Abstract

Background. Alirocumab and evolocumab, as protein convertase subtilisin kexin type 9 (PCSK9) inhibitors, have been reported to reduce cardiovascular risk. This meta-analysis is aimed at updating the safety data of PCSK9 inhibitors. Methods. We assessed the relative risk for all treatment-related adverse events, serious adverse events, diabetes-related adverse events, and neurocognitive and neurologic adverse events with PCSK9 inhibitors compared to controls (placebo or ezetimibe). In addition, we conducted a meta-analysis to quantitatively integrate and estimate the adverse event rates in long-term studies. Results. There were no significant differences between PCSK9 inhibitors and controls in the relative risk analysis. In a subgroup analysis of each PCSK9 inhibitor, alirocumab treatment significantly reduced the risk of serious adverse events compared to control treatment (; 95% confidence interval (CI), 0.896–0.980), but no significant difference was observed with evolocumab treatment (; 95% CI, 0.963–1.054). Moreover, alirocumab treatment afforded a significant reduction in the risk of diabetes-related adverse events compared to control treatment (; 95% CI, 0.845–0.987). The overall incidence (event rate) of long-term adverse events was 75.1% (95% CI, 71.2%–78.7%), and the incidence of serious long-term event rate was 16.2% (95% CI, 11.6%–22.3%). Conclusions. We suggest that alirocumab and evolocumab are generally safe and well tolerated and that their addition to background lipid-lowering therapy is not associated with an increased risk of adverse events or toxicity.

1. Introduction

Alirocumab and evolocumab are fully human monoclonal antibodies against the protein convertase subtilisin kexin type 9 (PCSK9) and modulate the upregulation of recycling and expression of low-density lipoprotein cholesterol (LDL-C) receptors at the cell surface, and increase LDL-C clearance from circulation [1]. Both PCSK9 inhibitors were approved by the Food and Drug Administration in 2015 and are indicated for patients with established cardiovascular disease to reduce the risk of myocardial infarction, stroke, and coronary revascularization [2]. In addition, they are used as an adjunct to diet, alone or in combination with other LDL-C-lowering therapies, in patients with primary hyperlipidemia.

In randomized controlled trials, alirocumab and evolocumab have been reported to reduce the risk of recurrent cardiovascular disease in patients following an acute coronary event and secondary prevention populations when added to background statin therapy [3, 4]. In terms of safety, PCSK9 inhibitors are well tolerated and favorable. However, injection-related adverse events, such as injection-site reactions and “flu-like” symptoms after injections, may be a limitation in some patients [1]. In addition, long-term follow-up data on the efficacy or safety of PCSK9 inhibitors are insufficient, and some issues regarding their potential impact on neurocognitive- or diabetes-related risk have not been clearly uncovered [5].

A previous meta-analysis including 25 randomized controlled trials found that alirocumab and evolocumab are generally safe. However, it was reported that alirocumab increased the rate of injection-site reactions, while evolocumab reduced the rate of abnormal liver function [6]. In systematic reviews that evaluated concerns related to diabetes mellitus, the PCSK9 inhibitors were not associated with the risk of new-onset diabetes and adverse events of diabetes mellitus [7, 8]. Similarly, there was no increased risk of neurocognitive adverse events [9]. Since then, more clinical studies of both PCSK9 inhibitors have been reported.

This meta-analysis was conducted to update the safety data for PCSK9 inhibitors to assess the relative risk of alirocumab and evolocumab compared with placebo (or ezetimibe) and to quantitatively integrate and estimate the incidence of adverse events in long-term studies.

2. Methods

2.1. Search Strategy and Study Selection

We searched for published articles reporting adverse events associated with alirocumab and evolocumab in MEDLINE (OVID and PubMed), EMBASE, the Cochrane Library, and http://ClinialTrials.gov. The search was completed on October 30, 2021. The following search terms were used: PCSK9 inhibitors, PCSK9 antibody, evolocumab, AMG 145, alirocumab, SAR236553, and REGN727. We reviewed the reference lists of the retrieved articles and searched the relevant reviews to identify additional eligible studies. There were no restrictions on any publication.

Two authors independently reviewed and selected studies for inclusion in the systematic review. The inclusion criteria were as follows: (1) phase 2, 3, or 4 clinical trials; (2) administration of alirocumab or evolocumab; and (3) safety or adverse drug events. Disagreement about the inclusion of an article in the evaluation was resolved through discussion. For a clinical trial described in multiple reports, we extracted data from the most complete account and used the other publications only for clarification.

The study protocol for this meta-analysis was registered in the International Prospective Register for Systematic Reviews (PROSPERO) CRD42022328637.

2.2. Data Extraction and Quality Assessment

Two authors independently reviewed detailed full-text articles. The data were extracted from each study: number and characteristics of participants, treatment administered (dose regimen and periods), and adverse events. The bias risk of the included studies was assessed by two authors using the Cochrane RoB 2 criteria: bias arising from the randomization process, bias due to deviations from intended interventions, bias due to missing outcome data, bias in the measurement of the outcome, and bias in the selection of the reported result [10]. Disagreements between the two authors were resolved by consensus after discussion.

2.3. Meta-Analysis and Statistical Analysis

To evaluate treatment safety, we compared the total number of adverse events and serious adverse events reported in participants treated with alirocumab or evolocumab vs. those treated with placebo or ezetimibe. Moreover, we assessed the total number of diabetes-related, neurocognitive, and neurologic adverse events reported in both treatment groups. Studies with a follow-up period of at least 48 weeks were included to estimate the incidence of long-term adverse events.

The test (employing Q statistics) and the calculating values were used to assess heterogeneity among including studies [11]. Based on the results of the heterogeneity test in each analysis, a fixed-effects model or a random-effects model was applied to the analysis [12, 13].

Publication bias was examined using Begg’s method and Egger’s regression test [14, 15]. Also, we performed sensitivity analyses by excluding the contribution of each study to the meta-analysis data in turn.

We performed all statistical analyses using the Comprehensive Meta-analysis Software version 2 (CMA 26526; Biostat, Englewood, NJ, USA). All values were two-sided, and values <0.05 was considered to indicate statistical significance.

3. Results

3.1. Study Characteristics and Risk of Bias Assessments

A total of 1,709 articles were identified in the literature search. The titles and abstracts of 743 articles were reviewed after excluding duplicates. Of these articles, 637 were excluded, and the full texts of 106 articles were assessed for meeting the eligibility criteria. A further 47 articles were excluded, and the data from the remaining 49 articles were finally included in the present meta-analysis (Figure 1). The general characteristics of included studies are shown in Table 1.

Risk of bias assessments for each study, including all domain judgments and support for judgment, are represented in the risk of bias section in Table 1. The risk of bias in outcomes across all studies was similar and predominately of ‘some concerns’ (Supplementary Table S1).

3.2. Meta-Analysis of All Adverse Events and Serious Adverse Events

Forty-seven studies were included to evaluate any treatment-related adverse events. A total of 35,358 participants treated with PCSK9 inhibitors (alirocumab or evolocumab) and 30,710 participants treated with controls (placebo or ezetimibe) were assessed. No significant differences were observed between the two treatments (; 95% confidence interval (CI), 0.992–1.055) (Table 2).

In the analysis of serious adverse events, 35,046 participants treated with PCSK9 inhibitors and 30,522 participants treated with controls from 44 studies were assessed. No significant differences were observed between the two treatments (; 95% CI, 0.944–1.003). In the subgroup analysis of each PCSK9 inhibitor, alirocumab treatment significantly reduced the risk of serious adverse events compared to the control treatment, but no significant difference was observed with evolocumab treatment (alirocumab: ; 95% CI, 0.896–0.980; evolocumab: ; 95% CI, 0.963–1.054) (Figure 2).

3.3. Meta-Analysis of Diabetes-Related Adverse Events

A total of 21 studies with 51,817 participants (27,770 treated with PCSK9 inhibitors and 24,047 treated with controls) were included. No significant difference was showed in the safety assessment of diabetes-related adverse events (; 95% CI, 0.914–1.023). In subgroup analysis of each PCSK9 inhibitor, alirocumab treatment afforded a significant reduction in the risk of diabetes-related adverse events compared to control treatment (; 95% CI, 0.845–0.987) (Figure 3).

3.4. Meta-Analysis of Neurocognitive and Neurologic Adverse Events

Nineteen studies, including 32,916 participants treated with PCSK9 inhibitors and 29,166 participants treated with controls, were assessed. There was no significant difference in the safety assessment of neurocognitive and neurological adverse events between the two treatments (; 95% CI, 0.913–1.163). There were no significant differences in the subgroup analysis of each PCSK9 inhibitor (Figure 4).

3.5. Incidence of Long-Term Adverse Events

A total of 13 studies were assessed for the long-term risk of all and serious adverse events in 20,969 participants treated with PCSK9 inhibitors. The overall incidence (event rate) of long-term adverse events was 75.1% (95% CI, 71.2%–78.7%), and the incidence of long-term serious event rate was 16.2% (95% CI, 11.6%–22.3%) using the random-effects model (Table 2).

Long-term risk of diabetes-related adverse events was assessed in 10 studies including 24,745 participants treated with PCSK9 inhibitors, and the incidence of diabetes-related adverse events was 4.50% (95% CI, 3.10%–6.50%), when applied the random-effects model (Table 2).

The long-term risk of neurocognitive and neurological adverse events was assessed in 12 studies, including 30,571 participants treated with PCSK9 inhibitors. The incidence of neurocognitive and neurologic adverse events was 1.70% (95% CI, 1.10%–2.70%), when applied the random-effects model (Table 2).

3.6. Publication Bias and Sensitivity Analyses

We evaluated the publication bias and the results of Begg’s and Egger’s tests are shown in Table 2. Sensitivity analysis was also performed by recalculating all findings after omitting the data from each study included in the meta-analysis. The results were not significantly altered throughout this process.

4. Discussion

We performed this meta-analysis to update the safety data for PCSK9 inhibitors to evaluate the relative risks of alirocumab and evolocumab compared to controls. In addition, we conducted a meta-analysis to quantitatively integrate and estimate the incidence of adverse events in long-term studies, which is a meaningful approach for the safety evaluation of PCSK9 inhibitors.

Based on the results of meta-analysis, we suggest that adding PCSK9 inhibitors to statins or other lipid-lowering therapies is not associated with an increased risk of adverse events or toxicity. That is, no significant differences were found in any of the comparisons analyzed, including serious adverse events, diabetes-related adverse events, or neurocognitive and neurological adverse events. Interestingly, alirocumab therapy seems to have a lower risk of diabetes and serious adverse events, which is consistent with a previous meta-analysis [7]. These results may be due to the unique characteristics of alirocumab or the effects of background lipid-lowering therapy.

In particular, diabetes mellitus is a cardiovascular risk and a significant adverse event of lipid-lowering therapies such as statins. Therefore, the use of PCSK9 inhibitors that do not increase the risk of diabetes is recommended. However, considering that most patients with dyslipidemia are treated with combination therapy, diabetes-related monitoring should not be excluded.

Previous studies have reported a higher incidence of neurocognitive events in patients receiving PCSK9 inhibitors than in those receiving standard therapy, but other clinical studies or systematic reviews did not show an increase in neurocognitive deficits in patients receiving these inhibitors [16, 17]. In addition, our meta-analysis showed results that were consistent with those described above. It is known that neither cholesterol nor PCSK9 can cross the blood-brain barrier under normal conditions, and alirocumab or evolocumab also cannot cross the blood-brain barrier [18, 19]. Therefore, we suggest that PCSK9 inhibitors do not cause or increase neurocognitive or neurological adverse events. However, cognitive problems in geriatric patients remain an important issue that requires close monitoring.

One meta-analysis reported that long-term treatment with alirocumab or evolocumab reduced LDL-C levels and improved cardiovascular outcomes while showing a similar safety profile to non-PCSK9 inhibitor therapy [20]. We performed the present meta-analysis to estimate the incidence of long-term adverse events. The overall incidence of long-term adverse events in PCSK9 inhibitor therapy is rather high at 75.1%, but it should be evaluated through (possibly indirect) comparison with the incidence of other comparative drugs. The incidence of diabetes-related adverse events, and neurocognitive and neurological adverse events was estimated to be approximately 4.50% and 1.7%, respectively.

In addition, a recent systematic review suggested that no major safety issues associated with PCSK9 inhibitors were observed, which is consistent with our results [21]. They also suggested that the use of PCSK9 inhibitors significantly reduced the risk of MI, ischemic stroke, and coronary revascularization in patients with dyslipidemia or atherosclerotic cardiovascular disease. These results, including our meta-analysis, are the evidences that support the role of PCSK9 inhibitors as treatments for dyslipidemia, and are expected to further increase their clinical use.

Our study had several limitations. First, we performed a meta-analysis based on previously reported articles which were not necessarily complete or accurate and the results may be partially different when applied to individual patients. Second, significant heterogeneity was present in the analyses, and dividing the studies into subgroups or performing a sensitivity analysis failed to identify the sources of heterogeneity. Despite these limitations, this meta-analysis is meaningful in that it provides clinical evidence for better pharmacotherapy in patients with dyslipidemia.

5. Conclusions

There were no significant differences between the PCSK9 inhibitors and controls, including serious adverse events, diabetes-related adverse events, or neurocognitive and neurological adverse events. PCKS9 inhibitors are relatively safe and well tolerated, and their addition to background lipid-lowering therapy is not associated with an increased risk of adverse events or toxicity.

Abbreviations

CI:Confidence interval
LDL-C:Low-density lipoprotein cholesterol
PCSK9:Protein convertase subtilisin kexin type 9
RR:Risk ratio.

Data Availability

The data supporting this meta-analysis are from previously reported studies and datasets, which have been cited. The processed data are available from the corresponding author upon request.

Disclosure

The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflicts of Interest

The authors declare that there is no conflict of interests.

Authors’ Contributions

Choi HD and Kim JH conceived the study, conducted the search, and collected the data. Choi HD performed the analysis. Choi HD and Kim JH wrote and reviewed the manuscript. All authors read and approved the final manuscript.

Acknowledgments

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government, Ministry of Science, and ICT (Grant No. 2022R1F1A1074583). A PRISMA2020 checklist for this meta-analysis was provided in Supplementary Table S2.

Supplementary Materials

Supplementary 1. Table S1: Risk of bias.

Supplementary 2. Table S2: PRISMA 2020 checklist.