Xinmailong Injection for Improvement of Cardiac Function in Patients with Heart Failure: A Systematic Review and Meta-Analysis

Background. Insect drugs have great potential for treating cardiovascular diseases. Xinmailong (XML) injection, a bioactive composite extracted from Periplaneta americana (a species of cockroach), was wildly used in treating heart failure in China. This meta-analysis aimed to assess the efficacy and safety of XML injection for the improvement of cardiac function in HF. Materials and Methods. Online literature search for relevant studies was performed using databases including PubMed, EMBASE, Cochrane Library, CNKI, and Wanfang. Left ventricular ejection fraction (LVEF), six-minute walk test (6MWT), and brain natriuretic peptide (BNP) were selected as target outcomes. The analysis was performed using Stata 12.0, and sources of heterogeneity were explored by subgroup analysis and metaregression. Results. 32 studies were included in this meta-analysis after meeting the inclusion/exclusion criteria. The results demonstrated that additional use of XML improved LVEF (WMD = 5.82, 95% CI: 5.52–7.13, 
 
 P
 <
 0.00001
 
 ) and 6MWT (WMD = 51.48, 95% CI: 35.83–67.13, 
 
 P
 <
 0.00001
 
 ) and reduced BNP (WMD = −172.84, 95% CI: −205.79 to −139.89, 
 
 P
 <
 0.00001
 
 ). The results of subgroup analyses and metaregression suggested that XML injection has more cardiac function improvement for middle-aged HF patients than youth, and greater LVEF and 6MWT improvement were associated with higher average age. Conclusions. XML plus conventional treatment demonstrated a significant effect in reducing cardiac dysfunction in HF patients, and age is a potential factor of higher efficacy. Given the heterogeneity and bias of the included RCTs, large, prospective, rigorous trials are still needed.


Introduction
Heart failure (HF) is a global healthcare issue, defined as a severe and terminal-stage symptom after heart disease, and has high patient mortality [1,2]. HF induces structural, neurohumoral, cellular, and molecular dysfunctions and leads to several organ and system dysfunction resulting in complex clinical manifestations [3]. About 12% of individuals over 80 have HF [4]. As of 2015, 8.9% of Chinese population over 35 years of age have HF [5].
Insect drugs, a type of traditional Chinese medicine (TCM), have great potential for the treatment of several cardiovascular diseases. Insect extracts are rich sources of pharmacopeias occurring naturally and have great scientific and medical value [6]. In the TCM theory system, insect drugs have the effect of activating blood circulation and removing blood stasis, which fits the TCM pathogenesis of HF.
Xinmailong injection (XML, commercialized by the Yunnan Teng Yao Pharmaceutical Co., Ltd, China) is a kind of Chinese patent medicine made by extracting effective substances from Periplaneta americana (a species of cockroach) with modern technology [7]. As a kind of insect drug, fresh adult P. americana bodies were dried and processed into powder for the treatment of disease. e P. americana powder has a history of applying for thousands of years that can be traced back to the Ming Dynasty (A.D. 1578). It was recorded in the "Compendium of Materia Medica" for its therapeutic effects of promoting blood circulation, detoxification, and urination [8].
Extracts of P. americana have many effects, such as promoting wound healing, antitumor activity, and treating HF and gastrointestinal ulcers. Research studies obtained the chemical compounds from P. americana, which were polyhydric alcohols, organic acids, alkaloids [9], divinyl sulfide, noradrenaline, ketone compounds [10,11], adenosine, inosine, protocatechuic acid, and pyroglutamate acid [12]. ese extracts play a specific role in the treatment of diseases with their function of antibacterial, antiviral, and antitumor activity and enhancement of immune function [8]. XML injection as a medicine has a positive inotropic activity, improves microcirculation, dilates pulmonary vessels, induces diuresis, has antiarrhythmia function, and inhibits free radical damage [13].
XML injection has achieved beneficial curative effects in clinical studies and can improve HF patients' cardiac function [14]. However, few studies systematically evaluate the effectiveness of XML injection and the quality of researches. erefore, we performed this systematic review of available randomized controlled trials to evaluate the efficacy of XML on the improvement of cardiac function in HF patients. Left ventricular ejection fraction (LVEF), sixminute walk test (6MWT), and brain natriuretic peptide (BNP) were three typical indicators of measuring cardiac functions and were selected as target outcomes in this review. We performed several subgroup analyses and metaregression to find out the facts that caused the heterogeneity among articles.

Materials and Methods
e present meta-analysis was performed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 2015 statement [15,16] and was registered on PROSPERO (CRD42020163716).

Database and Literature Search Strategies.
Online literature search for relevant studies was performed using five databases, which included PubMed, EMBASE, Cochrane Library, Chinese National Knowledge Infrastructure (CNKI), and Wanfang. Studies were reviewed and selected by two experienced investigators. Relevant data were extracted by an additional two independent investigators. Any discrepancies in data extraction or literature review were resolved by consensus or by consulting a third reviewer. All relevant data from the literature review were extracted from the aforementioned databases from inception to April 2020. e following keywords were searched in various combinations (Xinmailong OR XML OR Xin Mai Long OR XMLI OR Xinmailong Injection) AND (heart failure OR HF OR cardiac failure OR Chronic Heart Failure OR CHF) AND (randomized controlled trial OR RCT).

Inclusion
Criteria. All patients from the selected studies met the internationally accepted criteria for the diagnosis of HF. Only human studies for random clinical trials (RCTs) were considered. No etiology, ethnic group, severity, or course of disease were considered as exclusion criteria. In the experiment group, XML injection plus conventional therapy was considered as a treatment strategy. On the basis of the conventional therapy of western medicine, patients in the control group had placebo or not. Measurable outcomes were LVEF, 6MWT, and BNP.

Exclusion Criteria.
e exclusion criteria were (1) non-RCTs; (2) animal studies, mechanistic studies, case reports, and reviews; (3) patients with acute heart failure and severe liver and kidney disease; (4) use of other TCM formulas; (5) studies with unacceptable trail designs and inappropriate statistical methods; and (6) duplicate publications.

Data Extraction and Quality Assessment.
Clinical data and adverse events were collected and cross-checked by two independent investigators. Any disagreements were resolved through discussion or by consulting a third investigator.
e following data were extracted from studies: patient characteristics, details, differences in intervention between the control and experimental groups, outcome measures, and results from the items listed in the inclusion criteria. Methodological quality was assessed based on the Cochrane Handbook. e included RCTs were assessed for (1) random sequence generation, (2) allocation concealment, (3) blinding of participants and personnel, (4) blinding of outcome assessments, (5) selective outcome reporting, (6) incomplete outcome data, and (7) other potential sources of bias [17].

Data Synthesis.
Effect-size calculations and meta-analytical statistics were performed using Stata 12.0 obtained from its official website (http://www.stata.com). Weighted mean differences (WMD) with 95% CI were expressed for continuous variables and were analyzed using the inversevariance method. Heterogeneity across the trials was evaluated using the Cochran Q test and the Higgins I2 test [18]. Only fixed-effect models were used for both dichotomous and continuous variables, unless the results of pooled analyses showed significant heterogeneity (P < 0.10 and I2 > 50%). A random-effect model was used for significant heterogeneity. Potential publication bias was assessed by Begg's and Egger's tests [19]. A 2-tailed P value of less than 0.05 was set for statistical significance. Subgroup analyses and metaregression were performed for sensitivity analysis. Metaregression sensitivity analyses were performed using the "metareg" macro available in the Stata statistical package. Gender, average age, age section, pattern identification, duration time, course of disease, publication year, and sample size were set as the covariances of metaregression analyses.

Description of Selected
Studies. 388 potentially eligible publications were retrieved after the primary search from the eight databases. Of which 32 RCTs were included for the meta-analysis ( Figure 1). Except for one selected study that was published in English, the rest studies were published in Chinese. All selected studies were conducted in China and included 3730 patients. Characteristics of the included studies are listed in Table 1.
As shown in Table 2, all 32 selected studies mentioned random sequence generation and no selective reporting bias and 9 studies mentioned allocation sequence and concealment using random number tables.
ree studies cited blinding of participants and personnel, and no study mentioned blinding of outcome assessments. 4 studies reported patient attrition; among these, 2 studies reported no dropouts and 2 studies mentioned attrition rates.

Metaregression
Results. e results of metaregression (Table 5) suggested that the age section and average age could be two main sources that contributed to the heterogeneity of LVEF and 6MWT. Unfortunately, gender, pattern identification, duration time, course of the disease, publication year, and sample size showed no significant effect on LVEF and 6MWT, and none of these indicators listed had a statistically significant effect on the BNP level. e significant regression effect of age section (adj R 2 � 26.75%, P � 0.013 for LVEF; adj R 2 � 69.25%, P � 0.003 for 6MWT) met the results of subgroup analyses, which further indicated that middle-aged HF patients with XML administration could have higher LVEF and 6MWT improvement than the youth. e metaregression results based on average age (adj R 2 � 16.54%, P � 0.049 for LVEF; adj R 2 � 36.39%, P � 0.048 for 6MWT) suggested that greater LVEF and 6MWT improvements were associated with higher average age ( Figure 5).  Table 6, 7 adverse events were observed between the experimental and control groups. ese included cutaneous pruritus (3/0), palpitation (1/4), dizziness (2/0), headaches (0/1), nausea (2/1), leukocytosis (0/2), and hypotension (1/0). All symptoms reported were resolved during study observation, and no deaths related to XML administration were reported.

Summary of Main Results
. XML injection is a bioactive composite extracted from P. americana. e injection form can avoid people's visual aversion and improve the utilization ratio of bioavailability for its advanced extraction process. XML injection has been widely used in China for decades. Studies provided evidences to support XML as an effective administration to HF patients that XML injection can mitigate epirubicin-induced cardiotoxicity via activating the PI3K/Akt signaling pathway and inhibiting the Erk1/2 and P38 MAPK signaling pathways [51]. XML can also activate T-type calcium channels and inhibit Na+/K + -ATPase to increase intracellular calcium levels when treating HF [52]. Besides, the pharmacological research on XML has being carried out continuously and many components and functions of XML are being explored [12]. Despite its treatment effects on HF, it can also repair damaged skin and treat gastric ulcers [8]. is systematic review provided evidence that additional XML injection treatment demonstrates better therapeutic effects than conventional treatment on improving cardiac function in patients with HF evidenced by the increase in LVEF and 6MWT and reduction in BNP. Subgroup analyses provided evidence that the age section is a source of causing significant differences between groups. Under XML injection administration, the cardiac function of middle-aged patients showed better improvement than young patients (average > 65 as middle aged, average < 65 as youth). Further metaregression demonstrated that age section and average age showed significant heterogeneity in LVEF and 6MWT and  Evidence-Based Complementary and Alternative Medicine greater improvement of LVEF and 6MWT was connected to a higher age. e above evidences lead to the conclusion that XML injection may be more suitable for middle-aged patients and its efficiency increases by age, while it has a considerable improvement in HF patients of all age ranges. e pattern identification presented no potential heterogeneity in both subgroup analysis and metaregression, indicating XML injection, as a single-substance extract, has fixed therapeutic efficacy on HF without pattern identification, unlike other Chinese patent drugs that focus on TCM syndromes. Unfortunately, the rest covariances (gender, duration time, course of disease, publication year, and sample size) may contribute no significance to the heterogeneity.

Adverse Events.
No obvious adverse reactions of Xinmailong injection were found before marketing. With the expansion of users after marketing, many adverse reactions have been reported gradually. In previous reviews, the incidence of adverse reactions was 7.6% [53]. e most common symptoms are pruritus, slight blood pressure rise, nausea, and vomiting. In this review, adverse effects of XML injection were also collected from included studies, of which 2 studies mentioned skin itching [14,36], each study mentioned hypertension [37] and nausea [36]. Other studies also described palpitation, dizziness, headache, and leukocytosis of HF patients during XML administration. ere is no case of liver and kidney function damage in the clinical study of Xinmailong. On the contrary, there is a protective effect of XML injection on the early renal function damage caused by grass carp bile [54] and extracts of P. americana have a protective effect on liver damage, liver fibrosis, and hepatitis [55]. Although all the adverse reactions mentioned above can be relieved by rest or slowing down the intravenous drip, no lethal events or major adverse cardiovascular events were observed. XML seemed to be generally safe for HF treatment, but the evidence is still insufficient to make a decisive conclusion on safety. Attentions are still needed in practical application, as the safety of TCM herbal injections has been increasingly concerned by both medical workers and the public.

Limitations.
Study limitations in this review were mostly due to the risk of bias [56] of included studies. Many studies have no description of the necessary process of clinical trials, and therefore the evaluation of these studies   Evidence-Based Complementary and Alternative Medicine related to unclear risk of bias. Besides, current research studies focus on the total effect of XML injection, while paying less attention to comprehensive assessment based on basic data collection of different types of patients, which brought difficulty to follow up subgroup analyses and metaregression analysis. Studies in the future could be designed with an emphasis on confirmation and further exploration of age grouping, as average age had been explored to be a source of heterogeneity.
Concerning data extraction, the included trials used different diagnostic and measurement criteria for HF patients. ese different measurement criteria and the use of different instruments affect study comparisons and hence the meta-analysis. Treatment based on pattern identification is a key characteristic of TCM [57]. Application of Chinese patent medicines without the implementation of TCM pattern identification could be a matter of heterogeneity and subsequently reduce the reliability of meta-analysis results. Subgroup analyses in this review, unfortunately, found no heterogeneity based on pattern identification. Further studies are needed for conducting patient selection in which pattern identification is classified.
e results of Begg's and Egger's tests demonstrated no significant asymmetry and suggested no potential   10 Evidence-Based Complementary and Alternative Medicine publication bias in this review [58]. Future clinical studies investigating XML treatment for HF should be designed with appropriate controls and better study designs using large patient cohorts, multicenter and prospective. Heterogeneity within the study should be addressed and include appropriate subgroup classification.

Conclusion
is review demonstrated that combinational use of XML and conventional treatment may demonstrate better therapeutic effects on improving cardiac function in patients with HF, and age is a potential factor of higher efficacy. Study limitations included the low quality of the selected trials. Additional large, multicenter, prospective clinical trials are warranted to validate our findings.

Data Availability
e data used to support this systematic review are included within the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.

Authors' Contributions
Wang XL, Ruan XF, and Sun YL proposed the meta-analysis and designed the study. Qiang TT, Li YP, and Ruan XF performed publication review. Li YP, Sun YL, and Wang XL performed data extraction and analysis. Sun YL wrote the manuscript.