Efficacy and safety of statin therapy in kidney transplant recipients: a systematic review and meta-analysis

Abstract

Background

Dyslipidemia represents an important risk factor for cardiovascular diseases, although its optimal management after kidney transplantation remains unclear. The present meta-analysis aimed to shed light on the efficacy and safety of statins among kidney transplant recipients, evaluating their potential effects on the risk of cardiovascular events, mortality and graft survival.

Methods

Medline, Scopus, Web of Science, CENTRAL, Clinicaltrials.gov and Google Scholar were systematically searched from their inception through April 20, 2024. Both randomized controlled trials and observational studies evaluating the effects of statin administration after kidney transplantation were held eligible. Random-effects models were fitted using the maximum likelihood method, while the certainty of evidence was appraised following the GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) approach.

Results

Overall, 27 studies (10 randomized controlled trials and 17 observational studies) were included. Statin use compared to no use was associated with a lower risk of major adverse cardiovascular events [Relative risk (RR): 0.87, 95% confidence interval (CI): 0.67–0.96, moderate certainty] and overall mortality (RR: 0.84, 95% CI: 0.74–0.94, low certainty). The risk of graft loss did not differ between the compared groups (RR: 0.72, 95% CI: 0.48–1.08, very low certainty). Regarding safety endpoints, statin use was associated with a lower risk of hepatotoxicity (RR: 0.81, 95% CI: 0.70–0.93, moderate certainty), but with a greater risk of rhabdomyolysis (RR: 1.37, 95% CI: 1.10–1.70, low certainty) and cataract (RR: 1.22, 95% CI: 1.14–1.31, moderate certainty). No statistically significant differences between the compared groups with and without statin use were observed concerning the risk of creatine kinase elevation, post-transplant diabetes mellitus, hip fracture, venous thromboembolism, or cancer.

Conclusions

Among kidney transplant recipients, statin use is associated with a lower risk of cardiovascular events and better patient survival, presenting an acceptable safety profile. Further large-scale studies are needed to determine the optimal statin dosing strategy and lipid-lowering goals, depending on comorbidities and immunosuppression regimens.

Registration

10.17504/protocols.io.5qpvok3yzl4o/v1.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12944-024-02276-w.

Introduction

Cardiovascular disease constitutes a major source of morbidity after kidney transplantation, representing the leading cause of mortality with a functioning graft [1, 2]. Cardiovascular care accounts for one-third of hospitalizations among kidney transplant recipients and is associated with high healthcare costs [3]. The high incidence of cardiovascular events is based on the interplay of traditional and non-traditional risk factors [4]. Specifically, kidney failure patients who develop de novo disease or who have comorbid hypertension, diabetes mellitus or dyslipidemia can have increased cardiovascular risk. In addition, several factors unique to transplant recipients may also contribute to cardiovascular morbidity, such as oxidative stress, anemia, hyperhomocysteinemia and immunosuppression effects. The spectrum of cardiovascular disease, including ischemic heart disease, valvular disease, heart failure and pulmonary hypertension, is wide [5]. However, kidney transplant recipients are typically excluded from major cardiovascular outcome randomized controlled trials and are thus less likely to receive evidence-based goal-directed interventions [6].

Dyslipidemia is prevalent after kidney transplantation, as it is potentiated by the combined effects of various contributing factors, such as post-transplant diabetes mellitus, obesity, impaired renal function, proteinuria and glucocorticoid administration [7]. In addition, cyclosporine may exacerbate hyperlipidemia by reducing hepatocellular levels of low-density lipoprotein receptor [8], while mammalian target of rapamycin (mTOR) inhibitors have been linked to enhanced lipogenesis [9] and impaired catabolism of apolipoprotein B100-containing lipoproteins [10]. Currently, no established lipid targets exist for the kidney transplant population, but it is estimated that a significant proportion of kidney transplant recipients may be undertreated, not achieving the proposed targets for the general population [11].

Hydroxymethylglutaryl-CoA reductase inhibitors or statins are the most commonly prescribed hypolipidemic agents, as they potently reduce low-density lipoprotein levels and exert pleiotropic cardioprotective effects [12]. Although the 2013 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines provide a weak recommendation for the administration of statins in all adult kidney transplant recipients [13], concerns about toxicity and interactions with immunosuppressive agents may impede their widespread use. Previous meta-analyses [14–16] have suggested that statin use may be linked to cardiovascular risk reduction and survival improvement, although evidence regarding statin safety remains scattered. The present meta-analysis aimed to systematically accumulate the existing evidence in the field and shed more light on the association of statin use with the risk of cardiovascular events, mortality and adverse effects among kidney transplant recipients. An updated systematic literature review was conducted including both observational studies and randomized controlled trials in order to provide a critical evaluation of evidence regarding statin efficacy and safety following kidney transplantation.

Materials and methods

Study design

The meta-analysis protocol was prospectively registered and made publicly available (10.17504/protocols.io.5qpvok3yzl4o/v1). The study was reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines [17]. No ethical approval was needed as the meta-analysis was exclusively based on previously published data.

Eligibility criteria

The study population consisted of adult kidney transplant recipients. The intervention of interest was statin administration after kidney transplantation, given for primary or secondary cardiovascular prevention. The intervention was compared to placebo or standard care. The primary outcome of interest was the occurrence of major adverse cardiovascular events (MACE). The endpoint of MACE was a composite one and could include a combination of cardiovascular mortality, acute coronary syndrome, cerebrovascular accident, peripheral artery disease requiring revascularization and congestive heart failure. Secondary efficacy outcomes included patient overall survival and kidney allograft survival. The safety endpoints included the following: hepatotoxicity, rhabdomyolysis, creatine kinase elevation, post-transplant diabetes mellitus, cataract, venous thromboembolic events, hip fracture and cancer. Randomized controlled trials and observational (cohort and case-control) studies were considered potentially eligible. Studies evaluating only the effects of statins on blood lipids or on acute/chronic allograft rejection risk were not included. Pre-transplant statin use was not evaluated as the study focused only on the effects of statin therapy following kidney transplantation. Uncontrolled trials, cross-sectional and descriptive studies, review articles and in vitro studies were also excluded.

Literature search

The literature search was conducted by systematically searching Medline (via PubMed), Scopus, Web of Science, CENTRAL (Cochrane Central Register of Controlled Trials) and Clinicaltrials.gov. In addition, Google Scholar was screened to provide grey literature coverage, while the “snowball” method [18] was applied by examining the reference lists of the studies included in the review. All searches were performed from inception till June 20, 2024, without any language restrictions. The search algorithm was constructed using both MeSH terms and key-words. The main search algorithm was as follows: “(“Hydroxymethylglutaryl-CoA Reductase Inhibitors“[Mesh] OR statin* OR atorvastatin OR rosuvastatin OR fluvastatin OR lovastatin OR pitavastatin OR pravastatin OR simvastatin OR cerivastatin) AND (“Kidney Transplantation“[Mesh] OR “kidney transplant*” OR “renal transplant*”)”.

Study selection

The evaluation of studies as eligible or not eligible followed three consecutive stages. Firstly, the titles and abstracts of all identified electronic articles were screened for potential eligibility. Secondly, all studies that were considered to fulfil the inclusion criteria were retrieved in full-text form. Subsequently, studies that examined different populations, did not report any of the outcomes of interest or met any of the exclusion criteria were excluded. The process of study selection was independently carried out by two authors, resolving any discrepancies regarding the eligibility of the articles through the consensus of all the authors.

Data extraction

The following data were extracted from the included studies: year of publication, country, eligibility criteria, sample size, study design, type of statin, statin dose, time from transplantation, median age, body mass index, estimated glomerular filtration rate (eGFR), low-density lipoprotein cholesterol, percentage of male sex, diabetes mellitus status, hypertension status, history of cardiovascular disease, type of immunosuppression (calcineurin inhibitors/mTOR inhibitors) and all the necessary information regarding the outcomes of interest. Specifically, relative risks (RR) along with their 95% confidence intervals were extracted. In observational studies, adjusted effect sizes were extracted from multivariate regression models. In case relative risks were not reported, the number of events in the total of patients was extracted. Data extraction was independently performed by two authors, who resolved any disagreements through their consensus.

Quality assessment

The risk of bias in randomized controlled trials was judged by applying the RoB-2 tool [19]. Specifically, low risk, some concerns or high risk of bias were assigned in the domains of randomization, deviations from intended interventions, missing outcome data, measurement of the outcome and selection of the reported results. The quality of cohort studies was assessed with the ROBINS-I tool [20], which considers the following domains: confounding, selection of participants, classification of interventions, departures from intended interventions, missing data, measurement of outcomes and selection of the reported results. Risk of bias evaluation was performed by two authors and any potential discrepancies were resolved after discussion with all the authors.

Statistical analysis

Pooling of studies was performed by fitting random-effects meta-analysis models, due to the expected existence of methodological heterogeneity across studies. The restricted maximum likelihood method was used for between-study variance estimation (τ²) [21]. The weight assigned to each included study was the inverse of the sum of the within-study variance (σ_i²) and the estimated between-study variance (τ²). The confidence intervals (CI) were set at 95%. Inter-study heterogeneity was quantified by the inconsistency index (I²), with values greater than 50% indicating remarkable heterogeneity [22]. The 95% prediction intervals were estimated to evaluate the effects to be expected by future studies in the field [23]. Publication bias was assessed by the trim-fill method [24], while the statistical significance (P < 0.10) of Egger’s regression and Begg & Mazumdar’s rank correlation tests was taken into account in the case of 10 or more studies [23]. Subgroup analysis was conducted based on study design (randomized controlled trial or observational study), location (Europe, North America, South America Asia or Oceania), sample size (< 400 or ≥ 400 patients), risk of bias (low, moderate or high) and type of calcineurin inhibitor (cyclosporine or tacrolimus). Meta-regression analysis was performed for endpoints with 4 or more studies, based on the following variables: age, sex, BMI, eGFR, diabetes mellitus and history of cardiovascular disease. All analyses were performed in R-4.4.0 (package “metafor” [25]).

Certainty of evidence

The certainty of the existing evidence was appraised following the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach [26]. Specifically, the quality of evidence could be classified as very low, low, moderate or high by evaluating the domains of study limitations, inconsistency, indirectness, imprecision and publication bias.

Results

Study selection

The PRISMA flowchart depicts the process of study selection (Fig. 1). The database search identified 1952 electronic articles; after deduplication, a total of 1068 articles were screened for potential eligibility, 45 of which were retrieved as full-texts. Subsequently, 18 studies [11, 27–43] were excluded because they did not report the outcomes of interest (n = 7), lacked a control group (n = 4), were partial duplicates of studies already included in the review, resulting in population overlap (n = 3), evaluated different populations (n = 3) or assessed the effects of pre-transplant statin use (n = 1) (Suppl. Table 1, Appendix 1). Overall, 27 studies [44–70] (10 randomized controlled trials and 17 observational studies) were finally included in the meta-analysis.

Fig. 1.

PRISMA search plot diagram

Included studies

Table 1 presents the methodological characteristics of the included studies. Eleven studies were conducted in North America, 7 in Europe, 3 in Asia (South Korea), 1 in South America (Brazil), 1 in Oceania (Australia) and 4 were international ones. Three studies presented analyses from the ALERT (Assessment of Lescol in Renal Transplant) trial, while 3 studies derived their data from the U.S. Renal Data System. In 2 studies, statins were exclusively administered for primary cardiovascular prevention. The median participant age was 48.4 years, while the median percentage of males was 60%. In cohort studies, patients could have been treated with more than one statin at different doses and intensities. All studies used calcineurin inhibitors as maintenance immunosuppression, with cyclosporine being exclusively administered in 11 studies. On the other hand, mTOR inhibitors were exclusively used in only 1 study. The median follow-up of studies was 60 months (range: 3 to 180 months). The exclusion criteria, outcomes of interest and definitions of MACE in the included studies are shown in Suppl. Table 2 (Appendix 2). Suppl. Table 3 (Appendix 2) presents details about matching and covariate adjustment in the included cohort studies.

Table 1.

Methodological characteristics of the included studies

Study	Design	Country	Sample size	Statin	Time from Tx †	Age (years)†	Male sex (%)	DM (%)	HTN (%)	BMI (kg/m2) †	eGFR (ml/min/ 1.73 m2) †	LDL-C (mmol/L) †	CVD (%)	CNI (%)	mTORi (%)	Follow-up (months) †
2024; Nazoiri [54]	RC	France	318	Low-intensity: 58.3%; Moderate-intensity: 68.6%; High-intensity: 11.5%	12.1 months	45.7	64.2	7.5	88.1	23.8	52	2.60	0	CsA: 23.3 Tacrolimus: 76.4	1.9	72
2024; Yim [55]	PC	South Korea	714	Any	Up to 12 months	45.6	64.1	25.5	92.4	22.8	72.6	2.16	8.4	CsA: 5.3 Tacrolimus: 94.7	7.1	84
2023b; Bae-USRDS [53]	RC	USA	58,264	Simvastatin: 46.9%; Atorvastatin: 45.2%; Pravastatin: 14.5%; Rosuvastatin: 9.7%	NR	51.4	60.3	35.3	NR	27.5	NR	NR	NR	CsA: 4.4 Tacrolimus: 91.8	3.1	36
2023a; Bae-USRDS [52]	RC	USA	57,699	Simvastatin: 46.9%; Atorvastatin: 45.2%; Pravastatin: 14.5%; Rosuvastatin: 9.7%	NR	51.1	60.4	35.3	NR	27.5	NR	NR	NR	CsA: 4.3 Tacrolimus: 90.2	3.0	36
2021; Anderson [51]	RC	Germany	622	Atorvastatin: 65.2%; Simvastatin: 20.8%; Fluvastatin: 5.2%; Rosuvastatin: 4.8%; Pravastatin: 4.4%	> 12 months	53.9	57.2	23.2	NR	26.6	NR	2.98	14.8	CsA: 39.5 Tacrolimus: 18.8	0	64.8
2022; Frasco [50]	RC	USA	1384	Atorvastatin: 54.3%; Rosuvastatin: 18.8%; Simvastatin: 12.1%; Pravastatin: 12.1%	Post-transplant	54	57.9	37.1	NR	27.5	NR	NR	NR	Tacrolimus: 100	1.2	12
2020; Szili-Torok [49]	PC	Netherlands	190	Atorvastatin: 52.6%; Fluvastatin: 27.4%; Simvastatin: 14.7%; Pravastatin: 5.3%	78 months	50.3	50.2	0	82	24.9	50.8	3.5	NR	CsA: 66.4 Tacrolimus: 8.2	0	115.2
2017; Han [48]	RC	South Korea	165	Any	6.7 months	47.4	55.2	17.6	92.7	21.0	NR	NR	15.8	Tacrolimus: 100	0	60
2017; Vangala-USRDS [4]	Nested C-C	USA	15,806	Any	55.6 months	51.2	59.6	74.6	NR	27.3	NR	NR	75.3	CsA: 18.5 Tacrolimus: 62.8	9.9	12
2014; Choe [46]	RC	South Korea	394	Atorvastatin (20 mg):45.3%; Fluvastatin (80 mg): 54.7	NR	40.1	58.4	0	72.1	22.0	62.5	NR	NR	CsA: 34.8 Tacrolimus: 67.0	0	60
2010; Younas [70]	RC	USA	615	Any	Up to 12 months	50	63	16.9	NR	26	NR	2.65	7.2	CsA: 0.3 Tacrolimus: >90%	8.5	79.7
2008; Wiesbauer [69]	RC	Austria	2041	Any	NR	48	59.9	18.8	81.1	NR	NR	NR	67.1	NR	NR	144
2008; Seron [68]	RCT	Spain	89	Fluvastatin (80 mg)	Post-transplant	42	57.3	0	NR	NR	NR	NR	0	CsA: 100	0	72
2007; Lisik [67]	RC	USA	325	Pravastatin: 73.3%; Atorvastatin: 25%; Simvastatin: 2%	Up to 3 months	44.9	60	NR	NR	NR	NR	NR	NR	CsA: 100	100	75
2007; Lopau [66]	C-C	Germany	150	Pravastatin (29.7 mg†)	4 days	48.5	66.7	18	NR	NR	62.9	3.75	NR	CsA: 100	0	48
2004; Prasad [65]	RC	Canada	314	Atorvastatin: 85%; Pravastatin: 7%; Simvastatin: 4%; Fluvastatin: 4%	NR	44.6	56.8	0	NR	NR	NR	2.24	NR	CsA: 58.7 Tacrolimus: 41.3	0	36
2004; Fellström-ALERT [60]	RCT	Inter- national	2102	Fluvastatin (40 mg)	62.4	49.8	66	18.8	74.9	25.8	60.3	4.1	23.5	CsA: 100	NR	61.2
2001; Holdaas [57]	RCT	Inter- national	364	Fluvastatin (40 mg)	Post-transplant	48.4	71.4	12.4	NR	NR	NR	2.98	NR	CsA: 100	0	3
1999; Lepre [45]	RCT	Australia	49	Simvastatin (10 mg)	> 12 months	51.4	34.7	14.3	77.6	NR	NR	4.63	NR	CsA: 57.1	0	9
1996; Katznelson [40]	RCT	USA	48	Pravastatin (20 mg)	7 days	47.3	69	NR	NR	NR	NR	NR	NR	CsA: 100	0	4
2002; Cosio [63]	RC	USA	1574	Any	< 24 months: 32.4% 24–60 months: 20%	42	59	22	NR	26	NR	NR	NR	CsA: 100	0	98.4
2010; Moreso [62]	RC	Spain	3682	Any	NR	46.5	63.1	5.3	NR	NR	NR	NR	NR	CsA: 70 Tacrolimus: 20.9	NR	180
1994; Arnadottir [61]	RCT	USA	40	Simvastatin (10 mg)	50	51.5	65	20	NR	25.3	NR	5.45	NR	CsA: 100	0	4
2005; Holdaas-ALERT [58]	RCT	Inter- national	1652	Fluvastatin (40 mg)	62.4	48.5	66.3	16.9	74.6	26	60.3	4.1	23.5	CsA: 100	NR	72
2003; Holdaas-ALERT [56]	RCT	Inter- national	2102	Fluvastatin (40 mg)	62.4	49.8	66	18.8	74.9	25.8	60.3	4.1	23.5	CsA: 100	NR	72
2001; Santos [60]	RCT	Brazil	67	Simvastatin (10 mg)	60.5	43.3	44.8	NR	NR	NR	65.9	4.67	NR	CsA: 77.6	0	57.5
2001; Kasiske [59]	RCT	USA	105	Simvastatin (10 mg)	Post-transplant	46.5	59	NR	NR	NR	NR	NR	NR	CsA: 100	0	3

^†Median values. RC: retrospective cohort; PC: prospective cohort; C-C: case-control; RCT: randomized controlled trial; DM: diabetes mellitus; HTN: hypertension; BMI: body mass index; CNI: calcineurin inhibitor; mTORi: mammalian target of rapamycin inhibitor; CsA: cyclosporine; eGFR: estimated glomerular filtration rate; LDL-C: low-density lipoprotein cholesterol; CVD: cardiovascular disease; Tx: transplantation; ALERT: Assessment of Lescol in Renal Transplant; USRDS: United States Renal Data System; NR: not reported

Risk of bias

The outcomes of the RoB-2 evaluation of randomized controlled trials are illustrated in Suppl. Figure 1 (Appendix 3). The ALERT trial was evaluated to be at low risk of bias, while in the remaining trials, some concerns of bias were raised in the domain of randomization, due to the lack of adequate information regarding allocation sequence generation and concealment. Suppl. Table 4 presents the outcomes of the ROBINS-I assessment of observational studies. Overall, 14 studies were judged to be at moderate risk of bias and 3 studies were judged to be at serious risk of bias. No critical risk of bias was identified in any study. The main source of potential bias was considered to be confounding, due to concerns of inadequate adjustment for important covariates that could affect the association between statin use and the outcomes of interest.

Outcomes

Table 2 presents the meta-analysis outcomes for the primary and secondary endpoints. The stratified analysis based on the type of calcineurin inhibitor co-administered is shown in Table 3, while the outcomes of the subgroup analysis based on sample size, location, study design and risk of bias are shown in Table 4. Appendix 4 (Suppl. Figures 2–11) includes the forest plots of efficacy and safety outcomes, while the respective funnel plots are displayed in Appendix 5 (Suppl. Figures 12–22). The outcomes of meta-regression analyses are presented in Appendix 6 (Suppl. Table 5).

Table 2.

Summary of findings table regarding the association of statin use with primary and secondary endpoints

Endpoint	Studies no.	Sample size	Follow-up (months)†	RR (95% CI)	95% prediction interval	I2	Trim-fill method		GRADE assessment
							Missing studies	New RR (95% CI)	Certainty of evidence	Downgrading
Efficacy
MACE	6	3621	68.4 (48–72)	0.87 (0.67–0.96) *	0.67–0.96 *	0%	0	0.87 (0.67–0.96) *	Moderate	Study limitations
Mortality	15	70,750	72 (3-180)	0.84 (0.74–0.94) *	0.65–1.08	40.4%	1	0.84 (0.75–0.94) *	Low	Study limitations, inconsistency
Graft loss	9	10,255	72 (3-180)	0.72 (0.48–1.08)	0.23–2.24	90.2%	0	0.72 (0.48–1.08)	Very low	Study limitations, imprecision, inconsistency
Safety
Hepatotoxicity	6	60,641	54 (3–72)	0.81 (0.70–0.93) *	0.70–0.93 *	0%	2	0.80 (0.70–0.92) *	Moderate	Study limitations
Creatine kinase elevation	4	2624	40.5 (3–72)	0.97 (0.50–1.89)	0.50–1.89	0%	1	1.02 (0.53–1.96)	Low	Study limitations, imprecision
Rhabdomyolysis	2	59,801	54 (36–72)	1.37 (1.10–1.70) *	1.10–1.70 *	0%	NA	NA	Low	Study limitations, imprecision
Diabetes mellitus	4	58,597	48 (36-115.2)	1.11 (0.38–3.27)	0.11–11.07	92.4%	1	0.79 (0.26–2.39)	Very low	Study limitations, imprecision, inconsistency
Cataract	1	57,699	36	1.22 (1.14–1.31) *	NA	NA	NA	NA	Moderate	Study limitations
Hip fracture	2	15,846	8 (4–12)	0.90 (0.68–1.20)	0.68–1.20	0%	NA	NA	Low	Study limitations, imprecision
Venous thromboembolism	1	1384	12	0.92 (0.39–2.19)	NA	NA	NA	NA	Low	Study limitations, imprecision
Cancer	1	2102	72	0.94 (0.82–1.07)	NA	NA	NA	NA	Moderate	Imprecision

Asterisks denote statistically significant results

^†Median (range); MACE: major adverse cardiovascular events; RR: relative risk; CI: confidence intervals: I²: inconsistency index; GRADE: Grading of Recommendations, Assessment, Development, and Evaluations; NA: not applicable

Table 3.

Association of statin use with primary and secondary endpoints, stratified by the type of the co-administered calcineurin inhibitor

Endpoint	Calcineurin inhibitor				P
	Cyclosporine		Tacrolimus
	Studies no.	RR (95% CI)	Studies no.	RR (95% CI)
MACE	3	1.46 (0.38–5.60)	2	0.52 (0.14–1.88)	0.344
Mortality	3	0.96 (0.89–1.03)	9	0.82 (0.67-1.00)	0.393
Graft loss	7	0.83 (0.43–1.58)	1	0.98 (0.82–1.18)	0.812
Hepatotoxicity	4	0.64 (0.44–0.92) *	1	0.84 (0.72–0.98) *	0.178
Creatine kinase elevation	3	1.17 (0.56–2.42)	0	-	NA
Rhabdomyolysis	2	1.09 (0.57–2.10)	1	1.41 (1.10–1.81) *	0.478
Diabetes mellitus	1	1.11 (0.89–1.37)	1	1.13 (1.07–1.20) *	NA
Cataract	1	1.22 (0.93–1.62)	1	1.22 (1.14–1.31) *	NA
Hip fracture	1	3.00 (0.13–69.5)	0	-	NA
Venous thromboembolism	0	-	1	0.92 (0.39–2.19)	NA
Cancer	1	0.94 (0.82–1.07)	0	-	NA

Asterisks denote statistically significant results. P-values correspond to the significance of the interaction between the type of calcineurin inhibitor and statin effects

RR: relative risk; CI: confidence intervals; NA: not applicable

Table 4.

Subgroup analyses based on study sample size, location, design and risk of bias

Endpoint	Sample size			Location				Study design			Risk of bias
	< 400	≥ 400	P	Europe	North America	Asia	P	RCT	Observational	P	Low	Moderate	Serious	P
MACE	0.60 (0.28–1.32)	0.83 (0.68-1.00)	0.647	0.84 (0.59–1.20)	-	0.73 (0.14–3.92)	0.975	0.79 * (0.63–0.99)	0.80 (0.48–1.34)	0.971	0.79 * (0.63–0.99)	0.81 (0.45–1.47)	0.73 (0.17–3.15)	0.994
Mortality	0.66 (0.30–1.46)	0.86 * (0.78–0.96)	0.194	0.79 * (0.67–0.93)	0.83 * (0.70–0.99)	0.28 * (0.11–0.71)	0.108	1.02 (0.83–1.24)	0.80 * (0.70–0.92)	0.164	1.02 (0.83–1.25)	0.80 * (0.70–0.92)	2.92 (0.12–70.6)	0.284
Graft loss	0.76 (0.33–1.75)	0.64 (0.36–1.15)	0.766	0.75 * (0.65–0.85)	0.67 (0.27–1.67)	0.23 * (0.13–0.42)	< 0.001	1.10 (0.88–1.37)	0.54 * (0.32–0.90)	0.039	1.05 (0.83–1.33)	0.70 (0.43–1.14)	0.39 (0.08–1.95)	0.681
Hepatotoxicity	0.66 (0.35–1.24)	0.82 * (0.71–0.94)	0.456	1.33 (0.26–6.87)	0.82 * (0.71–0.95)	-	0.656	0.65 (0.40–1.06)	0.82 * (0.71–0.95)	0.377	0.71 (0.34–1.47)	0.81 * (0.70–0.93)	-	0.720
Creatine kinase elevation	0.76 (0.26–2.20)	1.20 (0.37–3.93)	0.512	0.33 (0.01–7.80)	-	-	0.322	0.97 (0.50–1.89)	-	NA	1.20 (0.37–3.93)	0.76 (0.26–2.20)	-	0.632
Rhabdomyolysis	-	1.37 (1.10–1.70)	NA	-	1.37 (1.10–1.70)	-	NA	1.00 (0.06-16.0)	1.37 (1.10–1.70)	NA	1.00 (0.06-16.0)	1.37 * (1.10–1.71)	-	NA
Diabetes mellitus	1.22 (0.22–5.61)	1.12 * (1.07–1.18)	0.998	3.86 * (1.21–12.29)	0.54 (0.12–2.47)	1.70 (0.85–3.40)	0.357	-	1.11 (0.38–3.27)	NA	-	1.81 (0.55–5.88)	0.64 (0.09–4.42)	0.348
Cataract	-	1.22 * (1.14–1.31)	NA	-	1.22 * (1.14–1.31)	-	NA	-	1.22 * (1.14–1.31)	NA	-	1.22 * (1.14–1.31)	-	NA
Hip fracture	0.89 (0.67–1.19)	3.00 (0.13–69.5)	NA	-	0.90 (0.68–1.20)	-	NA	0.89 (0.67–1.19)	3.00 (0.13–69.5)	NA	-	0.90 (0.68–1.20)	-	NA
Venous thromboembolism	-	0.92 (0.39–2.19)	NA	-	0.92 (0.39–2.19)	-	NA	-	0.92 (0.39–2.19)	NA	-	0.92 (0.39–2.19)	-	NA
Cancer	-	0.94 (0.82–1.07)	NA	-	-	-	NA	0.94 (0.82–1.07)	-	NA	0.94 (0.82–1.07)	-	-	NA

Asterisks denote statistical significance. P-values correspond to the significance of the difference of statin effects depending on study sample size, location, design and risk of bias

RR: relative risk; CI: confidence intervals; NA: not applicable; MACE: major adverse cardiovascular events; RCT: randomized controlled trials

Cardiovascular events

The association between statin use and MACE risk was examined in 6 studies (1 randomized controlled trial and 5 cohort studies). The administration of statins was associated with a significantly lower risk of MACE [Relative risk (RR): 0.87, 95% CI: 0.67 to 0.96, 3621 participants] (Fig. 2). No statistical heterogeneity was observed (I²: 0%) and thus the 95% prediction intervals were identical to the confidence intervals. No missing studies were identified by the trim-fill method (Suppl. Figure 12, Appendix 5). No statistically significant interaction with the type of calcineurin inhibitor was noted (P = 0.344) (Table 3). Similarly, the meta-analysis outcome was not significantly influenced by study sample size (P = 0.647), location (P = 0.975), design (P = 0.971) or risk of bias (P = 0.994) (Table 4). The meta-regression analysis showed no significant effects of age, sex, BMI, eGFR, history of cardiovascular disease and diabetes mellitus (Suppl. Table 5). The certainty of evidence was evaluated as moderate, downgrading for study limitations as most studies were observational ones with a moderate risk of bias.

Fig. 2.

Forest plot illustrating the association between statin use and risk of major adverse cardiovascular events (3621 participants). RR: relative risk; CI: confidence intervals; RE: random-effects; df: degrees of freedom

Patient survival

Overall patient survival was assessed in 15 studies (6 randomized controlled trials and 9 cohort studies). Statin use was associated with a significantly lower mortality risk (RR: 0.84, 95% CI: 0.74 to 0.94, 70,750 participants). Moderate statistical heterogeneity was observed (I²: 40.4%), while the 95% prediction intervals ranged from 0.65 to 1.08. After imputing potentially missing studies, the new trim-fill estimate remained statistically significant. No significant funnel plot asymmetry was present (Egger’s P = 0.113, Begg & Mazumdar’s P = 0.923) (Suppl. Figure 13). Subgroup analyses indicated no significant influence of calcineurin inhibitor type, study sample size, location, design or risk of bias (P > 0.05) (Tables 3 and 4). The meta-regression analysis suggested that the observed association was more pronounced in studies with participants of younger age and high eGFR on average (Suppl. Table 5). The certainty of evidence was judged to be low. Concerns about study limitations were raised as the majority of studies were to be at moderate risk of bias. In addition, the certainty of evidence was downgraded due to inconsistency, as statistical heterogeneity was moderate and the 95% prediction intervals crossed the null hypothesis.

Graft survival

The endpoint of graft survival was evaluated in 9 studies (5 randomized controlled trials and 4 cohort studies). The risk of graft loss was not significantly different between the statin group and the control group (RR: 0.72, 95% CI: 0.48 to 1.08, 10,255 participants). The level of statistical heterogeneity was remarkable (I²: 90.2%), and the 95% prediction intervals ranged from 0.23 to 2.24. No missing studies were identified by the trim-fill method (Suppl. Figure 14). Subgroup analysis indicated that the meta-analysis estimates significantly differed depending on the study location and design (Table 4). Specifically, statin use was linked to a significantly lower graft loss risk when separately pooling studies conducted in Asia (RR: 0.23, 95% CI: 0.13 to 0.42) or observational studies (RR: 0.54, 95% CI: 0.32 to 0.90). The meta-regression analysis suggested that participants’ BMI may have affected the pooled estimate (Suppl. Table 5). The certainty of evidence was assessed to be very low, downgrading for study limitations, imprecision and inconsistency. In particular, in most studies, the risk of bias was evaluated as moderate, while both 95% CI and prediction intervals were wide.

Hepatotoxicity

The risk of liver toxicity was examined in 4 randomized controlled trials and 2 cohort studies. A pooled analysis of the studies indicated that statin use was associated with a significantly lower risk of liver injury (RR: 0.81, 95% CI: 0.70 to 0.93, 60,641 participants). No statistical heterogeneity was noted (I²: 0%), and the 95% prediction intervals were similar to the confidence intervals. After the imputation of 2 potentially missing studies by the trim-fill method, the new estimate remained statistically significant (new RR: 0.80, 95% C: 0.70 to 0.93) (Suppl. Figure 15). The meta-analysis outcome was not significantly influenced by calcineurin inhibitor type, study sample size, location, design or risk of bias (P > 0.05) (Tables 3 and 4). The meta-regression analysis showed no significant effect of the examined covariates (Suppl. Table 5). The certainty of evidence was evaluated as moderate due to concerns of study limitations (moderate risk of bias in 5 out of 6 studies).

Creatine kinase elevation

The endpoint of creatine kinase elevation was reported by 4 randomized controlled trials. The risk of creatine kinase elevation was similar between statin-treated and nontreated patients (RR: 0.97, 95% CI: 0.50 to 1.89). No statistical heterogeneity was present (I²: 0%); hence, the 95% prediction intervals were identical to the confidence intervals. The trim-fill method identified 1 potentially missing study (new RR: 1.02, 95% CI: 0.53 to 1.96) (Suppl. Figure 16). The outcome did not differ among cyclosporine or tacrolimus users and was not significantly affected by study sample size, location, design or risk of bias (P > 0.05) (Tables 3 and 4). The meta-regression analysis proposed that the outcome was not significantly influenced by age, sex, BMI, eGFR, history of cardiovascular disease and diabetes mellitus (Suppl. Table 5). The certainty of evidence was appraised as low, downgrading for study limitations (moderate risk of bias in 3 out of 4 trials) and imprecision (wide 95% CI).

Rhabdomyolysis

The potential link between statin administration and rhabdomyolysis was investigated in 2 studies (1 randomized controlled trial and 1 retrospective cohort study). Bae et al [52]. identified rhabdomyolysis cases through Medicare Parts A and B claims using International Classification of Diseases (ICD) codes (ICD-9: 728.88 and ICD-10: M62.82). In the ALERT trial [56], rhabdomyolysis was developed in 2 patients (1 in each group) and was attributed to severe trauma. Overall, statin use was associated with a significantly greater risk of rhabdomyolysis (RR: 1.37, 95% CI: 1.10 to 1.70). No statistical heterogeneity was observed (I²: 0%), and thus, the 95% prediction intervals did not differ from the estimated confidence intervals. The stratified analysis indicated that the association remained significant only among tacrolimus users, although no significant subgroup effect could be ascertained (Table 3). Study sample size, location, design and risk of bias did not significantly influence the outcome (Table 4). The certainty of evidence was evaluated to be low, downgrading for study limitations and imprecision, reflecting the width of the estimated 95% CI.

Post-transplant diabetes mellitus

The potential association between statin use and the development of post-transplant diabetes mellitus was examined in 4 studies (1 prospective cohort study and 3 retrospective cohort studies). Two of them [49, 52] indicated that statin administration was linked to a significantly greater risk of diabetes mellitus, 1 study showed no significant association [46], while a significantly lower diabetes mellitus risk was suggested by Prasad et al. [65]. The pooling of all studies demonstrated no significant association between statin intake and post-transplant diabetes mellitus risk (RR: 1.11, 95% CI: 0.38 to 3.27, 58,597 participants), with high statistical inter-study heterogeneity (I²: 92.4%). The 95% prediction intervals ranged from 0.11 to 11.07. The trim-fill method identified 1 potentially missing study (new RR: 0.79, 95% CI: 0.26 to 2.39) (Suppl. Figure 18). Concerning the type of calcineurin inhibitor administered, the available data suggested a significant association between statins and diabetes mellitus among tacrolimus-treated patients (1 study, RR: 1.13, 95% CI: 1.07 to 1.20) (Table 3). The stratified analyses showed a positive association between statin use and diabetes mellitus development when studies conducted in Europe and those with a large sample size (≥ 400 participants) were pooled, although no statistically significant subgroup effects were calculated (Table 4). The certainty of evidence was appraised to be very low, downgrading for study limitations, imprecision and inconsistency. Specifically, the endpoint of post-transplant diabetes mellitus was evaluated only by observational studies, judged to be at moderate to serious risk of bias. Both methodological and statistical heterogeneity were noted, while the 95% CI and prediction intervals were wide, including both negative and positive effects.

Hip fracture

The endpoint of hip fracture was examined in 2 studies (1 randomized controlled trial and 1 nested case-control study). No significant association was observed between the administration of statins and hip fracture after kidney transplantation (RR: 0.90, 95% CI: 0.68 to 1.20, 15,846 participants). No statistical heterogeneity was noted (I²: 0%). Subgroup analysis indicated no significant effects of sample size or study design (Table 4). The certainty of evidence was appraised as low, downgrading for study limitations (2 studies at a moderate risk of bias) and imprecision (wide 95% CI).

Cataract, venous thromboembolism and cancer

The endpoint of cataract was assessed in 1 retrospective cohort study [52]. Cataract was identified through Medicare Parts A and B claims (ICD-9: 366.1x, 366.3x, 366.4x, 366.5x, 366.8, 366.9 or ICD-10: H25.x, H26.2x, H26.3x, H26.4x, H26.8, H26.9), which indicated that the use of statins may be significantly associated with cataract risk after kidney transplantation (RR: 1.22, 95% CI: 1.14 to 1.31, 57,699 participants). In particular, the estimated incidence of cataract was 22% at 5 years following statin use compared to 12% among statin non-users. The association remained statistically significant only among tacrolimus users (RR: 1.22, 95% CI: 1.14 to 1.31) (Table 3). The certainty of evidence was evaluated to be moderate, due to concerns about study limitations, as the endpoint was only evaluated by 1 observational study at a moderate risk of bias. One retrospective study [50] examined the risk of venous thromboembolism, and suggested no statistically significant association with the use of statins (RR: 0.92, 95% CI: 0.39 to 2.19, 1384 participants). Specifically, within 1 year, a venous thromboembolic event occurred in 9 out of 223 statin users and in 44 out of 1161 statin non-users. All patients were treated with tacrolimus. The certainty of evidence was judged as low, due to concerns of study limitations (1 observational study at a moderate risk of bias) and imprecision (wide 95% CI). Cancer risk was investigated in the ALERT trial [58], which indicated no statistically significant association with statin administration (RR: 0.94, 95% CI: 0.82 to 1.07, 2102 participants). In particular, malignancies were detected in 296 out of 1045 statin users compared to 316 out of 1049 patients treated with placebo. All kidney transplant recipients were treated with cyclosporine. The certainty of evidence was evaluated as moderate, downgrading for imprecision, as reflected by the width of the 95% CI.

Discussion

Based on 27 studies, the present systematic review and meta-analysis collected the available literature and evaluated the efficacy and safety of statins in kidney transplant recipients. Current evidence suggests that statin therapy may be associated with a cardiovascular benefit, reflected by the significant reduction in MACE risk among statin-treated patients. Statin use after transplantation was also linked to a significant improvement in overall survival, although graft survival may not be affected. The safety profile of statins appeared to be acceptable since no increased risk of severe adverse effects, especially liver toxicity, was detected. The rate of creatine kinase elevation was not associated with statin administration, although a minor increase in the risk of rhabdomyolysis was observed among statin-treated kidney transplant recipients.

The benefits of statins in the prevention of cardiovascular events after kidney transplantation may be mediated by their lipid-lowering action in conjunction with potential pleiotropic effects. A recent meta-analysis confirmed that statin therapy may effectively enhance the lipid profile of kidney transplant recipients by significantly decreasing serum triglyceride and low-density lipoprotein cholesterol levels [71]. Statin administration in kidney transplant recipients has been proposed to improve endothelial function not only through blood lipid reduction but also via an increase in nitric oxide bioavailability [72]. In addition, experimental data have shown that after kidney transplantation, statins may reduce the generation of reactive oxygen species, exerting antioxidant effects, as reflected by increased levels of glutathione peroxidase [73]. The beneficial impact of statins on cardiovascular health after kidney transplantation may also be promoted by their anti-inflammatory properties, resulting in lower levels of peptides involved in vascular inflammation [74].

Despite the cardiovascular benefits of statins, hesitancy in their prescription may arise due to concerns about possible interactions with calcineurin inhibitors that may increase the risk of toxicity. Specifically, most statins are metabolized by cytochrome P450 3A4, which is known to be inhibited by cyclosporine [75]. Moreover, cyclosporine may inhibit the hepatocellular uptake of statins via P-glycoprotein and organic anion-transporting polypeptides (OATP1B1), leading to increased statin exposure [76]. However, the present study raised no concerns about liver damage, as statin therapy was even linked to a significantly lower risk of hepatotoxicity. As in the general population, statin prescription may be limited by concerns about temporary transaminasemia, although the incidence of idiosyncratic drug-induced liver injury is low (1.2/100,000 statin users) [77]. The outcome of the present meta-analysis is in accordance with recent research evidence supporting the potential protective effects of statin against liver disease [78]. In particular, the possible hepatoprotective properties of statins are based on the inhibition of the prenylation of small guanosine triphosphate hydrolases, leading to reduced inflammation and oxidative stress [79].

On the other hand, rhabdomyolysis was more common among statin-treated kidney transplant recipients, although the increase in the risk was comparable to that estimated in studies conducted in the general population [80]. It should also be noted that the link between statin therapy and elevated rhabdomyolysis risk was based on the retrospective study of Bae et al [52]. , in which simvastatin, a statin that has been associated with the highest rhabdomyolysis risk in pharmacovigilance studies, was used by the majority of patients [81].

Remarkable interstudy heterogeneity exists concerning the risk of post-transplant diabetes mellitus. Studies in the general population have suggested a modest increase in impaired glucose metabolism among statin-treated individuals [82], which is based on the statin-induced increase in pancreatic low-density lipoprotein receptor expression, leading to beta-cell lipotoxicity and reduced insulin secretion [83]. In this context, Szili-Torok et al [49]. have indicated that the majority (73.3%) of post-transplant diabetes cases occurred among statin-treated patients, while Bae et al [52]. reported an increase in posttransplant diabetes risk of 13% among statin users receiving tacrolimus as maintenance immunosuppression. In contrast, the opposite outcome was suggested by the study of Prasad et al. [65] although baseline differences in the type of calcineurin inhibitor between the compared groups may have confounded the estimated effects. As a result, further evidence is needed to reach firm conclusions regarding the true impact of statin therapy on diabetes mellitus risk, depending on maintenance immunosuppression and comorbidities.

Current evidence concerning cataract comes from 1 retrospective study, which revealed that the risk of cataract increased by 22% among statin-treated kidney transplant recipients. A similar outcome has been estimated in the general population [84], while animal studies have confirmed the dose-dependent cataractogenic effects of statins [85]. It has been hypothesized that the statin-induced inhibition of cholesterol biosynthesis may affect the development of the lens epithelium, which requires a high cholesterol content to preserve its transparency [86]. Consequently, cataract may be considered as an adverse effect that should be expected but may not limit the prescription of statins in kidney transplant recipients.

Evidence regarding statin type, intensity and dosing after kidney transplantation remains limited. According to the KDIGO guidelines, all kidney transplant recipients should be evaluated for statin treatment, although follow-up blood lipid measurements are discouraged. As a result, a “fire and forget” strategy is recommended and no specific low-density lipoprotein cholesterol goals have been established [13]. A retrospective cohort study has indicated that the use of higher-intensity statins before kidney transplantation is associated with the greatest posttransplant survival benefit [36]. However, given the increased risk of adverse events among kidney transplant recipients and the lack of adequate safety data, lower statin doses are recommended. Specifically, for patients with eGFR below 60 ml/min/1.73 m², an equivalent atorvastatin dose of 20 mg is suggested [13]. The choice of statin type may be influenced by concerns about potential drug interactions. In this context, fluvastatin has been widely studied in the kidney transplant population due to the lack of significant interaction with cyclosporine. Specifically, fluvastatin is not a P-glycoprotein substrate, while it is metabolized not only by P450 3A4, but also by multiple cytochrome enzymes, such as 2C9 and 2C8 [87]. On the other hand, different pharmacokinetics are anticipated in patients receiving maintenance immunosuppression with cyclosporine and tacrolimus. In particular, tacrolimus has been proposed to have weaker inhibitory effects against cytochrome P450 and P-glycoprotein activity [88] and thus its combination with atorvastatin has been shown to be both safe and efficacious [89].

Strengths and limitations

The present study has several strengths. A comprehensive literature search with strict selection criteria was applied, limiting the possibility of any article loss. Inter-study heterogeneity was both quantified and explored, using various subgroup analyses. The risk of bias in the included studies has been evaluated using validated tools, while the certainty of evidence per outcome has been critically appraised, allowing a realistic assessment of existing data in the field. Downgrading of the quality of evidence occurred mainly due to study limitations since well-designed randomized controlled trials are currently limited and data are mainly derived from small trials with short follow-up periods or cohort studies prone to selection biases and residual confounding. Although meta-regression analysis was performed to assess the effects of potentially important covariates, it should be acknowledged that ecological fallacy may complicate the interpretability of meta-regression analyses examining individual patient-level characteristics that were reported as aggregate data in the original studies. Meta-regression analysis should be also interpreted with caution due to missing data and the small number of included studies. Heterogeneity may have also arisen from differenced in the definitions of endpoints across the included studies, especially concerning the definition of MACE. It should be also highlighted that the outcomes of some serious adverse effects, such as cataract, venous thromboembolism and cancer were only evaluated by one study and thus further validation is necessary. Furthermore, limited evidence was available regarding the potential differential statin effects among cyclosporine or tacrolimus users, while only few patients were treated with mTOR inhibitors. It is also important to note that statin dosing was not consistently reported in the included studies and thus the optimal prescriptions strategy remains unclear.

Implication for current clinical practice and future research

This meta-analysis suggests that clinicians should be encouraged to evaluate all kidney transplant recipients for statin therapy, given the potentially important clinical benefits in regards to cardiovascular risk reduction and improvement of overall survival. Nonetheless, statin prescription requires individualization in order to minimize the risk of toxicity and drug-drug interactions. Therefore, further large-scale research is needed to enable decision-making concerning statin intensity, dosing and treatment goals. In particular, it remains to be clarified whether the administration of high-intensity statins may be associated with more pronounced clinical benefits in terms of cardiovascular disease and patient survival, without increasing significantly the risk of adverse events. In addition, since most clinical trials have included patients receiving cyclosporin, future studies should focus on tacrolimus-treated kidney transplant recipients, aiming to shed more light on the actual risk of statin interaction with tacrolimus blood levels. This would also enable the drawing of clinical decisions about how early statin therapy could be initiated following kidney transplantation, especially in patients requiring statins for secondary cardiovascular prevention. The clinical utility of follow-up low-density lipoprotein cholesterol is currently unknown in the kidney transplant population and thus further high-quality evidence is needed to clarify whether the attainment of specific lipid targets exerts significant prognostic effects.

Conclusions

Among kidney transplant recipients, statin therapy is associated with significant benefits in terms of cardiovascular event reduction and survival improvement. Statin administration is well-tolerated, being associated with minor increases in the risk of rhabdomyolysis and cataract. Further research in large scale is needed to establish the favorable cardiovascular effects of statins and determine the subpopulation of kidney transplant recipients that may be safely targeted for higher-intensity statin treatment.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conception and design: I. Bellos. Analysis and interpretation of the data: I. Bellos, V. Benetou. Drafting of the article: I. Bellos, V. Benetou. Critical revision of the article for important intellectual content: P. Lagiou, S. Marinaki. Final approval of the article: I. Bellos, V. Benetou, P. Lagiou, S. Marinaki. Statistical expertise: I. Bellos, V. Benetou. Collection and assembly of data: I. Bellos, V. Benetou, S. Marinaki.

Funding

None.

Data availability

The data extracted from the original studies are available in the supplementary appendix. All data are available from the corresponding author upon request.

Declarations

Competing interests

The authors declare no competing interests.