Effects of perioperative statins on patient outcomes after noncardiac surgery: a meta-analysis

Abstract

Background: Cardiovascular complications are strongly correlated with a higher risk of mortality during follow-up after noncardiac surgery. However, controversy remains regarding whether perioperative administration of hydroxymethylglutaryl-CoA reductase inhibitors (statins) has a beneficial effect on patient outcomes.

Objective: We performed a meta-analysis to validate the hypothesis that perioperative statins improve patient outcomes after noncardiac surgery.

Methods: Electronic databases (PubMed, Web of Science, EMBASE, and the Cochrane Library) were searched for randomized controlled trials (RCTs) published up to 10 November 2017. RCTs were eligible for inclusion if they compared perioperative statin treatment with control treatment in patients scheduled for noncardiac surgery and reported data pertaining to clinical outcomes.

Results: Twelve RCTs involving 4707 patients (2371 in the perioperative statin group and 2336 in the control group) were ultimately included in this meta-analysis. The incidences of postoperative myocardial infarction, composite of death/myocardial infarction/stroke and new cases of atrial fibrillation were all lower in patients treated with statins than in control group patients, as shown by the fixed-effects model (odds ratio (OR) = 0.460, 95% confidence interval (CI) = 0.324–0.653, p = 0 for myocardial infarction; OR = 0.617, 95% CI = 0.476–0.801, p = 0 for composite of death/myocardial infarction/stroke; OR = 0.406, 95% CI = 0.247–0.666, p = 0 for new atrial fibrillation). No significant differences in the incidences of stroke or transient ischemic attack, all-cause mortality and cardiovascular mortality were observed between the statin and control arms.

Conclusions: This meta-analysis supports the hypothesis that perioperative statins effectively reduce the incidences of postoperative myocardial infarction, composite of death/myocardial infarction/stroke and new cases of atrial fibrillation in patients undergoing noncardiac surgery.

Key Messages

• Cardiovascular complications are strongly correlated with a higher risk of mortality during follow-up after noncardiac surgery.

• We performed a meta-analysis to confirm the hypothesis that perioperative statins improve patient outcomes after noncardiac surgery.

Keywords:

• Statins
• hydroxymethylglutaryl-CoA reductase inhibitor
• noncardiac surgery
• meta-analysis

Introduction

Three hundred million patients worldwide undergo major noncardiac surgery each year [1]. Despite great advancements in perioperative treatment, more than 1 million patients die within 30 days after surgery [2], and cardiovascular complications are strongly correlated with a higher risk of mortality during follow-up [3,4].

Several randomized controlled trials (RCTs) have recently assessed the effects of perioperative hydroxymethylglutaryl-CoA reductase inhibitors (statins) on patient outcomes after noncardiac surgery. However, all these studies were relatively small, which might have reduced the statistical power. Thus, we performed a meta-analysis to confirm the hypothesis that statins improve patient outcomes after noncardiac surgery.

Methods

Data sources and selection criteria

Electronic databases (PubMed, Web of Science, EMBASE and the Cochrane Library) were searched for RCTs published up to 10 November 2017. The following keywords were used: (“hydroxymethylglutaryl-CoA Reductase Inhibitors” OR = “HMG-CoA reductase inhibitors” OR = “statin” OR = “statins” OR = “lipid-lowering therapy” OR = “lipid-lowering treatment” OR = “lipid pharmacotherapy” OR = “atorvastatin” OR = “pravastatin” OR = “fluvastatin” OR = “simvastatin”) AND (“noncardiac surgery” OR = “non-cardiac vascular surgery” OR = “no-cardiac surgery” OR = “vascular surgery” OR = “vascular surgical procedures” OR = “peripheral arterial disease” OR = “peripheral vascular diseases” OR = “endarterectomy, carotid” OR = “carotid endarterectomy” OR = “aortic aneurysm, abdominal” OR = “aortic aneurysm repair” OR = “lower limb revascularization” OR = “lower extremity revascularization” OR = “lower limb bypass” OR = “lower extremity bypass” OR = “infrainguinal bypass”). Reference lists were also reviewed to identify relevant studies, regardless of the language in which the studies were published.

The following studies were included in the analysis: (1) RCTs; (2) studies including subjects undergoing noncardiac surgery; and (3) studies in which statins were administered to the treatment group and placebo to the control group. Non-RCTs, cohort studies, case reports, commentaries, and retrospective studies were excluded.

Study end-points and data extraction

The primary end-point was all-cause mortality, and the secondary end-points were cardiovascular mortality, myocardial infarction (MI), atrial fibrillation, stroke and MACE (composite of death, MI and stroke). Two reviewers independently extracted data for the following parameters from the identified studies: the first author, publication year, country, ages, clinical characteristics, type of procedure, type of statin therapy and number of patients. The risk of bias in the included RCTs was assessed with the method recommended by the Cochrane Collaboration [5].

Statistical analysis

Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for all outcomes and compared between the statin and control groups. Cochran’s test and I² statistical means were used to assess the heterogeneity among trials [6]. A random-effects model (the DerSimonian and Laird method) was used when p < .10. Otherwise, a fixed-effects model (Mantel–Haenszel method) was employed [7,8]. We used the Begg and Egger tests to determine whether publication bias was present with respect to all outcomes [9]. All statistical analyses were conducted with Stata v.12.0 (Stata Corp., College Station, TX) and RevMan 5.2 (the Cochrane Collaboration) software. A p value <.05 was considered statistically significant.

Results

A flow chart describing the study selection process is shown in Figure 1. Twelve RCTs involving 4707 patients (2371 in the perioperative statin group and 2336 in the control group) were ultimately included in this meta-analysis [10–21]. The characteristics of the included studies are summarized in Table 1. Atorvastatin was the intervention drug in eight studies; its dose ranged from 10 mg to 80 mg [1,3,12–16,20]. Rosuvastatin was the intervention drug in the other studies; it was administered at doses of 80 mg [5,17], 20 mg [21] and 10 mg [10]. The results of the risk-of-bias assessment for the included studies are reported in Figure 2.

Figure 1. Flow chart of the study selection process for this meta-analysis.

Figure 2. Risk-of-bias assessment summary.

Table 1. Baseline demographic and clinical characteristics.

									Total number of patients
Author	Country	Mean age, years	Male gender,%	CAD,%	Diabetes,%	Hypertension,%	Procedure	Duration, type and dose of statin therapy	Statin	No statin
Durazzo [13]	Brazil	65.92 vs 68.38 (p = .208)	80 vs 78 (p = 1.000)	30 vs 28 (p = 1.000)	18 vs 16 (p = 1.000)	76 vs 70 (p = .653)	Elective noncardiac arterial vascular surgery: aortic repair, infrainguinal arterial bypass, CEA	Atorvastatin 20 mg, 45 days	44	46
Schouten [19]	The Netherlands	66.0 vs 65.8	77.6 vs. 72.1	29.2 vs 26.7	22 vs 17	56.8 vs 57.9	Carotid artery surgery, open and endovascular abdominal aortic surgery, lower-limb arterial surgery	Fluvastatin 80 mg, 37 days	250	247
Dunkelgrun [12]	The Netherlands	65.4 vs 65.6	61.5 vs 57.1	NR	12.1 vs 9.0	NR	General surgery, urology, orthopaedics, ear-nose-throat, gynaecology, plastic surgery, dental surgery, ophthalmology	Fluvastatin 80 mg, 64 days	265	268
Puato [17]	Italy	NR	NR	NR	NR	NR	CEA	Atorvastatin 10 mg or 80 mg, 12 weeks ±2 days	39	19
Neilipovitz [15]	Canada	NR	NR	NR	NR	NR	Aortic surgery, infrainguinal revascularization	Atorvastatin 80 mg, 7 days	22	8
Luijendijk [14]	The Netherlands	35.8 vs 36.8 (p = .61)	56 vs 69 (p = .44)	NR	1 vs 1 (p = .72)	60 vs 70 (p = .49)	Surgical repair of aortic coarctation	Atorvastatin 80 mg, 3 months	68	65
Xia [20]	China	66 ± 10 vs 67 ± 11 (p = .29)	78 vs 81 (p = .51)	100 vs 100	37 vs 35 (p = .64)	77 vs 82 (p = .18)	Surgery for suppurative appendicitis, cholecystitis, cholangitis, pancreatitis, peptic ulcer perforation, urinary calculi	Atorvastatin 80 mg, 2 h before surgery	250	250
Amar [10]	USA	65 ± 9 vs 65 ± 9 (p = .88)	53 vs 60 (p = .54)	5 vs 2 (p = .61)	7 vs 7 (p = 1.00)	30 vs 38 (p = .46)	Pulmonary resection	Atorvastatin 40 mg, 2 weeks	43	45
Xia [21]	China	67 ± 10 vs 68 ± 11 (p = .26)	72 vs 74 (p = .63)	100 vs 100	33 vs 32 (p = .78)	65 vs 64 (p = .66)	Surgery for suppurative appendicitis, cholecystitis, cholangitis, pancreatitis, peptic ulcer perforation, urinary calculi	Rosuvastatin 20 mg, 2 h before surgery	275	275
Qin [18]	China	56.3 ± 5.4 vs 57.2 ± 5.2 (p = .756)	52.1 vs 48.1 (p = .825)	100 vs 100	NR	NR	Noncardiac surgery	Atorvastatin 20 mg, 37 days	96	104
Berwanger [11]	Brazil	66.8 ± 13.0 vs 66.7 ± 12.6	35.1 vs 40.9	4.6 vs 4.1	49.4 vs 48.1	88.4 vs 87.8	General, intraperitoneal, orthopaedic, gynaecologic, vascular, urologic, neurosurgery, breast (mastectomy), thoracic surgery	Atorvastatin 80 mg or 40 mg, 7 days	328	320
Parepa [16]	Romania	63.8 vs 64.7	59.1 vs 59.8	NR	1.30 vs 1.01	NR	Noncardiac, nonvascular surgery	Rosuvastatin 10 mg/day, 30 days	691	689

CAD: coronary artery disease; CEA: carotid endarterectomy; NR: not reported.

Patients in the statin and control arms did not exhibit significant differences in all-cause mortality (OR = 0.729, 95% CI = 0.464–1.147, p = .172). Significant heterogeneity was not noted between the trials (I² = 0%; p = .739), as shown in Figure 3 and Table 2, and we did not observe evidence of publication bias, as indicated by the Begg funnel plot (p = .462). Sensitivity analysis was performed using the leave-one-out approach. The pooled ORs were not significantly affected by the omission of individual trials from the analysis.

Figure 3. ORs for the associations between all-cause mortality and perioperative statin and placebo administration in patients undergoing noncardiac surgery. The sizes of the data markers are proportional to the weights of the individual studies.

Table 2. Meta-analyses of the effects of perioperative statins on improving patient outcomes after noncardiac surgery.

		Heterogeneity test			Publication bias		OR and 95% CI
Outcome	N	pH	I^2 (%)	Z test pZ	pB	pE	Fixed model	Random model
All-cause mortality	7	.739	0	.172	.462	.148	0.729 (0.464–1.147)	0.737 (0.466–1.166)
Cardiovascular mortality	10	.376	6.3	.346	1.000	.643	0.749 (0.410–1.367)	0.747 (0.380–1.467)
Myocardial infarction	9	.824	0	0	.266	.581	0.460 (0.324–0.653)	0.467 (0.328–0.667)
Stroke or TIA	7	.523	0	.265	.221	.778	0.584 (0.226–1.505)	0.476 (0.153–1.477)
MACE	7	.272	21.5	0	.707	.348	0.617 (0.476–0.801)	0.595 (0.427–0.828)
New cases of atrial fibrillation	4	.587	0	0	.308	.010	0.406 (0.247–0.666)	0.401 (0.242–0.666)

95% CI: 95% confidence interval; MACE: composite of death/myocardial infarction/stroke; N: number; OR: odds ratio; p_B: p Value of Begg’s test; p_E: p Value of Egger’s test; p_H: p Value of the heterogeneity test; p_Z: p Value of Z test.

During the follow-up period, 48 of 1514 patients (3.2%) in the perioperative statin group developed MI; the incidence of MI in the statin group was significantly lower than that (6.9%) in the control group (OR = 0.460, 95% CI = 0.324–0.653, p = 0), as shown in Figure 4. Incidences of postoperative MACE and new cases of atrial fibrillation were both lower in patients treated with statins than in control group patients, as shown by the fixed-effects model (OR = 0.617, 95% CI = 0.476–0.801, p = 0 for MACE; OR = 0.406, 95% CI = 0.247–0.666, p = 0 for new cases of atrial fibrillation). No significant differences in the incidences of stroke or TIA and cardiovascular mortality were observed between the statin and control arms (OR = 0.584, 95% CI = 0.226–1.505, p = .265 for stroke or TIA; OR = 0.749, 95% CI = 0.410–1.367, p = .346 for cardiovascular mortality).

Figure 4. ORs for the associations between MI and perioperative statin and placebo administration in patients undergoing noncardiac surgery. The sizes of the data markers are proportional to the weights of the individual studies.

Five of the 12 studies focused on noncardiovascular surgery, and another five studies assessed noncardiac vascular surgery. The other two studies investigated noncardiac surgery. We performed subgroup analyses based on the type of surgery. In the population that underwent noncardiovascular surgery, the incidences of MI, MACE and new cases of atrial fibrillation were all lower in patients treated with statins than in patients in the control group, as shown by the fixed-effects model (OR = 0.398, 95% CI = 0.248–0.638, p = 0 for MI; OR = 0.484, 95% CI = 0.340–0.688, p = 0 for MACE; OR = 0.370, 95% CI = 0.220–0.622, p = 0 for new cases of atrial fibrillation), as shown in Table 3. In the population that underwent noncardiac vascular surgery, a significant difference in the incidence of MI was observed between the statin and control arms (OR = 0.446, 95% CI = 0.220–0.901, p = .024). However, in the population that underwent noncardiac surgery, no significant differences in the incidences of MI, MACE and new cases of atrial fibrillation were observed between the statin and control arms.

Table 3. Subgroup analyses of the effects of perioperative statins on improving patient outcomes after noncardiac surgery.

			OR and 95% CI
Subgroup	Outcome	N	Fixed model	Random model	Z test pZ
Noncardiovascular surgery	All-cause mortality	2	0.670 (0.270–1.663)	0.682 (0.272–1.710)	.388
	Cardiovascular mortality	4	0.415 (0.146–1.182)	0.427 (0.148–1.237)	.100
	Myocardial infarction	4	0.398 (0.248–0.638)	0.405 (0.251–0.651)	0
	Stroke or TIA	1	0.331 (0.034–3.201)	0.331 (0.034–3.201)	.340
	MACE	2	0.484 (0.340–0.688)	0.488 (0.339–0.701)	0
	New cases of atrial fibrillation	2	0.370 (0.220–0.622)	0.371 (0.221–0.624)	0
Noncardiac vascular surgery	All-cause mortality	4	0.564 (0.273–1.165)	0.566 (0.273–1.173)	.122
	Cardiovascular mortality	4	0.546 (0.181–1.648)	0.548 (0.180–1.670)	.283
	Myocardial infarction	4	0.446 (0.220–0.901)	0.444 (0.218–0.908)	.024
	Stroke or TIA	5	0.284 (0.067–1.198)	0.290 (0.068–1.246)	.086
	MACE	2	0.474 (0.028–8.011)	0.474 (0.028–8.011)	.605
	New cases of atrial fibrillation	2	1.796 (0.189–17.073)	1.784 (0.187–17.046)	.610
Noncardiac surgery	All-cause mortality	1	1.046 (0.484–2.262)	1.046 (0.484–2.262)	.909
	Cardiovascular mortality	2	2.215 (0.675–7.266)	2.215 (0.675–7.266)	.190
	Myocardial infarction	1	0.753 (0.337–1.685)	0.753 (0.337–1.685)	.491
	Stroke or TIA	1	6.849 (0.352–133.123)	6.849 (0.352–133.123)	.204
	MACE	2	0.857 (0.576–1.274)	0.857 (0.576–1.274)	.445

95% CI: 95% confidence interval; MACE: composite of death/myocardial infarction/stroke; N: number; OR: odds ratio; p_Z: p Value of the Z test; TIA: transient ischemic attack.

Discussion

The present meta-analysis was designed to assess the effects of perioperative statins on patient outcomes after noncardiac surgery. Patients who received statin therapy had a significantly lower risk of developing postoperative MI, MACE and new cases of atrial fibrillation than did controls.

Statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) are used to treat hyperlipidaemia and exert pleiotropic effects [22]. Statins have been shown to exert anti-arrhythmic effects by modulating systemic inflammatory marker expression [23,24], attenuate fibrosis after MI by alleviating the inflammatory response [25], induce the vasodilation of coronary microvessels and exert direct antithrombotic effects [26,27]. Many studies have assessed the effects of perioperative statins on outcomes after cardiac surgery. Statin use may prevent postoperative atrial fibrillation, particularly in patients undergoing coronary artery bypass graft (CABG) surgery [28–30]. Curtis et al. performed a retrospective study and showed that preoperative statin administration for 24 hours or less reduced the rate of 30-day all-cause mortality after CABG [31].

More than 300 million noncardiac surgeries are performed annually. Perioperative myocardial injury (PMI) has been identified as an important complication of noncardiac surgery [32]. Strategies for improving the treatments and outcomes of PMI may provide major medical benefits [32]. A previous meta-analysis investigated whether perioperative statin therapy improves patient outcomes during and after noncardiac vascular surgery [33]. Five eligible studies were included in the previous meta-analysis. However, this study did not obtain sufficient evidence to conclude that statin use improved patient outcomes after surgery. Another meta-analysis by Antoniou et al. was published in 2015 [34]. This analysis aimed to investigate the role of perioperative statin therapy in noncardiac vascular and endovascular surgery. Four RCTs and 20 observational cohort or case–control studies were included in the analysis; the results suggested that statin therapy improves operative and interventional outcomes and lowers the risks of all-cause mortality, MI and MACE. As the conclusions of these studies were inconsistent, we collected nearly all relevant, published RCTs and conducted a meta-analysis to examine the effects of statin use after noncardiac surgery. In 2017, Bass et al. assessed the effect of atorvastatin on perioperative myocardial injury and inflammation after orthopaedic surgery. Twenty-two patients (11 in the perioperative statin group and 11 in the control group) were included, none of whom experienced MI, stroke, transient ischemic attack, atrial fibrillation or death. Nonetheless, the study was underpowered to reveal the effects of perioperative atorvastatin on patient outcomes after orthopaedic surgery [35], and it was therefore not included in our analysis. Ultimately, our analysis included 12 studies, five of which focused on populations that underwent noncardiovascular surgery, five of which assessed populations that underwent noncardiac vascular surgery, and two of which investigated populations that underwent noncardiac surgery. Our analysis involved a greater number of participants than previous analyses; thus, its statistical power was increased. Based on our results, statin therapy has significant beneficial effects on patient outcomes after noncardiac surgery.

Our analysis had some limitations. First, although 12 studies were included in our meta-analysis, the total number of participants was still small. Second, various types of surgery, such as vascular, urological and dental surgery, were performed in all studies. We performed subgroup analyses based on the type of surgery, but we were unable to draw a precise conclusion about the impact of the type of surgery on postoperative outcomes because of the limited number of included studies. Another limitation was that varying doses of two different intervention drugs (atorvastatin and rosuvastatin) were employed in the studies included in our meta-analysis. Because the number of included participants was small, we could not perform subgroup analyses to assess the effects of different types and doses of statins on outcomes.

In conclusion, the current limited evidence supports the hypothesis that perioperative statins effectively reduce the rates of MI, MACE and new cases of atrial fibrillation in patients undergoing noncardiac surgery. Future studies should focus on the optimal statin type and dose and other clinical end-points, such as lengths of stay and infection complications.

Disclosure statement

No potential conflict of interest was reported by the authors.