Meta-analysis of therapeutic effect of Xiaoxuming decoction on stroke

Xin Wang, Ze-Ying Wang

1Shandong University of Traditional Chinese Medicine, Jinan, China;2Weifang Hospital of Traditional Chinese Medicine, Weifang,China.

Abstract

Keywords:Xiaoxuming decoction;Stroke;Meta-analysis.

Introduction

Xiaoxuming decoction(XXMD) originated from Chen Yanzhi's "Essay prescription" [1], Southern and Northern dynasties.The version of XXMD in Simiao Sun's "Thousand-Golden-Prescriptions" is widely used in this study.However, since the Tang and Song dynasties, apoplexy is mostly based on internal wind, and the prescription for treating external wind is almost abandoned but not used.Modern medical scientists put forward the etiological theory of "external wind triggering internal wind" for the incidence of stroke, which has re-improved the status of external wind prescriptions.The treatment of stroke in modern medicine is mainly based on intervention,thrombolysis, anti-platelet aggregation, lipid regulation and neuroprotection [2].Surgical interventional therapy has strict requirements for time window and tissue window, and the target of internal medicine is single.The compound prescription of traditional Chinese medicine supported by unique theoretical system provides a new target and new idea for treatment.In the past five years, there is no updated data for the evaluation of the therapeutic efficacy of XXMD in the treatment of apoplexy.This study is mainly through the literature published in the past five years to systematically evaluate the clinical effectiveness of XXMD in the treatment of stroke in order to provide a basis for clinical application.

1 Materials and methods

1.1 Inclusion and Exclusion criteria

1.1.1 Inclusion criteria

(1) Study type:randomized controlled trial (RCT);(2)Participants:all patients had clear diagnostic criteria for stroke [3] or diagnostic criteria for cerebral infarction,such asChinese guidelines for diagnosis and treatment of Acute Ischemic Stroke[4],Chinese guidelines for Prevention and treatment of Cerebrovascular Diseases[5] and so on;(3) Intervention:the control group was treated with placebo or basic western medicine,including general support, including oxygen inhalation and respiratory support, cardiac monitoring and cardiopathy management, body temperature control,blood pressure and blood glucose control, acupuncture,thrombolysis, antiplatelet therapy, antifibrinolysis,etc.To improve cerebral blood circulation and complications and symptomatic treatment;if necessary,the treatment of lowering intracranial pressure,etc [4].The experimental group was treated with XXMD (oral preparation, unlimited dosage form) or XXMD plus western medicine routine treatment.(4)Outcome indicators:at least one of the following:score of the National Institutes of Health Stroke scale(NIHSS), stroke score, stroke efficacy [3], neurological impairment efficacy or Barthel index (BI).

1.1.2 Exclusion criteria

(1) Literature on animal experiments, reviews,summaries of experience, and case reports;(2) Repeated publications (only one article is selected,and the data integrity is preferred);(3) The results do not provide data that can calculate relative risk(RR), 95% confidence interval (CI) and standardized mean difference (MD);(4) literature with inconsistent intervention measures.

1.2 Retrieval strategy

Computer literature retrieval is carried out according to the principle of evidence-based medicine retrieval strategy (PICOS), searching China knowledge Network (CNKI), China Biomedical Literature Database (CBM), VIP Chinese Journal Database (VIP),Wanfang Database, Pubmed data Base.According to the previous literature research, in order to update the latest evidence-based results, the retrieval time is set from January 1, 2015 to March 1, 2020.Chinese search subject words:Xiao Shu Sheng Tang, Stroke.English search strategy see supplementary material 1.

1.3 Document screening and data extraction

The repetitive literature was eliminated by endnote, the literature was studied and managed by knowledge net,and the data extraction table was made by Excel.The two researchers independently completed the literature screening according to the pre-established inclusion and exclusion criteria, and carried out literature screening and data extraction and input.the input data mainly include the first author, publication time, sample size,baseline (age, male-to-female ratio, etc.), diagnostic criteria, intervention measures, course of treatment,outcome indicators and the results are cross-checked.If there are differences, the communication authors will participate in the discussion and settlement.

1.4 Document quality evaluation

Cochrane 5.2 bias risk assessment tool was used to evaluate the methodological quality of the included study.The evaluation contents included:1 random sequence generation, 2 allocation concealment, 3 blindness to subjects and testers, 4 blindness to outcome evaluators, 5 incomplete outcome data, 6 selective reports and 7 other biases.Each bias risk is divided into three grades:"low risk", "unclear" and "high risk".The included literature is evaluated independently by two researchers, and disputes are discussed and resolved by the correspondent authors.

1.5 Statistical methods

The research uses the RevMan5.3 software provided by Cochrane collaboration network for Meta analysis,the counting data uses the RR value, the measurement data uses the MD value, and calculates the 95%CI.The heterogeneity of the included study was tested ifI2≤ 50% andP＞ 0.1, it shows that the heterogeneity between groups is small, and the fixed effect model is used for analysis;ifI2＞ 50% orP≤ 0.1, it shows that the heterogeneity between groups is large, and the source of heterogeneity is found by subgroup analysis or sensitivity analysis.if the heterogeneity can not be eliminated, the random effect model is used to analyze the effect.Funnel chart was used to analyze publication bias.If the funnel chart is roughly symmetrically distributed, it is stated that there is no publication bias or the publication bias is small;if the funnel chart distribution is asymmetrical, it indicates that there is publication bias.

2 Result

2.1 Results of literature retrieval

A total of 109 articles were obtained by computer search, including 24 articles in CNKI, 28 articles in VIP, 34 articles in Wanfang, 0 articles in Pubmed,23 articles in CBM, and 0 articles in other ways.70 repetitive literatures were screened out, 2 animal experiments and 1 clinical experience were screened out by reading topics and abstracts.The full text read the remaining 36 articles, and finally included 22 articles for Meta quantitative analysis.See Figure 1.

Figure 1 Flow diagram of the meta-analysis

2.2 General characteristics of the included study

A total of 1947 cases were enrolled in 22 studies[6-27], including the experimental group (n = 975) and the control group (n = 972).The basic characteristics included in the study are shown in Table 1.

Table 1 Characteristics of 22 included trials

*T:experimental group;C:control group;W:Western medicine basic treatment;XXMD:Xiaoxuming decoction;GGD:Gegen Decoction.① Chinese Guidelines for the Diagnosis and Treatment of Acute Ischemic Stroke 2010; ②1996 Stroke Disease Diagnosis and Efficacy Evaluation Standards;③Cerebral Infarction and Cerebral Hemorrhage Integrated Traditional Chinese and Western Medicine Diagnostic Standards (Trial) 2006;④ Chinese Guidelines for the Prevention and Treatment of Cerebrovascular Diseases; ⑤ Acute Ischemia Guidelines for the Early Treatment of Patients with Stroke:A Guide for Health Professionals of the American Heart Association/American Stroke Association; ⑥ 1996 Diagnosis Essentials of Various Cerebrovascular Diseases;⑦ 1994 Stroke Disease Diagnosis and Efficacy Evaluation Criteria;⑧Stroke Disease Diagnosis and Efficacy Evaluation" Standards 2015;⑨Guidelines for the Diagnosis and Treatment of Acute Ischemic Stroke in China, 2014.a.NIHSS score b.TCM disease score c.TCM clinical efficacy d.Neurological impairment to improve clinical efficacy e.Barthel index BI f.Modified Rankin scale (mRS).

2.3 Quality evaluation

12 of the 22 literatures [6,9,10,14-19,21,24,26]applied the random number table method, one [22]was grouped according to the odd and even numbers of the date of admission, one [23] used the lottery method, and 8 [7,8,11-13,20,25,27] did not explicitly mention the specific application of the random method.It is only described as "randomly divided into".All the literatures did not describe the application of distribution concealment and blind method, considering that there was no application of placebo in the design,blind evaluation was high risk.Four [9,10,13,25]literatures described data shedding and dealt with them accordingly, and finally did not affect the statistical results, and the data of the remaining 18 articles were complete.There was no obvious bias of selective outcome in all the literatures, and other sources of bias were not discussed.See Figure 2,3.

2.4 Results of Meta analysis

2.4.1 NIHSS scale

A total of 16 literatures [6-8,11,12,14,15,18-24,26,27]were included for heterogeneity test, and the results showed thatP＜0.00001,I2=87%.The sensitivity analysis was carried out by eliminating the literature one by one, and the residual effect was combined,which did not change the meaning of the data, and no obvious heterogeneity sources were found, indicating that the results were robust.Therefore, the random effect model was used, and the results showed thatP＜0.00001, 95%CI (-3.37,-2.19).The results showed that the NIHSS score of stroke patients in the experimental group was significantly better than that in the control group, and the results were statistically significant.See Figure 4.

Figure 2 Risk of biassummary

Figure 3 Risk of biasgraph

Figure 4 Forest plot of NIHSS

2.4.2 Stroke points

The heterogeneity test of 9 studies [7,9,13,14,21-23,25,26] showed that there was a high degree of heterogeneity among the literature (P=0.000 01,I2=81%),.Sensitivity analysis was used to find the source of heterogeneity, one by one was removed and combined with the residual effect, and the results were stable.the results showed that the heterogeneity of one article was significantly changed, as shown in Table 2, the patient in this study was in the sequelae period.Considering the different stages of the disease, there is a difference in the score of stroke.After excluding the literature, the fixed effect model was used.The results showed that the stroke score of the experimental group was significantly better than that of the control group(95%CI (-2.71 ,-1.85),P＜0.00001).The results were statistically significant.See Figure 5.

Table 2 Sensitivity analysis for stroke points

Figure 5 Forest plot of stroke points

2.4.3 BI index

The heterogeneity of 7 RCT included in BI index[9,10,12,13,17,24,26] was analyzed.The results showed that the heterogeneity between groups was small (P＝0.40,I2＝3%).The fixed effect model was used to combine the effect, 95%CI (6.15,8.50) and there was little difference after using random effect model (95%CI (6.02, 8.50)).It shows that the result is robust.The results showed that XXMDplus western medicine basic treatment significantly improved BI index compared with simple western medicine basic treatment, and the difference was statistically significant (P＜0.00001).See Figure 6.

Figure 6 Forest plot of BI index

2.4.4 Clinical effective rate

12 articles of RCT [7-11,14,16,19-22.25] were included in the heterogeneity analysis of the efficacy of stroke.The results showed that there was no significant heterogeneity between the two groups（P=1,I2=0%).The fixed effect model combined with effect was used.The results showed that the improvement rate of stroke score in the experimental group was 3.58 times higher than that in the control group, and the difference was statistically significant (RR=3.58, 95%CI (2.40,5.36),P＜0.00001).

According to the stage of the disease, the patients were divided into two groups:acute stage (within two weeks after onset) and non-acute stage.The results showed that there was no significant heterogeneity between the two groups (I2=0%), in the non-acute group (RR=3.74,95%CI (1.69，8.28)) and in the acute group (RR=3.80,95%CI (2.12，6.79)).According to the subgroup analysis of the disease stage, 7 of the 12 articles were not classified, 4 were classified as wind-phlegm stasis type, and 1 was wind-phlegm upward disturbance type.Among them, the non-typing group RR=3.86,95%CI (2.20, 6.78) and the wind-phlegm stasis group RR=3.44.95%CI (1.81, 6.56).There was no significant heterogeneity in both groups (P＜0.00001).See Figure 7.

2.5 Bias analysis

A bias analysis was conducted on 12 RCT articles that were included in the clinical efficiency of stroke[7-11,14,16,19-22,25].The funnel chart was used to identify publication bias.It can be seen from figure 9 that the funnel chart is basically symmetrical, and most of them are located above the funnel chart, suggesting that there is a small risk of bias in the report.See Figure 8.

Figure 7 Forest plot of clinical effective rate

Figure 8 Bias risk funnel plot

2.6 Safety evaluation

It was found that 12 of 22 articles [6,8,11,12,15,18,19,21,22,24,25,27] did not describe the results of safety evaluation, 7 [7,9,10,13,16,17,26] reported no obvious adverse reactions in terms of liver and kidney function and three major routines, and 3 RCT[14,20,23] reported adverse reactions such as diarrhoea,nausea, dizziness and abdominal pain.However, there was no significant statistical difference between the experimental group and the control group.Based on the above results, it is considered that there are few known adverse reactions, but more complete and comprehensive studies are needed to improve the data of adverse reactions.

Discussion

The essence of apoplexy is based on deficiency and reality [28].In XXMD, there are Mahuang (Ephedrae Herba), Fangji (Stephaniae Tetrandrae Radix),Xingren(Semen Armeniacae Amarum), Renshen(Ginseng Radlx), Huangqi (Astragali Radix), Gui xin (Cinnamomi Ramulus), Gancao (Glycyrrhizae Radix), Shaoyao (Radix Paeoniae Alba), and Chuanxiong (Chuangxiong rhizoma), Fangfeng(RadixSaposhnikoviae), and 12 herbs containing Mahuang decoction and Guizhi decoction to take its "switch"meaning.

Modern studies have shown that the efficacy of XXMD is closely related to its ability to regulate lipid metabolism, which can significantly reduce the level of blood lipids in patients [20,30].Its chemical composition acts on the vasomotor G protein coupled receptor (GPCR) target to achieve the effect of vasodilation [31].XXMD can protect the structure and function of the blood-brain barrier by regulating the function of mitochondria, reducing the activity of nitric oxide synthase, the content of NO and the levels of SOD and MCP-1, so as to play a neuroprotective effect and reduce cerebral ischemia-reperfusion injury [32-35].

The previous study [36,37] showed that the curative effect of XXMD group was 1.17 and 1.15 times higher than that of the control group.In this paper, through the literature published in the past five years and the combination of data.It is found that compared with the routine treatment of western medicine, XXMD can significantly improve the neurological deficit score,improve the BI index, reduce the stroke score and improve the clinical effective rate, and the results are statistically significant.In terms of clinical effective rate, the effective rate of XXMD in acute stage was 1.02 times higher than that in non-acute stage.After the fine syndrome type was divided into groups, it was not found that the effective rate of the wind-phlegm stasis type group was significantly higher than that of the non-type group, suggesting that XXMD may have a higher curative effect in the syndrome types other than the wind-phlegm stasis type.

However, this study also has corresponding limitations.First, TCM decoction is difficult to produce placebo with the same odor and color, and there are difficulties in the realization of blinding method,which makes the literature quality evaluation difficult to further improve.The application of capsules and other preparations may solve this problem, but there is still doubt and research on the difference in curative effect between capsule and decoction.Second, in the course of the study, it is found that the current research inclusion criteria and other experimental design are not rigorous, such as the lack of clear diagnostic criteria,which limits the number of included literature, which may have an impact on the final results.Third, there is less description of the safety of TCM to quantify the safety of TCM with clear laboratory indicators,resulting in incomplete evaluation of the compound prescription of traditional Chinese medicine.Future clinical research should improve and fine safety indicators to make the evaluation of TCM compound prescription more comprehensive.Fourth, the core of the treatment of TCM is syndrome differentiation and treatment, and most randomized controlled trials have no restriction on syndrome types.XXMD is more suitable for which syndrome type or in which period of the disease can achieve better curative effect.More quality, strict design and reasonable research are still needed to improve its evidence-based basis.