Efficacy and safety of Nephritis rehabilitation tablet combined with Valsartan for chronic glomerulonephritis: A system review and metaanalysis

Qiao Liu, Tai-Wei Dong, Feng Gao, Yan-Li Guo, Feng Miao, Pei-Feng Wei,?

1. Shannxi University of Chinese Medicine, Xianyang 712000, China

2. The Second Affiliated Hospital of Shaanxi University of Chinese Medicine, Xianyang 712000, China

ABSTRACT Objective: To systematically evaluate the efficacy and safety of Nephritis Rehabilitation Tablet(NRT) combined with valsartan in treatment of chronic glomerulonephritis (CGN). Methods:Computer search databases such as CNKI, CBM, VIP, wanfang, Embase, PubMed and Cochrane library, and find all randomized controlled trials (RCTs) comparing NRT combined with valsartan versus valsartan in treatment of chronic glomerulonephritis. The search time limit is to build the database until October 2020. RCTs were screened according to the inclusion and exclusion criteria. After data extraction and quality assessment, the Cochrane risk of bias tool was used to evaluate the methodological quality of these studies. Meta-analysis was performed by Review Manager 5.2, and GRADE system for evidence quality evaluation. Results: We have identified a total of 24 eligible RCTs with 2082 participants and completed a metaanalysis based on these RCTs. The results of the meta-analysis showed that compared with valsartan, NRT combined with valsartan treatment showed effective curative effect in terms of effective treatment rate (OR = 4.72; 95%CI, 3.67, 6.08; P ＜ 0.00001), 24h urine protein quantification (MD = -0.52; 95%CI, -0.59, -0.44; P ＜ 0.00001), serum creatinine (Scr) (MD= -10.33; 95%CI, -14.00, -6.66; P ＜ 0.00001), Systolic blood pressure (SBP) (MD = -11.42;95%CI, -17.67, -5.17; P = 0.0003), Diastolic blood pressure (DBP) (MD = -6.28; 95%CI, -9.14,-3.42; P ＜ 0.0001), Blood urine nitrogen (BUN) (MD = -0.02; 95% CI, -0.41, 0.37; P = 0.93),plasma albumin (ALB) (MD = 5.05; 95% CI, 4.27, 5.84; P ＜ 0.00001), and adverse reactions(OR = 0.93, 95% CI, 0.54, 1.60; P = 0.78). No serious adverse events were mentioned in these studies. And based on the results of the systematic review, the GRADE system recommended ranking method was used to evaluate the quality of evidence and the recommendation level.The results showed that the level of evidence was moderate and the recommendation intensity was weak recommendation. Conclusions: NRT combined with valsartan has a significant effect on the treatment of CGN, and the treatment effect is better than valsartan alone. There are no obvious adverse reactions during the treatment process. However, due to the generally low quality of the literature included in these studies, and the variability of the evaluation methods of each study, morelarge samples, multi-center, high-quality samples are still needed RCTs are further verified.

Keywords:Nephritis rehabilitation tablet Valsartan Chronic glomerulonephritis Randomized controlled trial Meta-analysis

1. Introduction

CGN is a common disease in the nephrology department, which refers to the morphological and functional damage of glomerular capillaries. Clinically, it mainly appears in the form of nephritis syndrome. The patients usually start with hematuria and proteinuria,accompanied by varying degrees of edema and hypertension. If the disease cannot be effectively controlled in time, it may not be cured,and it will eventually develop chronic renal failure is a serious threat to life and health [1]. Clinically, patients with CGN are mainly treated with angiotensin II receptor antagonist (ARB) to control the progression of their disease [2]. And valsartan is a commonly used ARB clinically, this drug can block the renin-angiotensin system at the receptor level, effectively reduce the pressure in the glomerular blood vessels, reduce proteinuria and slow down the rate of glomerular sclerosis, thereby delaying the development of chronic kidney disease [3]. Although valsartan has a certain therapeutic effect on CGN, due to the influence of many factors, the effect of single medication is not good [4]. Combined with current clinical practice,traditional Chinese medicine has certain advantages in the treatment of CGN, and most patients also use traditional Chinese medicine on the basis of modern medical treatment.

NRT belong to traditional Chinese medicine preparations, mainly containing Chinese herbal medicines such as Alisma orientalis,Chinese yam, black beans, ginseng, Imperata cylindrica and motherwort [5]. In modern pharmacology, NRT combined with valsartan have a synergistic effect, which can regulate the body's immune function, promote the synthesis of albumin in the human liver, repair patients' glomerular cells, and improve patients' renal function. The current clinical practice shows that most patients with CGN are treated with NRT combined with valsartan to achieve significant results, which can promote faster recovery of patients and alleviate the clinical symptoms of patients. Compared with the effect of single medication, it has many advantages [6-7]. However, because of the small sample size, it is difficult to effectively evaluate its efficacy. Therefore, this study used the method of meta-analysis to assess the effectiveness and safety of NRT combined with valsartan in the treatment of CGN, in order to provide objective evidencebased medical evidence for clinical treatment programs.

2. Materials and Methods

2.1. Inclusion Criteria

The inclusion criteria were as follows. (1) All patients in this group were diagnosed with CGN. The diagnostic criteria were:proteinuria, hematuria, edema and hypertension were the main clinical manifestations, and the course of the disease exceeded 3 months. (2) Research protocol: The treatment group was treated with NRT combined with valsartan to treat patients with CGN, and the control group was treated with valsartan or conventional treatment and valsartan for CGN. (3) The main outcome indicators include:Effective treatment rate, 24h urine protein quantification, serum creatinine (Scr), Blood pressure (BP), Blood urine nitrogen (BUN),Plasma albumin (ALB) and Adverse reactions. (4) Research Type:RCTs. (5) The document language is limited to Chinese and English.

2.2. Exclusion Criteria

Studies that meet the following criteria were excluded

(1) The selected candidates have other serious diseases. (2) There are pregnant patients in the study. (3) In addition to NRT and valsartan, the treatment group also received other treatments not applied in the control group. (4) Duplicate publications. (5) Nonrandomized controlled trial. (6) Unable to obtain the full text or the literature with missing data on the above outcome indicators.

2.3. Search Strategy

Two reviewers (Qiao Liu and Taiwei Dong) searched Chinese and English literatures related to RCTs of NRT combined with valsartan in the treatment of CGN. The computer searched 7 databases including CNKI, CBM, VIP, wanfang, PubMed, Embase and Cochrane library. Search for the following terms: “Nephritis rehabilitation tablet” [Title/Abstract] OR “Shen yan kang fu tablet”[Title/Abstract] AND “valsartan” [Title/Abstract] AND “Chronic glomerulonephritis” [Title/Abstract] OR “Chronic nephritis” [Title/Abstract].

2.4. Literature screening and data extraction

Combined with the relevant inclusion and exclusion criteria of the literature, two researchers (Qiao Liu and Feng Gao) independently screened the literature, extracted and sorted out the required data and checked each other. In case of differences, they can be discussed and resolved or a third researcher (Feng Miao) can assist in the judgment. The extracted information includes: the first author and publication Year, random method, number of cases in treatment group and control group, age, gender, specific intervention measures and treatment process, outcome measures.

2.5. Risk of bias assessment and quality assessment

Two researchers (Feng Gao and Taiwei Dong) evaluated the risk of bias in the included literature according to the risk of bias evaluation tool for RCTs recommended by the Cochrane System Reviewer Manual. The methodological criteria and methods of evaluation are as follows: (1) Random sequence generation. (2) Allocation concealment. (3) Blinding of participants and personnel. (4) Blinding of outcome assessment (5) Incomplete outcome data. (6) Selective reporting. (7) Other biases. We conduct bias risk assessments for each RCT and classify them as "low risk", "uncertain risk" or "high risk". It was completed by two researchers independently and checked each other. If necessary, they could intervene through a third evaluator (Peifeng Wei) to negotiate and finally reach a consensus.

2.6. Evidence quality and recommendation level

Based on the results of systematic reviews, the GRADE system recommendation ranking method [8] was used to evaluate the quality of evidence and the recommendation level. The quality of evidence is graded as follows: (1) High quality: further research will not change the credibility of the efficacy evaluation results; (2) Moderate quality: further research may change the evaluation results, and the credibility of the efficacy evaluation results will be reduced. (3)Low quality: further research is very likely to change the evaluation results, and the credibility of the efficacy evaluation results is greatly reduced. (4) Very low quality: The results of any efficacy evaluation are very uncertain. Although the evidence based on RCTs was initially rated as high quality, our confidence in this type of evidence may be reduced due to the following five factors: (1) Risk of bias. (2)Inconsistency. (3) Indirectness. (4) Imprecision. (5) Publication bias.The recommendation level is divided into "strong recommendation"and "weak recommendation": strong recommendation indicates that the evaluator is convinced that the intervention has more advantages than disadvantages or disadvantages than benefits, and weak recommendation indicates that the advantages and disadvantages are uncertain. Finally, GRADEpro 3.6 software was used to analyze and chart the quality of evidence, and the recommendation level was given based on the quality of evidence combined with the research theme. The evaluation of the promotion and demotion factors is the responsibility of Qiao Liu and Feng Gao. If there is a dispute,the third researcher (Peifeng Wei) is required to review and reach agreement through discussion.

2.7. Statistical analysis of data

Reviewer Manager 5.2 software provided by Cochrane was used for meta-analysis of the literature. Binary variables use the odds ratio (OR) statistical effect quantity, for the continuous variables with the same measurement method or unit of the same intervention effect, the mean difference (MD) is used. The effect sizes all provide 95% confidence interval (CI). The heterogeneity between the included study results is analyzed using the χ2test and the amount of heterogeneity is quantitatively determined in conjunction with I-square (I2). If the heterogeneity between the studies is small (I2＜50%), the fixed effect model is used; if the heterogeneity is high (I2≥ 50%), the random effect model is used. In addition, if there is obvious heterogeneity between the results of each study, methods such as subgroup analysis or sensitivity analysis should be used for treatment, or only descriptive analysis. A funnel plot is used to assess publication bias.

3. Results

3.1. Basic Information

3.1.1. General Information about Inclusion of These Studies

A total of 221 related documents were obtained during the initial search, including 69 articles in CNKI, 9 articles in CBM, 54 articles in VIP, and 89 articles in Wanfang. After using NoteExpress to eliminate duplicate documents, 120 articles were obtained. After reading the title and abstract, 69 irrelevant articles were eliminated.Then look for and read the full text, and further exclude the participants who did not meet the inclusion criteria, non-randomized controlled trials and non-defined outcomes. Finally, 24 RCTs were included [9-32]. All included trials were conducted in China and published in Chinese. The flow diagram and results of literature screening are shown in Figure 1.

Figure 1 Flow diagram for literature retrieval

3.1.2. Basic characteristics of included literature

The 24 articles are all RCTs, and a total of 2082 patients were enrolled, including 1044 in the treatment group and 1038 in the control group. And the sample size of all studies ranged from 45 to 160. The course of treatment was mostly 4 to 12 weeks, and only one study [18] did not mention the specific course of treatment.Except for one study [29], all other studies described that the treatment combination control group was comparable in gender, age,condition, and course of disease. Among the treatment measures of the control group, 8 studies mentioned conventional treatment,of which 5 studies added anti-infective drugs, and 3 studies only described as conventional treatment, without specifying specific drugs. The intervention measures of the treatment group were added with NRT on the basis of the treatment measures of the control group. And common outcome indicators include effective treatment rate, 24h urine protein quantification, Scr, BP, BUN, ALB and adverse reactions. The detailed basic characteristics of the included studies are shown in Table 1.

3.2. Risk Assessment of Bias in Included Studies

According to the bias risk assessment method recommended by Cochrane, all the studies are described as random grouping, of which 4 studies [11, 17, 30-31] used random number table method to group patients and evaluated as "low risk", and 1 study [15] used computer randomized grouping method to group patients and evaluated as"low risk". One study [24] divided the patients into groups by lottery and was evaluated as "low risk", and one study [29] was described as grouping patients according to the date of admission and was evaluated as "high risk". The rest of the studies did not describe the specific randomization plan and were evaluated as "unclear". All studies did not describe allocation concealment, did not use blinding,and there were no cases of loss to follow-up during treatment and follow-up. The specific details are shown in Figure 2.

Figure 2 Risk of bias assessment of included studies.

Table 1 Basic characteristics of RCTs included in the meta-analysis

3.3. Results of meta-analysis

3.3.1. Effective treatment rate

A total of 21 studies used effective treatment rate as an indicator of curative effect, including a total of 1875 patients [11-13, 15-32]. Metaanalysis was used to compare the effective treatment rate between the treatment group and the control group. After the heterogeneity test, there was no difference between the studies (P =0.97, I2=0%).Meta-analysis using a fixed-effect model shows that the effective treatment rate of NRT combined with valsartan in the treatment of CGN is better than that of valsartan alone (OR = 4.72; 95%CI, 3.67,6.08; P ＜ 0.00001), the difference is statistically significant (Figure 3).

Figure 3 Effective treatment rate: NRT combined with valsartan vs valsartan

3.3.2. 24h urine protein quantification

21 studies with 1851 patients used 24h urine protein quantification as an indicator of efficacy [9-14, 17-29, 31-32]. After the heterogeneity test, there is a high heterogeneity among the studies (P ＜ 0.00001,I2= 87%). Meta-analysis using a random-effect model showed that the combination of NRT and valsartan in the treatment of CGN can lower 24h urine protein quantification more than valsartan alone (MD= -0.52; 95%CI, -0.59, -0.44; P ＜ 0.00001) (Figure 4), the difference is statistically significant. Sensitivity analysis found that excluding 20 studies, the combined effects were still statistically significant and the direction of the forest plot results did not change.

Figure 4 24h urine protein quantification: NRT combined with valsartan vs valsartan

3.3.3. Scr

A total of 21 studies with 1863 patients used Scr as an indicator of efficacy [9-15, 17-23, 25-29, 31-32]. It has considerably high heterogeneity in Scr (P ＜ 0.00001, I2= 92%). Meta-analysis using a random-effect model showed that the combination of NRT and valsartan in the treatment of CGN can lower Scr more than valsartan alone (MD = -10.33; 95%CI, -14.00, -6.66; P ＜ 0.00001) (Figure 5),the difference is statistically significant. Sensitivity analysis found that excluding 20 studies, the combined effects were still statistically significant and the direction of the forest plot results did not change.

Figure 5 Scr: NRT combined with valsartan vs valsartan

3.3.4. BP

10 studies with 942 patients reported the impact on BP [12-13, 17,20-22, 24, 26, 31-32]. It has considerably high heterogeneity in SBP(P ＜ 0.00001, I2= 99%) and DBP (P ＜ 0.00001, I2= 95%), Metaanalysis using a random-effect model showed that in reducing SBP,the combination of NRT and valsartan was better than valsartan alone in the treatment of CGN (MD = -11.42; 95%CI, -17.67, -5.17;P = 0.0003) (Figure 6), the difference is statistically significant.In reducing DBP, the treatment of CGN with NRT combined with valsartan is better than valsartan alone (MD = -6.28; 95%CI, -9.14,-3.42; P ＜ 0.0001) (Figure 7), and the difference is statistically significant. Sensitivity analysis found that excluding 9 studies, the combined effects were still statistically significant and the direction of the forest plot results did not change.

Figure 6 SBP: NRT combined with valsartan vs valsartan

Figure 7 DBP: NRT combined with valsartan vs valsartan

3.3.5. BUN

Eight studies with 683 patients assessed the therapy of NRT plus valsartan versus valsartan in decreasing BUN in patients with CGN[9-11, 14, 19, 23, 29, 32]. It has considerably high heterogeneity in BUN(P ＜ 0.001, I2= 78%), so a random-effect model was conducted for analysis. Compared with valsartan alone in the treatment of CGN with NRT combined with valsartan, there was no significant difference (MD = -0.02; 95% CI, -0.41, 0.37; P = 0.93) (Figure 8).Sensitivity analysis found that excluding 7 studies, the combined effects were still statistically significant and the direction of the forest plot results did not change.

Figure 8 BUN: NRT combined with valsartan vs valsartan

3.3.6. ALB

A total of 4 studies with 454 patients evaluated ALB [10-13]. It has considerably high heterogeneity in ALB (P = 0.09, I2= 53%),after excluding studies that may cause heterogeneity [10], the heterogeneity reduced (P = 0.48, I2= 0%), so a fixed-effect model was conducted for analysis. The trials reported that NRT plus valsartan was superior to atorvastatin treatment to increase ALB, the difference was statistically significant (MD = 5.05; 95% CI, 4.27,5.84; P ＜ 0.00001) (Figure 9).

Figure 9 ALB: NRT combined with valsartan vs valsartan

3.3.7. Adverse reactions

Nine studies with 840 patients reported the occurrence of adverse reactions [11, 16-18, 21-22, 26-27, 30]. After the heterogeneity test, there is no heterogeneity among the studies (P = 0.99, I2= 0%). The results of Meta-analysis using a fixed-effect model show that the difference between the treatment group and the control group is not statistically significant (OR = 0.93, 95% CI, 0.54, 1.60; P = 0.78)(Figure 10).

Figure 10 Adverse reactions: NRT combined with valsartan vs valsartan

3.4. Publish biased reviews

Using a funnel plot to test the publication bias of the outcome measure of effective treatment rate, it can be seen that the distribution of each study is basically symmetrical, indicating that the possibility of publication bias is small (Figure 11).

Figure 11 Funnel plot of NRT plus valsartan versus valsartan on effective treatment rate in patients with CGN.

3.5. GRADE evidence quality

This study has 7 outcome indicators, which are 4 primary outcome measures: effective treatment rate, BP, 24h urine protein quantification and Scr. 3 secondary outcome measures: BUN,ALB and adverse reactions. The GRADE system recommended classification and reasons for each outcome are shown in Table 2.

4. Discussion

4.1. Clinical efficacy analysis

Traditional Chinese medicine believes that CGN belongs to the categories of "edema" and "turbid urine". The main cause of its symptoms is that the spleen and stomach are damaged when external pathogens invade the body, which causes the weakness of righteous qi and kidney qi. Insufficient, resulting in hematuria, edema and other symptoms. Clinically, it should be based on invigorating the spleen and kidney, promoting blood circulation and removing blood stasis, so as to reduce the relevant symptoms of patients and achieve the effect of treating both the symptoms and the root causes [33]. NRT are currently commonly used prescriptions for the treatment of nephropathy in traditional Chinese medicine.They are mainly composed of Chinese medicines such as Chinese yam, Alisma orientalis, ginseng, black bean, Imperata cylindrica root and motherwort. The effect of spleen and kidney. Modern pharmacology shows [34-35] that the drug can enhance the body's immune function, reduce swelling and diuresis, regulate blood pressure, and improve renal function. In this study, 24 articles werefinally included. The results of Meta-analysis showed that the NRT combined with valsartan group was significantly better than the valsartan group in terms of treatment efficiency, 24h urine protein content, Scr, BP and ALB. There are statistical differences, and there is no statistical difference between the two indicators of BUN and adverse responders. Sensitivity analysis showed that, except for ALB, the results of other outcome indicators did not change.The results of the funnel chart analysis are basically symmetrical,which indicates that the possibility of publishing deviation is small,indicating that the results are reliable. Therefore, based on the results of the study, it can be considered that the combined use of NRT and valsartan has certain advantages in the treatment of CGN and can improve the clinical efficacy. However, due to the small sample size and many confounding factors, it needs to be confirmed by further large-sample, high-quality, multi-center studies. Of the 24 studies,only 9 mentioned the occurrence of adverse reactions. Among them,there were 27 cases of adverse reactions in the treatment group and 29 cases in the control group. Adverse reactions mainly include:nausea, headache, diarrhea, gastrointestinal reactions, cough, mild hyperkalemia, skin itching, ecchymosis, dizziness, vomiting, etc.There was no statistical difference between the two groups, and no other serious adverse reactions. In addition, more than half of the included studies did not report adverse reactions. It is recommended that future studies should pay attention to drug safety and improve the observation and reporting of safety indicators to improve the clinical reference value of the study.

Table 2 GRADE Summary Table of Outcome Indicator Evidence Quality

4.2. Limitations of this study

Statistical heterogeneity still exists among some of the outcome indicators included in the trial. Five problems still remain in these RCTs from the results: (1) Some studies did not report the method of randomization, and there may be selection bias. (2) None of the included studies mentioned allocation concealment and blinding, and there may be biases in implementation and measurement. (3) The number of studies on some outcome indicators is small. Although a statistically significant conclusion has been reached, the statistical power may be insufficient, and there is still a need to increase the sample for further research. (4) Although the language search is not restricted in this research, 15 articles were included in the Chinese literature and 0 articles in English after the search, which may affect the extrapolation of the research results. (5) Most of the studies did not report adverse reactions.

4.3. GRADE evidence quality and recommendation level

In the GRADE system, although the evidence based on RCTs is initially rated as high quality, our confidence in this type of evidence may be reduced for some reasons. According to the GRADE methodological quality evaluation, 7 of the above 8 results belong to medium-level evidence, and one is low-level evidence. The reason is the lack of concealment and blinding, so there are research limitations. Some studies included fewer patients and observations,resulting in inaccurate results [36, 37]. At the same time, the main outcome indicator of this study is the effective treatment rate and 24h urine protein quantification, Scr, BP, BUN, ALB and adverse reactions are all "important" and "critical" outcome indicators, but in view of the low quality of the original study, the authenticity of the conclusions was affected. And the included documents are all domestic, which affects the extrapolation of the conclusion, so the recommendation level is "weak recommendation ".

In this study, a Meta-analysis of 24 included articles showed that compared with valsartan alone, the combination of NRT and valsartan can significantly reduce the patient's proteinuria and blood creatinine levels and improve renal function. However, in view of the small number of included studies and the overall low quality of the literature, more high-quality, multi-center, large-sample, and long-term follow-up clinical trials are needed to support the results,in order to provide more scientific and reliable clinical treatment recommendations for the clinic .

Abbreviations

NRT: Nephritis Rehabilitation Tablet

CGN: Chronic Glomerulonephritis

Scr: Serum creatinine

BP: Blood pressure

SBP: Systolic blood pressure

DBP: Diastolic blood pressure

BUN: Blood urine nitrogen

ALB: Plasma albumin

RCTs: Randomized Controlled Trials

OR: Odds ratio

CI: Confidence interval

MD: Mean difference

T: Treatment group

C: Conventional group

CT: Conventional treatment

NR: Not report

Cap: Capsule

Conflicts of Interest

The authors of this manuscript state that they do not have any conflicts of interest and that there is nothing to disclose.

Authors’ contributions

Qiao Liu, the first author, participated in literature retrieval and screening, sorting out and analyzing related data, and writing the first draft. Peifeng Wei, the corresponding author, participated in the work of paper design and article content review. The second author,participated in the selection of literature and data extraction. The third author, involved in data analysis. The fourth and fifth authors participated in correcting the first draft of the article.

Acknowledgments

This work was financially supported by grants from Innovative team plan of the Second Affiliated Hospital of Shaanxi University of Chinese Medicine (2020XKTD-A04) and Discipline Innovation Team Plan of Shaanxi University of Chinese Medicine (2019-QN02).