
- Greenberg B, Barnard D, Narayan S. Management of heart failure. Sydney: John Wiley and Sons; 2010.
- Zhang Z, Chang Q, Zhu M, Huang Y, Ho WKK, Chen ZY. Characterization of antioxidants present in hawthorn fruits. J Nutr Biochem. 2001;12: 144–152.
- Aalen OO, Borgan Ø, Gjessing HK. Survival and event history analysis: a process point of view. Berlin: Springer; 2008.
- Holubarsch CJ, Colucci WS, Meinertz T, Gaus W, Tendera M. Survival and prognosis: investigation of Crataegus Extract WS 1442 in congestive heart failure (SPICE)—rationale, study design and study protocol. Eur J Heart Fail 2000;2:431–437.
- Tankanow R, Tamer HR, Streetman DS, et al. Interaction study between digoxin and a preparation of hawthorn (Crataegus oxyacantha) J Clin Pharmacol. 2003;43:637–642.