Efficacy of statins on renal function in patients with chronic kidney disease: a systematic review and meta-analysis

Abstract Background Studies have shown that the use of statins could significantly improve lipid profiles; however, it remains controversial whether the use of statins could improve renal function in patients with chronic kidney disease (CKD). Therefore, we conducted a meta-analysis of randomized controlled trials (RCTs) to evaluate the effects of statins on renal function in patients with CKD. Methods We systematically searched PubMed, EMBASE, and the Cochrane Library databases for eligible RCTs from inception to October 2020. Pooled effect estimates were assigned as weighted mean differences (WMDs) with 95% confidence intervals (CIs) using the random-effects model. Results We selected 33 RCTs that recruited 37,391 patients with CKD patients. The summary results suggested that statin use significantly reduced urinary albumin (WMD: −2.04; 95%CI: −3.53 to −0.56; p = .007) and protein (WMD: −0.58; 95%CI: −0.95 to −0.21; p = .002) excretions and increased creatinine clearance (WMD: 0.86; 95%CI: 0.32–1.41; p = .002). However, there were no significant differences between statin and control groups in terms of changes in estimated glomerular filtration rate (WMD: 0.38; 95%CI: −0.04 to 0.79; p = .075), and serum creatinine levels (WMD: −0.07; 95%CI: −0.25, 0.12; p = .475). Conclusions We found that statin use in patients with CKD may slow CKD progression by lowering urinary albumin and protein excretions or increasing creatinine clearance. Further large-scale RCTs should be conducted to evaluate the long-term effects of statins on renal outcomes. Abbreviations: CKD: chronic kidney disease; RCT: randomized controlled trials; WMD: weighted mean differences; CI: confidence intervals; ACEI: angiotensin-converting enzyme inhibitors; eGFR: estimated glomerular filtration rate


Introduction
Chronic kidney disease (CKD) is a common disorder that can cause cardiovascular disease, kidney failure, and other complications. CKD, with an increasing prevalence, affects more than 10% of the global population [1]. In the United States of America (USA), an estimated 9.6% of non-institutionalized adults suffer from CKD [2,3]. Studies conducted in Europe, Australia, and Asia confirm the high prevalence of CKD [4][5][6]. However, the prevalence of CKD in many developing countries remains unknown [7]. CKD has many potential causes that vary in frequency among populations [8]. Renovascular disease is one of the risk factors for developing CKD and worsening renal outcomes [9]. Some proposed mechanisms for progressive CKD in patients with renovascular disease include endothelial dysfunction, oxidative stress, and systemic inflammation of the glomerular capillary wall [10]. Renal replacement therapy is currently the most effective CKD treatment modality; however, the administration of renal replacement therapy in low-and middle-income countries is limited because it is not readily available in these countries. Further, most patients die of kidney failure without receiving dialysis or transplantation [11]. Thus, it is necessary to find alternative strategies to improve the prognosis of CKD.
Lipid-lowering therapies may improve renal function and lower albuminuria as reported in some previous meta-analyses [12,13]. They found that lipid reduction could improve renal function and decrease proteinuria in patients with CKD [13]. Although the abovementioned meta-analyses reported the beneficial effects of statins on pathologic albuminuria, larger studies are required to assess the validity of these findings and determine if statins can also reduce cardiovascular or end-stage renal disease occurrence [12]. Another study found that a combined approach using angiotensinconverting enzyme inhibitors (ACEI) and statins could represent a therapeutic option for patients with advanced renal disease in whom ACEI alone failed to substantially reduce proteinuria and renal injury [14]. Some scholars believe that this may be due to the antiinflammatory effect of statins and improved endothelial function [14]. Previous meta-analyses that evaluated the effect of statins on renal outcomes did not focus on patients with CKD [15]. Therefore, we conducted this meta-analysis, based on randomized controlled trials (RCTs), to evaluate the effects of statins on renal function in patients with CKD.

Search strategy and eligibility criteria
According to the transparent reporting of systematic reviews and meta-analyses 2009 (PRISMA checklist) [16], two reviewers independently identified relevant studies in the PubMed, EMBASE, and Cochrane library databases from inception to October 2020. This review included only studies published in English, and the following terms were applied in our search: CKD, chronic renal disease, chronic nephropathy, statin, atorvastatin, simvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, cerivastatin, mevastatin, pitavastatin, dyslipidemia, hyperlipidemia, hypercholesterolemia, hyperlipoproteinemia, hypertriglyceridemia, human, and RCTs. The details of the search strategies for each database are shown in Supplemental 1. If a dataset was published multiple times, the most relevant publication was included, and the others supplemented it.
Studies were included if they met the following criteria: (1)  Kinetic method), and urinary protein excretion. Reviews, case reports, letters, mechanism studies, and non-human studies were excluded. After an initial screening of the study titles and abstracts, the full texts of potentially eligible studies were read to assess whether the study could be included in the meta-analysis. This process was performed by two reviewers (SL and YG), and inconsistent results between reviewers were settled by group discussion until a consensus was reached.
Data extraction and quality assessment Data were independently extracted from the selected studies by two reviewers (SL and YG) using standardized criteria. The following items were extracted: first author's surname, year of publication, country, number of participants, mean age, and baseline eGFR in the intervention and control groups, intervention, control, follow-up duration, and reported outcomes. Quality assessment was performed simultaneously by two reviewers (SL and YG) using the Jadad scale (with scores ranging from 0 to 5), and based on randomization, blinding, allocation concealment, withdrawals and dropouts, and the use of intention-to-treat analysis [17]. Any conflicts between reviewers for data abstraction and quality assessment were settled by a third reviewer (LZ), who reviewed the full-text of retrieved studies.

Statistical analysis
The treatment efficacy of statins on renal function improvement in patients with CKD was assigned as continuous data, and the pooled weighted mean differences (WMDs) with 95% confidence intervals (CIs) were calculated using the random-effects model, which considering the varies underlying included studies [18,19]. The heterogeneity among included studies was assessed using the I 2 statistic and Q statistic [20,21]. Sensitivity analysis was conducted by leave-one-out to assess the robustness of a pooled conclusion [22]. Subgroup analysis was conducted according to year of publication (before 2010, 2010, or after), country (Asia, other), sample size (!100, <100), mean age (!65.0, <65.0 years), statin type (atorvastatin, fluvastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin), follow-up duration (!12.0, <12.0 months), and study quality (high [Jadad score 4 or 5], low [Jadad score 0-3]); the difference between subgroups was assessed using the interaction p test [23]. The funnel plot, Egger, and Begg test results were used to assess potential publication bias [24,25]. The inspection level was 2-sided, and statistical significance was set at p<.05. All statistical analyses were performed using STATA software version 10.0 (StataCorp, College Station, TX).

Literature search
A total of 3741 published studies were initially identified, of which 2246 potentially relevant studies were retained after duplicates were removed. Subsequently, 2169 were excluded owing to irrelevant topics. The remaining 77 studies were retrieved for further full-text evaluations, and 33 RCTs met the inclusion criteria . By reviewing the reference lists of these studies, we found three potentially eligible studies, and these studies were contained in initial electronic searches. Details of the study selection process are shown in Figure 1.

Study characteristics
We included 33 RCTs involving 37,391 patients with CKD; the baseline characteristics of included studies are summarized in Table 1. The duration of follow-up ranged from 10 d to 72.0 months and 28 to 16,245 patients were included in each individual trial. The intervention strategies included the administration of atorvastatin, cerivastatin, fluvastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin. Study quality was assessed using the Jadad scale; 3, 10, 13, and 7 trials scored 5, 4, 3, and 2, respectively.

Estimated glomerular filtration rate
Twenty-one studies reported the effect of statins on eGFR, and the pooled result indicated that the use of statins was not associated with a change in eGFR as compared with the control (WMD: 0.38; 95%CI: À0.04 to 0.79; p ¼ .075; Figure 2). Moreover, we observed significant heterogeneity across included trials (I 2 ¼98.3%; p< .001). Sensitivity analysis suggested that the pooled conclusion was unstable because of the marginal 95%CI (Supplemental 2). Subgroup analysis revealed that statin use was associated with high eGFR for pooled studies published before 2010, mean patient age !65.0 years, atorvastatin use, pravastatin use, and low-quality studies. However, we noted that the use of fluvastatin was associated with a lower eGFR than that in the control group (Table 2). There was no significant publication bias for eGFR (p value for Egger: .277; p value for Begg: .309; Supplemental 3).

Urinary albumin excretion
Ten studies reported the effect of statins on urinary albumin excretion and the pooled results suggested that the use of statins was associated with lower urinary albumin excretion than that in the control group (WMD: À2.04; 95%CI: À3.53 to À0.56; p ¼ .007; Figure 3). Moreover, there was significant heterogeneity in urinary albumin excretion across the included trials (I 2 ¼ 99.2%; p < .001). The conclusion was not robust when individual studies were excluded one by one (Supplemental 2). Although significant differences between statin and control groups were observed in most subgroups, we noted that statins had no significant effect on urinary albumin excretion for pooled studies conducted in Asia or other countries, sample size !100, mean age of patients <65.0 years, and fluvastatin or simvastatin use. Conversely, we noted that the use of statins was associated with high urinary albumin excretion for pravastatin use, follow-up duration !12.0 months, and high-quality studies ( Table 2). No significant publication bias was observed for urinary albumin excretion (p value for Egger: .695; p value for Begg: .858; Supplemental 3).

Creatinine clearance
Ten studies reported the effects of statin use on creatinine clearance. We noted that statin use was associated with higher creatinine clearance compared with that in  Figure 4), and significant heterogeneity was seen among the included trials (I 2 ¼92.8%; p < .001). Sensitivity analysis indicated that the pooled conclusion was not stable after the sequential exclusion of individual studies (Supplemental 2). Subgroup analysis revealed that the use of statins was associated with higher creatinine levels for pooled studies conducted in other countries, mean age <65.0 years, the use of atorvastatin, and follow-up duration !12.0 months ( Table  2). There was no significant publication bias for creatinine clearance (p value for Egger: .269; p value for Begg: .858; Supplemental 3).

Serum creatinine
Seven studies reported the effect of statin use on serum creatinine and the pooled result suggested no significant difference in serum creatinine levels between statin and control groups (WMD: À0.07; 95%CI: À0.25 to 0.12; p ¼ .475; Figure 5). Moreover, there was significant heterogeneity among the included studies (I 2 ¼94.1%; p < .001). The pooled conclusion was robust and not altered by the sequential exclusion of individual studies (Supplemental 2). Subgroup analysis revealed that the use of pravastatin was associated with lower serum creatinine levels ( Table 2). No significant publication bias for serum creatinine was observed (p value for Egger: .876; p value for Begg: .548; Supplemental 3).

Urinary protein excretion
Ten studies reported the effect of statins on urinary protein excretion. Statin use was associated with lower urinary protein excretion (WMD: À0.58; 95%CI: À0.95 to À0.21; p ¼ .002; Figure 6); moreover, we observed a significant heterogeneity for urinary protein excretion (I 2 ¼97.8%; p <.001). Sensitivity analysis indicated that the use of statins was not associated with urinary protein excretion after excluding the study conducted by Gheith et al. [31], which specifically included patients with persistent idiopathic nephrotic syndrome, and most of the patients had focal segmental glomerulosclerosis (Supplemental 2). Subgroup analysis indicated that statin use was associated with lower urinary protein excretion for pooled studies before 2010, studies conducted in Asia or other countries, sample size <100, mean age <65.0 years, the use of pitavastatin, followup duration !12.0 months, and low-quality studies ( Table 2). There was no significant publication bias for urinary protein excretion (p value for Egger: .094; p value for Begg: .107; Supplemental 3).

Discussion
This study assessed the effects of statins on renal function in patients with CKD using a meta-analytic approach. A total of 33 RCTs and 37,391 patients with CKD were identified in our study, and patient characteristics were varied. The findings of this study indicated that patients with CKD treated with statins could experience an improvement in urinary albumin excretion, creatinine clearance, and urinary protein excretion. However, statin therapy had no significant effect on eGFR and serum creatinine levels. Sensitivity analysis revealed that statin therapy might play a protective role on eGFR as compared with the control. Results of subgroup analyses suggested that the treatment efficacy of statins and control could be affected by the year of publication, country, sample size, mean age, statin type, follow-up duration, and study quality. In a review of previous meta-analyses, Palmer et al. found that statins produce significant benefits on mortality and cardiovascular events in patients with earlystage CKD, but had no significant effects in patients on dialysis or after kidney transplantation. Moreover, they reported no significant difference between the use of statins and placebo or no treatment on eGFR [59]. However, this study reported the eGFR based on data at the end of the study, and the baseline eGFR between statin and placebo groups was not consistent. Furthermore, Shen et al. conducted a meta-analysis of 14 studies and found that the mean changes in albuminuria and urinary excretion rates in the statin group were greater than those in the placebo group, whereas no significant differences were seen between statins and placebo in terms of changes in eGFR, serum       creatinine levels, or blood urea nitrogen levels [60]. However, this study included patients with diabetic nephropathy. In addition, the results of this study are consistent with those of an important meta-analysis in that participants with CKD did not require dialysis, and the effect of statins on eGFR changes was clear compared to that of the control group [61], which suggested that statin therapy may slow CKD progression by ameliorating the eGFR in patients with CKD. However, the evidence of progression to end-stage kidney disease relies on data from the SHARP Study 2010 alone, and the treatment effects of statins on this outcome remain uncertain despite the report of more than 2000 events [62]. Su   high-intensity statins [65]. However, these studies did not provide comprehensive results regarding renal function, and several published articles were not included in the meta-analysis. Therefore, the current meta-analysis was conducted to clarify the treatment efficacy of statins on renal function in patients with CKD. The findings of this analysis were not unexpected as statins have been shown to demonstrate pleiotropic effects [15]. Statins, independent of their cholesterollowering effect, could ameliorate endothelial function and reduce inflammatory and fibrogenic processes in the renal interstitium [14], thereby improving renal function. However, we could not find any improvement in eGFR and serum creatinine levels. Trials investigating the effect of statins on kidney function and protein excretion yielded controversial results, with some confirming the renoprotective effect and proteinuria reduction [12] and others showing no effect [66]. In this study, we did not find a significant effect of statins on serum creatinine, although statin use was associated with a high level of urinary albumin excretion compared with that in the control group. However, this result was mainly based on a study conducted by Asselbergs et al. [36], which needed further large-scale RCT verification.
Subgroup analyses found that the beneficial effect of statins was mainly detected for studies published before 2010; studies conducted in other countries; sample size < 100; mean age of patients !65.0 years; the use of atorvastatin, pitavastatin, or pravastatin; or studies with low quality. The statistical power, type I error in individual trials, and quality of the included studies could explain the above results. Moreover, the baseline eGFR in elderly patients was higher than that in younger patients from several included trials, and the beneficial effects of statins might have been focused on patients with mild CKD [26,43,50]. Additionally, these results could guide further direction for the accurate evaluation of the effects of statins on renal function in patients with CKD.
This study had several limitations. First, in this review, studies that were unpublished or published in a language other than English were not included; this might have led to publication bias. Second, the heterogeneity across included studies was high and not fully interpreted through sensitivity and subgroup analyses. Third, analysis based on data from the study level and individual patient data were not available, and the detailed analyses stratified by patient characteristics were restricted. Finally, this study was not registered, and the transparency of this study was restricted.

Conclusion
We found that patients with CKD treated with statins could experience renal function improvement by lowering the urinary albumin and protein excretions or by increasing creatinine clearance, especially with the use of atorvastatin, pitavastatin, or pravastatin. Further large-scale RCTs should be conducted to assess the long-term effects of statins on renal outcomes in patients with CKD.