Efficacy and Safety of Salvia miltiorrhiza for Treating Chronic Kidney Diseases: A Systematic Review and Meta-Analysis

Objective This meta-analysis evaluated the effects and potential harms of Salvia miltiorrhiza or its extracts Salvianolate and Tanshinone for the treatment of population with a chronic kidney disease (CKD). Methods We searched for the randomized clinical trials (RCTs) through databases including the Cochrane Library, PubMed, Embase, Web of Science, Current Controlled Trials, China National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform (Wanfang Data), China Biology Medicine Disc (SinoMed), and Chinese Clinical Trial Registry (ChiCTR). Meta-analysis was performed with STATA 16 software after data extraction. The risk of bias was assessed with the Cochrane risk-of-bias tool (RoB 2.0), and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework was employed to evaluate the quality of evidence. Result A total of 32 studies were included involving 2264 participants. Compared to the control group, the treatment group significantly decreased serum creatinine (SCr) (SMD −0.60, 95% CI −0.79 to −0.41, P < 0.0001), blood urea nitrogen (BUN) (SMD −0.66, 95% CI −0.81 to −0.50, P < 0.0001), Cystatin C (CysC) (SMD −5.16, 95% CI −14.84 to 4.53, P=0.297), 24 hour urine protein (24 h UPE) (SMD −0.70, 95% CI −1.21 to −0.19, P=0.008), time to initiation of dialysis (Log RR 0.43, 95% CI 0.23 to 0.81, P=0.0089), serum total cholesterol (TC) (SMD −0.53, 95% CI −0.88 to −0.17, P=0.0042, P=0.0035), plasma fibrinogen (FIB) (SMD −0.79, 95% CI −1.12 to −0.46, P < 0.0001), C-reactive protein (CRP) (SMD −0.56, 95% CI −0.93 to −0.19, P=0.0029); increased creatinine clearance (Ccr) (SMD 0.92, 95% CI 0.43 to 1.41, P=0.0002), glomerular filtration rate (GFR) (SMD 0.56, 95% CI 0.30 to 0.83, P < 0.001), effective rate (Log RR 0.30, 95% CI 0.23 to 0.37, P < 0.0001), and hemoglobin (Hb) (SMD 0.42, 95% CI 0.13 to 0.71, P=0.0042). Moreover, the incidences of adverse effects were similar between the two groups. Conclusions Salvia miltiorrhiza or its extracts Salvianolate and Tanshinone, as a complementary therapy to conventional medicine, presents potential impacts to improve kidney functions and delay the progression of CKD without obvious adverse effects. However, the certainty of the evidence and the risk of bias are suboptimal and further clinical studies are still required to determine the underlying effects.


Introduction
A chronic kidney disease (CKD) arises from various heterogeneous diseases. e diagnosis of CKD rests on establishing a chronic reduction in the kidney function and structural damage. e prevalence of CKD for stages 1-5 is 13.4% and 10.6% for stages 3-5 [1]. Contrary to diabetes or other metabolic diseases as prevalent as CKD, renal function impairment is often asymptomatic until very late stages [2]. According to the National Kidney Foundation, 30 million adults in the United States had CKD in 2017, and only 10 percent knew they had it, at a medical cost of $103 billion. In addition, CKD ranks fourteenth in the list of leading causes of death, which accounts for 12.2 deaths per 1,00,000 people, and the death rate of CKD will continue to increase to reach 14 per 1,00,000 people by 2030 [3]. In short, CKD has the characteristics of high incidence, high cost, high mortality, and low recognition rate.
Patients with CKD need efficient treatments to delay disease progression and improve the quality of life and the survival rate. In China, Chinese herbal medicines (CHMs) are widely used for the treatment of CKD. ere are many prescriptions containing varieties or single CHM for CKD. It could also be said that every doctor's prescription may be different and is constantly modified during patients' followup. Hence, comparing the efficacy of diverse prescriptions is inherently heterogeneous and is not conducive to promotion outside China. Furthermore, according to the traditional Chinese medicine (TCM) theory, promoting blood circulation and removing blood stasis should be adopted throughout the treatment of CKD. Salvia miltiorrhiza (Danshen) is one of the most commonly used CHMs. Studies have demonstrated that Salvia miltiorrhiza is the top single CHM prescribed for CKD in China [4]. Medicinal parts of Salvia miltiorrhiza (Danshen) is the dried root and rhizomes of Salvia miltiorrhiza Bge. Salvia miltiorrhiza has specific preparations for clinical applications, such as Salvia miltiorrhiza tablet or injection and its extracts Salvianolate injection as well as Tanshinone injection, which all have strict quality control standards and the procedures are fully reproducible. Salvia miltiorrhiza and its extracts Salvianolate and Tanshinone are extensively used for CKD.
Clinically, a number of studies have displayed that Salvia miltiorrhiza can improve kidney function in CKD patients by increasing the glomerular filtration rate (GFR) and creatinine clearance (Ccr) and reducing serum creatinine (SCr) and proteinuria, but this conclusion is yet to be verified [5].
Research studies also reveal that Salvia miltiorrhiza could alleviate kidney injury via inhibiting oxidative stress and apoptosis [6] and exerting prominent renal protective effects in iron-overloaded mice by decreasing of iron deposition and suppression of lipid peroxidation and apoptosis in the kidney [7]. Tanshinone IIA significantly attenuates kidney fibrosis by inhibiting the recruitment of fibrocytes into the kidney [8] and decreases renal damage in diabetic rats via inhibiting oxidative stress and inflammation [9]. Salvianolate might alleviate the renal damage in chronic renal failure rats through antioxidant stress [10], accordingly attenuating glomerular injury, including albuminuria secretion, mesangial matrix expansion, foot process effacement in the kidneys of db/db mice, and ameliorated oxidative podocyte injury [11].
ere is one previous meta-analysis that evaluated the efficacy and safety of Tanshinone for CKD [12], which includes 21 studies published before June 1, 2019. In our meta-analysis, we include Salvia miltiorrhiza and its extracts Salvianolate besides Tanshinone because they have a strong connection with each other and they are all widely used in China for CKD. All of the studies we included were published before November 9th, 2021. e subjects in the previous meta-analysis were diagnosed with diabetic nephropathy (3 studies), hypertensive renal damage (4 studies), renal vascular hypertension (1 study), and cardiorenal syndrome (1 study) rather than CKD (12 studies).
Whereas, we focused on the subjects of patients diagnosed with CKD. e inclusion of subjects was more rigorous, whereas we may miss some patients with CKD who were diagnosed with hypertensive nephropathy or other diagnoses. e current meta-analysis was performed to comprehensively evaluate the efficacy and safety of Salvia miltiorrhiza and its extracts Salvianolate and Tanshinone for the treatment of patients with CKD, with a view to provide substantial evidence for supporting its clinical application in CKD patients.

Protocol and Registration.
is meta-analysis had been registered in PROSPEPO with registration number CRD42021291786.

Types of Participants
(1) Inclusion Criteria. Adults and children with CKD at all stages.
(2) Exclusion Criteria. Studies stating that participants had renal damage, but without baseline GFR, Ccr, or SCr; participants with diabetic nephropathy. ese issues had been investigated in a previous study [13]; studies involving Salvia miltiorrhiza as one of several active components in a compound recipe were not included.

Types of Interventions. Treatment group received
Salvia miltiorrhiza or its extracts Salvianolate and Tanshinone. e control group received placebo, no treatment, or conventional treatment.

Search Methods.
is meta-analysis was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020 [14,15] Table S2).

Study Selection.
e search strategy described was used to obtain titles and abstracts of studies that may be relevant to this review. Titles, abstracts, and full texts were screened independently by two authors who determined which met the inclusion criteria and excluded studies that were not appropriate.

Data Collection Process.
Data extraction was carried out independently by the same two authors using a pre-tested data extraction form. If more than one publication of one study existed, the publication with the most complete data was used. Any discrepancy between published versions was to be highlighted. Disagreements between authors were resolved by consensus and with a third author.

Quality Assessment and Statistical Methods.
e publications included in this meta-analysis were subject to quality assessment according to the Cochrane criteria [15]. e risk of bias was assessed using the Cochrane risk-of-bias tool (RoB. 2.0) [16]. In addition, the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework was employed to evaluate the quality of evidence contributing to each estimate [15]. e STATA 16 software was used for data analysis. For dichotomous outcomes, results were expressed as Log risk ratio (Log RR) with 95% confidence intervals (CI). For continuous outcomes, the standard mean difference (SMD) was presented with 95% confidence intervals (CI).
Heterogeneity was analyzed using a Chi 2 test on N − 1 degrees of freedom, with an alpha of 0.1 used for statistical significance and with the I 2 -test [15]. I 2 > 50% corresponds   to high levels of heterogeneity, respectively. A subgroup or sensitivity analysis was conducted to explore the underlying causes of heterogeneity in treatment outcomes.

Evidence-Based Complementary and Alternative Medicine
To assess small-study effects, we generated Egger's test or funnel plots [15] including at least 10 trials of varying size. If asymmetry was detected in the funnel plot, a contour- enhanced funnel plot was generated to assess whether the asymmetry was likely due to publication bias or other factors of the trials.

Additional Analyses.
We conducted subgroup analyses to explore the impact of Salvia miltiorrhiza and its extracts Salvianolate and Tanshinone preparations.

Study Selection.
Our initial search found 332 records. After excluding 47 duplicate reports and 232 irrelevant records based on identification of titles and abstracts, we reviewed 53 full-text studies for inclusion and then 21 studies were further excluded. Finally, a total of 32 studies were included in the meta-analysis (Figure 1).

Study
Characteristics. 32 included studies involved 2264 participants and were conducted in hospitals of China and published in Chinese. All studies were parallel arm. Participants' age ranged from 18-96 years. Authors, year of publication, treatment plan, sample size, duration of treatment, and index of evaluation of each study are presented in Table 1.

Risk of Bias.
e 32 studies included were all RCTs, yet only 9 had detailed descriptions of the methods. In one study, computer-generated random numbers were used in the sequence generation process, and in 6 studies, random number tables were adopted. Nevertheless, in another 2 studies, patient record numbers were used. Only one study specified the method of blinding, and that was single-blind. All studies were assessed according to the RoB. 2.0 tool, of which 8 (25%) were assessed as "low risk of bias," 22 (68.75%) as "some concerns," and 2 (6.25%) as "high risk" (Figure 2).

Kidney Function
(1) SCr. A total of 29 studies compared Scr levels between the treatment and control group. We classified these studies into following subgroups based on the different preparations: Salvia miltiorrhiza, Salvianolate, and Tanshinone. As indicated in Figure 3, the random-effect model was used due to the high heterogeneity. Scr levels in the treatment group were significantly reduced compared with the control group (SMD −0.60, 95% CI −0.79 to −0.41, P < 0.0001, e Salvianolate group was the main source of heterogeneity. en, we divided the studies of the Salvianolate group into various subgroups based on bias, when we removed the RCT of Wang et al. [21] which had the highest Scr and the Liu [24] with the smallest sample size, both low risk group and some concerns group had low heterogeneity (I 2 � 0 and I 2 � 6.03%).
Egger's test exhibited that there was no publication bias P < 0.001.
(2) Ccr. A total of 13 studies compared Ccr levels between the treatment group and control group. We classified studies   Fixed-effects inverse-variance model      (3) GFR. A total of 3 studies compared GFR levels between the treatment group and control group. As shown in Figure 5, the heterogeneity was low (I 2 � 0%) and the fixedeffect model was employed to analyze the data. GFR levels in the treatment group were significantly increased compared with the control group (SMD 0.56, 95% CI 0.30 to 0.83, P < 0.001, I 2 � 0%).
(5) CysC. A total of 4 studies recorded CysC levels between the treatment group and control group. As presented in Figure 7, CysC levels in the treatment group were decreased but without significance compared with the control group (SMD −5.16, 95% CI −14.84 to 4.53, P � 0.297, I 2 � 99.93%).
(6) Effective Rate. A total of 15 studies compared the effective rate between the two groups. As listed in Figure 8(a), the effective rate in the treatment group was significantly higher compared with the control group (Log RR 0.30, 95% CI 0.23 to 0.37, P < 0.0001, I 2 � 0%), and the heterogeneity was low. Subgroup analysis indicated that the effective rates were remarkably higher in Salvia miltiorrhiza and Tanshinone groups compared to the control group (Log RR 0.40, 95% CI 0.30 to 0.50, P < 0.0001, I 2 � 0%; Log RR 0.28 95% CI 0.11 to 0.44, P < 0.001, I 2 � 0%), the effective rate was higher in the Salvianolate group compared with the control group (Log RR 0.30, 95% CI -0.07 to 0.68, P � 0.113, I 2 � 76.77%). Liu et al. [40] used alprostadil injection combined Salvianolate injection in the treatment group, which might bias the heterogeneity. Furthermore, removal of this study resulted in a considerable reduced I 2 (I 2 � 0%).
Egger's test revealed that there was publication bias (P < 0.05). As shown in Figure 8(b), 7 studies were imputed and 2 studies were in the area of 5% <P < 10%.

24 h UPE.
A total of 10 studies compared 24 h UPE levels between the treatment group and control group. As suggested in Figure 9, 24 h UPE levels in the treatment group were observably reduced compared with the control group (SMD −0.70, 95% CI −1.21 to −0.19, P � 0.008, I 2 � 90.36%). We classified these studies into three subgroups based on the degrees of 24 h UPE: ≤1.0 g, 1.0-3.5 g, ≥3.5 g. Subgroup analysis indicated that the effect of Salvia miltiorrhiza was inversely proportional to the degree of proteinuria. Egger's test hinted that there was no publication bias (P � 0.5167).

Time to Initiation of Dialysis.
ere was only one study [23] that reported the time to initiation of dialysis. As demonstrated in Figure 10, in comparison with the control group, there were less CKD patients into initiation of dialysis in the treatment group (Log RR 0.43, 95% CI 0.23 to 0.81, P � 0.0089).

Adverse Effects.
In all, 13 studies reported adverse effects. As presented in Figure 11(a), adverse effects in the treatment group did not differ significantly from that of the control group (Log RR −0.52, 95% CI: −1.16 to 0.12, P � 0.1112, I 2 � 0%). Random-effects Hedges model

Evidence-Based Complementary and Alternative Medicine
Egger's test indicated that there was publication bias P � 0.0349. As shown in Figure 11(b), 7 studies were imputed, 2 studies were in the area of 1% < P < 5%, and one study was in the area of 5% < P < 10%.

Nutritional Status
(1) ALB. A total of 4 studies documented ALB levels between the treatment group and control group. As displayed in Figure 12, there was no significance between the Salvia miltiorrhiza group and control group (SMD 0.23, 95% CI −0.28 to 0.75, P � 0.3775, I 2 � 76.82%).
(2) TC. A total of 2 studies recorded TC levels between the treatment group and control group. As expressed in Figure 13, TC levels in the treatment group were distinctly reduced compared with the control group (SMD −0.53, 95% CI−0.88 to −0.17, P � 0.0035, I 2 � 0%).

Anemia Measured: Hb.
A total of 6 studies reported HB levels. As manifested in Figure 14, Hb levels in the treatment group were significantly enhanced compared with the control group (SMD 0.42, 95% CI 0.13 to 0.71, P � 0.0042, I 2 � 50.39%).

Hemorheology Index: FIB.
A total of 2 studies had data of FIB levels. As presented in Figure 15, FIB levels in the treatment group were significantly reduced compared with the control group (SMD −0.79, 95% CI −1.12 to −0.46, P < 0.0001, I 2 � 0%).

Inflammatory Factor: CRP.
A total of 3 studies compared CRP levels between the treatment group and control group. As displayed in Figure 16, CRP levels in the treatment group were notably reduced compared with the control group (SMD −0.56, 95% CI −0.93 to −0.19, P � 0.0029, I 2 � 56.31%).

Certainty of Evidence.
All outcome indicators were evaluated by GRADEpro GDT. e quality of evidence was downgraded for the risk of bias or publication bias. After comprehensive analysis, the summary table was formed, and it was found that 2 outcome indicators (14.29%) were of high quality, 12 outcome indicators (71.43%) were of moderate quality, and 2 outcome indicators (14.29%) were of low quality (Supplementary Table S3).

Publication Bias.
Egger's test declared that no publication bias in the indicators of SCr, CCr, BUN, and 24 h UPE were observed. Whereas, there was publication bias in the indicator of the effective rate; the contour-enhanced funnel plot suggested that 7 studies were imputed and 2 studies were in the area of 5% < P < 10%. In the indicator of adverse effects, Egger's test hinted that there existed a publication bias, 7 studies were imputed, 2 studies were in  the area of 1% < P < 5%, and another one study was in the area of 5% < P < 10%.

Discussion
In this meta-analysis, we intended to explore potential effects of Salvia miltiorrhiza for people with CKD on disease progression and complications. 32 studies that involved 2264 participants with CKD were included. In the aspects of kidney functions (SCr, Ccr, GFR, BUN, CySC, and effective rate), the comparison results revealed that the treatment group significantly reduced SCr, BUN and CysC, increased Ccr, GFR, and effective rate, indicating that this complementary therapy may have good effects for kidney functions. Proteinuria is a common clinical feature in patients with CKD, which is also an important factor in the CKD progression. e comparison    Heterogeneity: T 2 = 0.00%, I 2 = .%, H 2 = .
Test of θ = 0: z = -2.62, P = 0.01   CKD patients into initiation of dialysis in the treatment group. Unfortunately, the sample size of this study was too small (n � 30). 13 studies reported adverse effects. e incidence of adverse effects was similar between the two groups. Meanwhile, some indicators of complications were compared. e treatment group alleviated CKD-associated complications, and the complementary therapy may have effects on reducing TC, FIB, and CRP levels and increasing Hb levels.

Evidence-Based Complementary and Alternative Medicine
According to the TCM theory, Salvia miltiorrhiza (Danshen) is one of the most commonly used CHMs with the effect of promoting blood circulation to remove blood stasis, which should be used throughout the treatment of CKD. Studies have indicated that Salvia miltiorrhiza is the top single CHM prescribed for the treatment of CKD in China. A number of studies have testified that Salvia miltiorrhiza can improve kidney functions in CKD patients by increasing GFR and Ccr and reducing SCr and proteinuria [4]. Research studies also revealed that Salvia miltiorrhiza alleviated kidney injury via inhibiting oxidative stress and apoptosis [5]. Tanshinone IIA obviously attenuated kidney fibrosis by inhibiting the recruitment of fibrocytes into the kidney [7]. Besides, Salvianolate alleviated the renal damage and attenuated glomerular injury through antioxidant stress [9].
ere are certain limitations of the evidence that should be considered. According to the RoB 2.0 tool, 75% of the studies were assessed as "some concerns" or "high risk." Although all of the included studies claimed to have a randomized controlled design, but only 9 had detailed descriptions of the methods. Methodological deficiencies are related to the lack of a clear description of randomization, allocation concealment, and binding. In 7 studies, computer-generated random numbers or random number tables were used. Whereas, patient record numbers were used in 2 studies. Only one study specified the method of blinding, whereas, that was single-blind. Egger's test implied that there was publication bias in the indicator of effective rate and adverse effects.
In addition, the heterogeneity was high in the results of some indicators. As for SCr, Salvianolate group was the main cause of heterogeneity. When we removed the RCT of Wang et al. [21] with the highest baseline Scr and the Liu et al. [24] with the smallest sample size, the heterogeneity significantly declined. In the indicator of the effective rate, Liu et al. [40] used alprostadil injection combined with Salvianolate injection for the treatment group, which could bias the heterogeneity. ese abovementioned statements indicate that we should expand the sample size, to improve the quality of RCT. Meanwhile, limited by language barriers, only Chinese and English databases were searched, and all the included studies were conducted in China, which might affect the final results to a certain degree. Hence, there exist doubts about the applicability of evidence in other countries.

Conclusions
Current evidences indicate that Salvia miltiorrhiza may have certain benefits for CKD patients as a complementary therapy, which could improve kidney functions, reduce proteinuria, delay the progression of CKD, and improve   Test of θ i = θ j : Q (2) = 4.33, P = 0.11 Test of θ = 0: z = -2.98, P = 0.00 several complications resulting from CKD. However, the certainty of the evidence and the risk of bias are suboptimal and further clinical studies are still needed to determine the potential effects of Salvia miltiorrhiza for patients with CKD.

Data Availability
e data used to support the findings of this study are available on request from the first and corresponding authors.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Niansong Wang, Youhua Xu, and Dingkun Gui conceived and designed this study. Wei Zhang, Jun Li, Pan Yang, Gaoqiang Wang, and Yuanfang Zhong performed the data extraction, analysis, and interpretation and wrote the initial draft. Hanyin Liu and Yanli Yue assisted with data interpretation.