Efficacy and Safety of Fuzi Formulae on the Treatment of Heart Failure as Complementary Therapy: A Systematic Review and Meta-Analysis of High-Quality Randomized Controlled Trials

Objective Heart failure is a major public health problem worldwide nowadays. However, the morbidity, mortality, and awareness of heart failure are not satisfied as well as the status of current treatments. According to the standard treatment for chronic heart failure (CHFST), Fuzi (the seminal root of Aconitum carmichaelii Debx.) formulae are widely used as a complementary treatment for heart failure in clinical practice for a long time. We are aiming to assess the efficacy and safety of Fuzi formulae (FZF) on the treatment of heart failure according to high-quality randomized controlled trials (RCTs). Methods RCTs in PubMed, Cochrane Library, China National Knowledge Infrastructure (CNKI), Chinese Scientific Journals Database (VIP), and Wanfang Database were searched from their inception until June 2019. In addition, the U.S. National Library of Medicine (clinicaltrials.gov) and the Chinese Clinical Trial Registry (http://www.chictr.org.cn) were also searched. We included RCTs that test the efficacy and safety of FZF for the treatment of heart failure, compared with placebo, CHFST, or placebo plus CHFST. The methodological quality of included studies were evaluated by the Cochrane Collaboration's tool for assessing risk of bias. RCTs with Cochrane risk of bias (RoB) score ≥4 were included in the analysis. The meta-analysis was conducted through RevMan 5.2 software. The GRADE approach was used to assess the quality of the evidence. Results Twelve RCTs with 1490 participants were identified. The studies investigated the efficacy and safety of FZF, such as FZF plus the CHFST vs placebo plus CHFST (n = 4), FZF plus CHFST vs CHFST (n = 6), FZF plus digoxin tablets (DT) plus CHFST vs placebo plus DT plus CHFST (n = 1), and FZF plus placebo plus CHFST vs placebo plus DT plus CHFST (n = 1). Meta-analysis indicated that FZF have additional benefits based on the CHFST in reducing plasma NT-proBNP level, MLHFQ scores, Lee's heart failure scores (LHFs), and composite cardiac events (CCEs). Meanwhile, it also improved the efficacy on TCM symptoms (TCMs), NYHA functional classification (NYHAfc), 6MWD, and LVEF. Adverse events were reported in 6 out of 12 studies without significant statistical difference. However, after assessing the strength of evidence, it was found that only the quality of evidence for CCEs was high, and the others were either moderate or low or very low. So we could not draw confirmative conclusions on its additional benefits except CCEs. Further clinical trials should be well designed to avoid the issues that were identified in this study. Conclusion The efficacy and additional benefits of FZF for CCEs were certain according to the high-quality evidence assessed through GRADE. However, the efficacy and additional benefits for the other outcomes were uncertain judging from current studies. In addition, the safety assessment has a great room for improvement. Thus, further research studies are needed to find more convincing proofs.


Background
Heart failure is a public health problem in clinical cardiology nowadays. There are about 3% to 5% people suffering from heart failure (HF) according to epidemiological surveys worldwide. The morbidity of HF is about 2% in developed countries and 1.3% in China, which means nearly 18 million people are having HF in China. Meanwhile, the proportion of people over 65 years with HF is about 10%, indicating that people are more likely to suffer from heart failure when they get older. The 1-year mortality rate is ranged from 20% to 40% in different countries in patients who are readmitted for heart failure, about 50% of HF patients died within five years after diagnosis, and the 10-year mortality is more than 90%. It is amazing that the mortality of HF is even higher than breast cancer, prostate cancer, colorectal cancer, and other common cancers. The mortality of lung cancer is only 18.4%, which is the highest in cancers. However, most of the patients with HF are still having a superficial knowledge on HF. Although HF is more common in the elderly, about 30% of the patients mistake the symptoms of HF as normal aging phenomena [1][2][3][4][5].
In general, current treatments for HF are relatively fixed, including diuretics, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, beta-blockers, aldosterone receptor antagonists, digitalis, and vasodilating agents, according to the guidelines for CHFST. However, these drugs only achieve good short-term effects. This is why the number of deaths and readmission resulting from HF continues to rise despite of the advances in drug treatment strategies for HF. From the perspective of TCM, the primary cause of HF is the yang deficiency of heart that results from Qi inadequacy and blood stasis in general consideration. Based on the primary cause of HF, many Chinese herbs have demonstrated safety and efficacy in the management of HF in both animal models and humans [6][7][8][9][10][11][12]. Fuzi is widely used in the treatment of HF as an adjuvant therapy in our long-term clinical practice whether in decoction or other dosage forms. Since Fuzi is a major Chinese herb for restoring yang for resuscitation, it is contained in FZF, which are made into various kinds of forms such as granule, capsule, pill, oral decoction, and injection for convenient use. It is beneficial to HF patients in relieving symptoms and improving indicators despite of its toxicity as known. This study aimed at investigating the efficacy and safety of FZF on the treatment of HF and providing reference for clinical diagnosis and treatment.

Methods
This systematic review and meta-analysis are based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement search strategy [13].

Inclusion Criteria
(1) Type of participants: researches involving adult patients with any type of HF.
(2) Type of study: only RCTs that assessed the efficacy and safety of Fuzi for the treatment of HF were eligible. (3) Type of intervention: Fuzi must be included in the herbal formula used in the experimental group. There were no restrictions on the dosage forms of the drug (e.g., decoction, injection, pill, and capsule), dosage, frequency, or treatment time. Medications of the control group medications including placebo, CHFST, and placebo plus CHFST were also accepted. (4) Type of results: the efficacy of Fuzi on the treatment of HF was evaluated through primary outcomes of plasma NT-proBNP level and the efficacy of TCM symptoms (TCMs). Secondary outcomes included the efficacy of NYHA functional classification (NYHAfc), LVEF, 6-minute walk distance (6MWD), composite cardiac events (CCEs) such as death and readmission, the Minnesota Living with Heart Failure Questionnaire (MLHFQ) scores, and Lee's heart failure scores (LHFs). The safety was evaluated through adverse events and laboratory indexes. (5) RoB scored≥4 points [14].

Exclusion Criteria.
If the above conditions were not met, the study was excluded. In addition, the following literatures were also excluded: (1) Duplicate publications (2) Animal experiments, mechanism, studies, reviews, protocols, experience, and case reports otherwise, the random effect model (REM) was a more plausible match. However, since there were different ingredients in FZF, discrepancies in effect sizes should not be ignored. Therefore, whatever the heterogeneity index I 2 was, we conducted REM to balance the effects of each study. Sensitivity analysis was performed by changing analysis combination to explore the impact of confounding factors.
Meanwhile, in consideration of the differences in interventions and treatments, the subgroup analysis was performed using the Z-test. The differences between treatment groups and control groups were considered to be statistically significant when P < 0.05. If an outcome was reported in more than ten studies, funnel plots and Egger's test were used to examine their publication bias. Finally, we used the GRADE approach to access the strength of the evidence so as to make our results more credible.  [18][19][20][21][22][23][24][25][26][27][28][29]. A PRISMA flow diagram depicted the search process and study selection ( Figure 1).

Primary Outcomes
3.5.1. Plasma NT-proBNP Level. 6 studies evaluated the plasma NT-proBNP level, and a reduction was showed in FZF plus CHFST, compared with CHFST (SMD � − 1.76, 95% CI: − 2.87 to − 0.66, P � 0.002, heterogeneity χ2 � 132.51, P < 0.00001, I 2 � 96%, Figure 2) [18,21,[25][26][27][28]. However, the  quality of the evidence was low (Table 5). So further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. The result of the heterogeneity test was 96%, which indicated they had high heterogeneity, so the statistical analysis was conducted with REM.

Study or subgroup
Dong [25] Liu et al. [20] Wang [21] Wei [27] Zhu [28] Zou [18] Total (95% CI) Heterogeneity: tau 2 = 1.80; chi 2 = 132.51, df = 5 (P < 0.00001); I 2 = 96% Test for overall effect: Z = 3.13 (P = 0.002)   873 per 1000 (756 to 1000) * The basis for the assumed risk (e.g., the median control group risk across studies) is provided. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio. GRADE working group grades of evidence: high quality: further research is very unlikely to change our confidence in the estimate of effect; moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; very low quality: we are very uncertain about the estimate.  [25] Liu et al. [20] Wang [21] Wang et al. [29] Wei [27] Zhu [28] Total events

Subtotal (95% CI) Total (95% CI)
Heterogeneity: tau 2 = 0.00; chi 2 = 0.67, df = 2 (P = 0.72); I 2 = 0% Test for overall effect: Z = 3.06 (P = 0.002) Heterogeneity: tau 2 = 0.01; chi 2 = 15.75, df = 8 (P = 0.05); I 2 = 49% Test for overall effect: Z = 4.54 (P < 0.00001) Test for subgroup differences: chi 2 [20,23,24]. Moreover, the quality of evidence was high according to the GRADE approach (Table 6). [18-20, 23, 24, 29] reported adverse events occurring during the treatment period, and a total of 9.4% (69/735) patients in the experimental groups and 13.4% (101/755) patients in control groups suffered from adverse events (Table 7). Some patients had more than one event. 6 studies [21, 22, 25-28] stated no adverse events happened during the treatment period. 4 studies [18][19][20]23] provided adequate information of the adverse events. 3 studies reported erythra as adverse event in control groups [19,20,29], and the erythra was considered to be anaphylactic reaction to uncertain western medicine. 1 study reported cough as an adverse event in the control group [23], and the cough could be self-remission after medicine withdrawal. 1 study reported chest tightness and heart palpitation as adverse events in the control group [18], and the reason of these adverse events was not mentioned in detail. 1 study reported chills as an adverse event in the control group [29]. The last study did not mention adverse events, but there was no detailed description [24]. No significant abnormal was found in the blood routine, urine routine, liver function, and kidney function test. 1 study reported adverse events related to study drugs, and 20 cases in the experimental group and 23 cases in the control group were without a detail description [24]. This study also described arterial occlusive diseases (1 case in the control group), worsening heart failure (4 cases in the experimental group, and 7 cases in the control group), stroke (1 case in the experimental group and 1 case in the control group), lumbar fracture (1 case in the experimental group), and unknown reasons (2 cases in the experimental group and 3 cases in the control group) as serious adverse events. There was no report of any serious adverse events related to the study drugs. Three studies [21,28,29] reported safety with specific laboratory index, and there was no statistical significance. However, meta-analysis showed the safety of FZF was not satisfied, which needs to be improved in further studies (RR � 0.71, 95% CI: 0.48 to 1.06, P � 0.09, heterogeneity χ2 � 5.11, P � 0.40, I 2 � 2%, Figure 9) [18,20,[22][23][24]29].
3.9. Sensitivity Analysis. The sensitivity analysis did not indicate that the results of any individual study would change the final outcome, which meant that none of the studies can significantly affect the pooled OR and 95% CI.

Summary of Evidence.
The studies of TCM for the treatment of HF have been carried out for quite a long while and most of them, including animal experiments, clinical trials, or pharmacological studies, have indicated the efficacy of TCM at the basis of CHFST [30][31][32]. Fuzi are widely used as a TCM herb in the formulae of TCM based on the essence of heart-yang deficiency of heart failure, such as various decoctions made according to the principles of TCM syndrome, injections, and Chinese patent medicines [18,19,24,25,[33][34][35][36][37]. However, the poor methodological quality and small sample sizes prevented the author from making firm conclusions. Despite the current evidence and wide application of FZF in clinical practice, our systematic review tried to determine the efficacy and safety of FZF on heart failure through analyzing 12 high-quality RCTs with 1490 participants. The present study indicates that FZF provide statistical benefits in improving the efficacy on NYHAfc and LVEF and reducing plasma NT-proBNP level as well. Furthermore, they can also improve the patients' prognosis and life quality and reduce the risks of patients in death and readmission for heart failure. However, despite the low strength of current evidence, the benefits were almost uncertain except the benefits for death and readmisssion. In addition, FZF appeared to be generally safe and well tolerated with mild adverse reactions. Although 6 reported adverse events, only 45 out of 1490 cases (3.0%) had adverse events possibly related to FZF without powerful evidence. No significant differences were found on laboratory indicators. Current evidence supported that FZF could be an adjuvant therapy for the treatment of heart failure on the basis of CHFST in improving death and readmission.

Limitations.
There are several limitations in our primary studies as well. Firstly, the participants included were all with chronic heart failure. Therefore, the efficacy of FZF for acute heart failure are not clear. Thus, further research studies on FZF for the treatments of acute heart failure are needed. And the sample sizes of some studies were small, which might have influence on the results. Meanwhile, because all the participants in the studies came from China, the ethnic differences and regional differences were indefinite. Secondly, the components of FZF varied in producing area, species, processed methods, dosages, forms, and decocting. For the processed methods, 1 study used Heishunpian [24], 1 study used Baifupian [27], and 1 study used Fupian [22] in FZF, while the other studies used Fuzi [19,25,26,28,29], Shufuzi [18,20,23], and Paofuzi [21] as described. As for the decoction methods, 3 FZF needed to be decocted, 1 FZF was not described specifically [27], and only 1 study described the decoction method clearly [28]. Current studies indicate that decocting methods have some effects on the toxicity of Fuzi and different dosages may also have unknown influences uncover [38,39]. Thirdly, although we included the high-quality RCTs according to a cumulative score of at least 4 out of 7 based on the Cochrane RoB tool domains and revised Jadad scale, the methodological details were still not adequate in some studies. Only 1 study [23] described a proper method of allocation concealment, and 6 studies [19,20,23,24,26,29] employed the blinding procedure. Some studies were unable to be blinded, due to the fact that TCM is special in color, smell, and taste, which were difficult to be changed or covered. And most kinds of FZF were so different from Dong [25] Wang [21] Wei [27] Wang [21] Zhu [28] Subtotal (95% CI)

Subtotal (95% CI)
Total (95% CI) 10 32 Total events Total events Heterogeneity: tau 2 = 0.00; chi 2 = 0.44, df = 3 (P = 0.93); I 2 = 0% Test for overall effect: Z = 3.23 (P = 0.001) Heterogeneity: tau 2 = 0.00; chi 2 = 0.14, df = 2 (P = 0.93); I 2 = 0% Test for overall effect: Z = 4.19 (P < 0.001) Heterogeneity: tau 2 = 0.00; chi 2 = 1.52, df = 6 (P = 0.96); I 2 = 0% Test for overall effect: Z = 5.20 (P < 0.00001) Test for subgroup differences: chi 2 = 0.92, df = 1 (P = 0.34); I 2 = 0% 0. 45     glucose injection. However, no study used a doubledummy technique to reduce the difference of drugs between the experiment and control groups. Improper blinding methods made it difficult to get certain results intentionally or unintentionally, and it is hard to ensure the credibility of study conclusions [40]. In addition, the intervention of trials with inadequate allocation concealment is 18% more "beneficial" than trials with adequate concealment [41]; however, lacking allocation concealment was common. Fourthly, due to traditional culture and the barrier of language, all RCTs were in English or in Chinese and were conducted in Chinese population, which restricts the generalizability of the findings. Finally, the strength of evidence was almost poor according to the GRADE approach, which made the final results less credible.

Implications for Practice.
Modern pharmacological studies on Fuzi were performed (the seminal root of Aconitum carmichaelii Debx) to explain its mechanisms of actions. Fuzi was first recorded in "Shen Nong Ben Cao Jing," which was known as one of the TCM classics for over thousands of years. The herb is good at restoring yang for resuscitation, tonifying fire, and helping yang, removing rheumatism and relieving pain because of its properties: pungent and sweet in flavor, pretty hot in nature, and extremely poisonous especially when the fresh herb was used [42]. So processed products of Fuzi with less toxic effect are commonly used nowadays. The common processed products including Yanfuzi, Heishunpian, Baifupian, Paofuzi, Danfuzi, Shufuzi, Weifuzi, and other decocting pieces processed by ginger juice, glycyrrhiza juice, tofu, etc. The main chemical components of Fuzi are aconitum alkaloids including C-19 diterpenoid alkaloids (aconitine, mesaconitine, and hypaconitin), C-20 diterpenoid alkaloids (songorine and songoramine), and nonalkaloids such as urical, β-sitosterol, daucosterol, and glyceryl monopalmitate. And the biological activities of Fuzi include enhancing myocardial systole, resisting inflammation, relieving pain, resisting tumor, promoting immunity, and influencing metabolism [43]. Most processed products of Fuzi could reduce its' poisonousness, and the chemical components of Fuzi are also changed indeed. Firstly, the toxicity of aconite is significantly reduced. Because aconitum alkaloids contain ester bonds and have thermal instability, and they are hydrolyzed to form monoester alkaloids and protoalkaloids with less toxicity after boiling. Meanwhile, aconitum alkaloids are prone to pinacol rearrangement and pyro-type alkaloids under acidic and heating conditions. The toxicity of aconitum alkaloids is relatively low, and their analgesic and anti-inflammatory effects are still obvious. In addition, the number of fatty alkaloids in Fuzi increased after running and the substitution of long chain fatty acyl and acetyl groups reduced the toxicity greatly [44,45]. In a word, the toxicity of dicarboxylic alkaloids with high toxicity is reduced, and other alkaloids with low toxicity or nontoxicity are increased, so the toxicity of processed Fuzi is greatly reduced. Furthermore, the processed products of Fuzi will not only reduce its cardiotonic, analgesic, and anti-inflammatory effects but also increase the safe dose of Fuzi [46,47]. Although many scholars have proved that Fuzi has a cardiotonic effect on different animal models of HF with its different preparations; further research studies are still needed to clarify the nature of the ingredients of the mixture and the mechanisms of action of different processed Fuzi products.

Implications for Further Studies.
At first, we suggest that the protocol of clinical trials must be registered in clinical trials registry platform such as The U.S. National Library of Medicine (clinicaltrials.gov) and The Chinese Clinical Trial Registry (http://www.chictr.org.cn). Meanwhile, CONSORT 2010 statement should be applied in trial reporting and publication in order to draw normative conclusions for further studies. Secondly, in order to facilitate more reliable comparison of study results, the clinical trials must be well designed according to international standards. The enrollment of participants should be more wide, sample size needs to be appropriately calculated, the randomization principle and allocation concealment should be implemented with more attention, and the standards of measurement results need to be as uniform as possible, as well as the test medication, in order to strengthen the evidence and make results more reliable. The type of acute heart failure should be further studied, which could give precise evidence for clinic. Thirdly, Fuzi, Renshen, Danshen, Fuling, Huangqi, and Tinglizi were the most frequently used herbs in treating HF, which should be considered firstly when formulating optimal formula. Finally, there are large spaces on the exact pathomechanism of migraine, and the pharmacological mechanism of Fuzi remains largely unknown, which should be further investigated.

Conclusion
The efficacy and additional benefits of FZF for CCEs were certain according to the high-quality evidence assessed through GRADE. However, the efficacy and additional benefits for the other outcomes were uncertain based on current studies. Furthermore, the safety assessment has a great room for improvement. Thus, further researches are needed to find more convincing proof.