FormalPara Key Points

This retrospective observational study investigated the association between protein convertase subtilisin/kexin type 9 (PCSK9) inhibitors and infection-related adverse events using the World Health Organization’s (WHO) VigiBase. In the patient cohort receiving PCSK9 inhibitors, 82% were reported to have been administered evolocumab.

In addition to the potential respiratory infections listed on the label, a broad range of infection safety profiles associated with PCSK9 inhibitors, including gastrointestinal, renal, skin, and respiratory infections, were identified. The most significant adverse drug reactions occurred within approximately 5 months of treatment.

Given that most patients prescribed PCSK9 inhibitors are older adults, monitoring for infectious disorders should be performed during the early stages of treatment in this vulnerable population.

1 Introduction

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality worldwide [1,2,3,4]. According to the recent guidelines of high-profile cardiovascular societies, maintaining optimal low-density lipoprotein cholesterol (LDL-C) levels can help prevent the occurrence of CVD [5,6,7,8,9]. Previous studies have demonstrated the efficacy of lipid-lowering therapy (LLT) in reducing the risk of future cardiovascular events [10,11,12,13,14,15]. Recently, the newest class of LLT, i.e., protein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, demonstrated a remarkable reduction in LDL-C levels among patients treated with statins that are tolerant to treatment [16, 17]. This is consistent with trials where PCSK9 inhibitors were associated with a significant reduction in LDL‐C levels (up to 62%) and a 48% decrease in cardiovascular outcomes [18,19,20,21].

Despite these clinical benefits, the association between PCSK9 inhibitors and infectious complications remains controversial. In the ODYSSEY trial, the PCSK9 inhibitor-treated group showed a higher risk of infection-related adverse events (AEs) than that of the placebo group, including respiratory tract diseases such as influenza and nasopharyngitis [22,23,24,25,26]. In the FOURIER trial, patients treated with evolocumab exhibited a higher risk of nasopharyngitis than those in the placebo group (7.8% vs 7.4%) [27]. Several biological mechanisms indicate that PCSK9 inhibitors may cause bacterial infection through the excessive reduction of plasma cholesterol [28]. Moreover, PCSK9 inhibitors likely induce viral entry by enhancing the expression of LDL receptor (LDL-R) and CD81 levels to facilitate viral propagation [29]. Although it is biologically plausible that PCSK9 inhibitors may be linked to infections, it is imperative to assess the association between PCSK9 inhibitors and infection. A recent systematic review and meta-analysis study demonstrated that there is no detrimental association between PCSK9 inhibitors and the risk of severe infection, including sepsis [30]. However, no studies have explored the relationship between PCSK9 inhibitors and comprehensive infection-related AEs utilizing global databases, and epidemiological evidence is still lacking [31, 32].

With the increasing application of PCSK9 inhibitors in clinical settings, particularly for older adults, it is important to elucidate their infection-related safety profiles. Therefore, we performed an observational pharmacovigilance study to investigate the association between PCSK9 inhibitors and infection-related AEs using the World Health Organization’s (WHO) VigiBase.

2 Methods

2.1 Data Source

VigiBase is a global database maintained by the WHO that records all potential AEs associated with medicinal products. It is the largest unique database worldwide, with more than 30 million individual case safety reports (ICSRs) on suspected drug side effects from more than 130 countries since January 1967. The database contains information on patient demographic characteristics (age, sex, and region), drug usage (suspected and concomitant drugs), and reported effects (date of occurrence and seriousness) from both regulatory and voluntary sources. It is composed of spontaneous reports from multiple groups, including healthcare professionals, patients, and pharmaceutical companies in the post-marketing setting. The seriousness of the safety reports in VigiBase is classified by reporters according to the seriousness criteria of the WHO–Uppsala Monitoring Centre (WHO-UMC). The ICSR database in VigiBase is linked to medical classifications such as the Medical Dictionary for Regulatory Activities (MedDRA) for AEs. These VigiBase classifications enable structured data entry, retrieval, and analysis with different levels of precision. The ICSR reports in VigiBase are primarily sourced from the VigiFlow database, which aggregates data from various stakeholders such as healthcare professionals, the pharmaceutical industry, and patients across WHO member states. VigiFlow serves as the national repository for ICSRs and is then utilized in the generation process of the VigiBase database. Unlike VigiFlow and VigiLyze, which are part of the WHO's Pharmacovigilance tools, VigiBase is a raw database that focuses more on clinical information and adverse event data [33]. We used adverse drug reactions (ADRs) coded using MedDRA terms (version 24.1 at the time of the search).

2.2 Study Design

We conducted an observational, retrospective pharmacovigilance study of patients worldwide who were administered PCSK9 inhibitors, including alirocumab and evolocumab, using VigiBase. We utilized all available Vigibase sources from 1968 to 2022 for pharmacovigilance studies, given that the signals are detected based on existing information. To identify safety reports related to PCSK9 inhibitors, we excluded all ICSRs medication information missing and suspected duplicates. Additionally, reports identified as suspected or interacting drugs were excluded, along with safety reports without records of potential AEs in VigiBase (Fig. 1). The study drug was defined by obtaining a unique drug-chemical code for the drug of interest using the terms ‘alirocumab,’ ‘Praluent,’ ‘evolocumab,’ and ‘Repatha.’ Administrative information (reporter type and country), demographic data (age and sex), drug use (time to infectious AE onset, drug indication, and seriousness of infectious AE), and AE datasets (infection-related and co-reported AEs with infectious ADRs in each safety report) were collected.

Fig. 1
figure 1

Flowchart for selection of reports on PCSK9 inhibitors. PCSK9 protein convertase subtilisin/kexin type 9, WHO World Health Organization

Full size image

2.3 Outcome

This study included all infection-related AEs classified by group queries according to the MedDRA. All ICSRs that reported alirocumab or evolocumab as a suspected drug and at least one infectious ADR, as defined by the preferred terms (PTs), were retrieved. Given the PTs for various infection-related AEs, we categorized the infectious diseases based on the underlying primary system organ class (SOC) ‘infections and infestations’ using the MedDRA classification (version 24.1). To distinguish between the organs affected by infection-related diseases, we used secondary SOC (Supplementary Table 1, see electronic supplementary material [ESM]).

2.4 Statistical Analysis

Descriptive analyses were conducted for patients receiving PCSK9 inhibitors, and the proportions were calculated using categorical variables. Disproportional statistical analysis was used to evaluate drug AEs in the pharmacovigilance studies.

The association of infection-related AEs with PCSK9 inhibitors compared with other drugs was assessed using a disproportionality method specific to case–non-case studies. The disproportionality signal indicates a valid connection between PCSK9 inhibitors and specific side effects if the proportion of AEs of interest is significantly higher for the drug of interest than for any other drug [34].

We used the reporting odds ratio (ROR) and information component (IC) to report the findings of the disproportional analysis. The ROR is a frequentist measure of association derived from the contingency table of the drug and the number of AEs. The higher the ROR, the more statistically relevant the pharmacovigilance AE signal. Because the ROR estimate has been demonstrated to be the most definitive in at least three cases of AEs of interest [35], we limited the number of reported AEs to at least three. Thus, the ROR and 95% confidence interval (CI) were used in this study, and a lower bound of 95% CI ≥1 was considered a signal. IC is a Bayesian neural network-based disproportionality method that is calculated using the observed and expected numbers of reports for a drug–ADR combination. A positive IC value (≥0) indicates a higher number of reports than expected.

In addition to the infection-related AEs associated with PCSK9 inhibitors, we identified co-reported AEs. In this study, we performed a descriptive analysis of AEs other than infection-related AEs within the same report. The frequencies of the top 10 SOCs co-reported with infection-related AEs are demonstrated [36].

Time-to-onset (TTO) is the time interval between the date of PCSK9 inhibitor administration and the start of the reaction. The time to infection-related AE onset was determined by eliminating the TTO values <0 or the number of infection-related AEs <5.

Finally, to assess whether the higher risk of infection-related AEs can be attributed to the comparator, we conducted a sensitivity analysis by substituting the comparator with statins (atorvastatin, simvastatin, pravastatin, rosuvastatin, and fluvastatin). All statistical analyses and VigiBase data management were performed using the SAS package (SAS Institute, Cary, NC, USA), version 9.4.

3 Results

3.1 Characteristics of Adverse Events (AEs)

Among the 31,923,227 ICSRs, 114,293 were likely associated with PCSK9 inhibitors, including 20,603 (18% of associated PCSK9 inhibitors) reports on alirocumab and 93,690 (82% of associated PCSK9 inhibitors) reports on evolocumab. Among patients receiving PCSK9 inhibitors, 41.3% were ≥65 years old and 53.7% were women. Most ICSRs for the study drugs reported in VigiBase were from the United States (87.3%). The number of reports on the usage of PCSK9 inhibitors increased between 2019 and 2020 (40.1%). Healthcare professionals reported 70,180 cases (61.4%). The number of safety reports for these two drugs increased continuously between 2015 and 2022 (Table 1).

Table 1 Baseline demographic characteristics of the study population
Full size table

3.2 Association Between PCSK9 Inhibitors and Infection-Related AEs

Among 114,293 reports (258,099 drug–AE pairs) on PCSK9 inhibitors, 8464 reports (31,951 drug–AE pairs) were identified as infection-related AEs. Disproportionality analysis of the VigiBase dataset of safety reports associated with PCSK9 inhibitors highlighted 19 significant signals based on both the ROR and IC. The ROR of infectious AEs in the skin and respiratory, gastrointestinal, and renal systems was determined for alirocumab and evolocumab. The significant infection-related AEs were bronchitis (ROR 2.49, 95% CI 2.27–2.72), coronavirus infection (ROR 1.59, 95% CI 1.16–2.18), influenza (ROR 2.89, 95% CI 2.74–3.05), nasopharyngitis (ROR 4.47, 95% CI 4.30–4.64), sinusitis (ROR 2.91, 95% CI 2.70–3.14), diverticulitis (ROR 1.87, 95% CI 1.56–2.24), gastric infection (ROR 2.47, 95% CI 1.63–3.75), kidney infection (ROR 1.36, 95% CI 1.06–1.73), urinary tract infection (ROR 1.42, 95% CI 1.31–1.54), and dermatitis infected (ROR 2.94, 95% CI 1.23–7.04) (Fig. 2).

Fig. 2
figure 2

Results of signal detection for infection-related AEs of PCSK9 inhibitors. AE adverse event, CI confidence interval, IC information component, PCSK9 protein convertase subtilisin/kexin type 9, ROR reporting odds ratio

Full size image

3.3 Co-Reported AEs with Infection-Related AEs

Concurrent reports of infection-related AEs were classified as SOCs. The top 10 overlapping datasets based on each patient report after PCSK9 inhibitor administration are shown in Fig. 3. Owing to the numerous combinations of infectious AEs and other concurrent classes, only the 38 most frequent overlaps were included. The most common overlapping SOCs were ‘musculoskeletal and connective tissue disorders’ (31.1%); ‘respiratory, thoracic, and mediastinal disorders’ (29.4%); and ‘gastrointestinal disorders’ (18.2%). Additionally, 4109 infection-related AEs were exclusively reported. The most reported SOC combination was ‘respiratory, thoracic, and mediastinal disorders’ (697 cases) with infection-related AEs. The second most reported overlapping combination was ‘musculoskeletal and connective tissue disorders’ (599 cases).

Fig. 3
figure 3

Co-reported AEs with infection-related AEs. ADR adverse drug reaction, AE adverse event. *Owing to the numerous combinations of infectious AEs and other concurrent classes, we present only the 38 most frequent overlaps

Full size image

3.4 TTO Analysis

Among the significant infection-related AE signals, only a portion of ICSRs with respiratory tract infections were available for TTO analysis because of the small number of reported cases. Most initial respiratory tract infection AEs (70.0%) developed within 5 months of drug administration (Fig. 4). The median and interquartile range (IQR) TTO of initial infection-related AEs for respiratory tract diseases, including only significant signals in the main results, were 72 (23.5–202.5), 76 (13–140), and 104 (22–480.5) days for nasopharyngitis, rhinitis, and influenza, respectively. The median and IQR TTO for urinary tract infections were 34.5 (7.8–121.3) days.

Fig. 4
figure 4

TTO of infection-related AEs of PCSK9 inhibitors. AE adverse event, ICSR individual case safety report, IQR interquartile range, PCSK9 protein convertase subtilisin/kexin type 9, TTO time-to-onset. *Among the significant infection-related AE signals, only some ICSRs with respiratory tract infections were available using TTO with more than five reported cases

Full size image

3.5 Sensitivity Analysis

3.5.1 Association Between Alirocumab and Infection-Related AEs

The number of infection-related reports associated with alirocumab was 1502. Disproportionality analysis was conducted on the VigiBase dataset of safety reports related to alirocumab, a PCSK9 inhibitor, focusing specifically on signals that were significantly prominent within the main result (Fig. 2). Utilizing both the ROR and IC, a total of 12 significant signals were identified within the existing set of 19 signals. Significant AEs that were excluded in comparison to the main results encompassed gastric infection, gastroenteritis viral, pharyngitis streptococcal, prostate infection, respiratory tract infection, and upper respiratory tract infection. Notably, no occurrences of ear, nose and throat infection were identified in the reported data. The significant infection-related AEs were bronchitis (ROR 3.28, 95% CI 2.76–3.90), coronavirus infection (ROR 2.63, 95% CI 1.53–4.53), influenza (ROR 3.07, 95% CI 2.73–3.44), nasopharyngitis (ROR 3.61, 95% CI 3.29–3.96), sinusitis (ROR 3.33, 95% CI 2.85–3.89), diverticulitis (ROR 2.43, 95% CI 1.72–3.43), kidney infection (ROR 1.74, 95% CI 1.08–2.80), urinary tract infection (ROR 1.96, 95% CI 1.68–2.28), and dermatitis infected (ROR 5.68, 95% CI 1.43–22.65). Furthermore, upon comparing the risk ratios for each AE associated with alirocumab in the ODYSSEY trial, a higher risk was observed compared with our analysis (Supplementary Table 2, see ESM).

3.5.2 Association Between Evolocumab and Infection-Related AEs

The number of infection-related reports associated with evolocumab was 5619. Disproportionality analysis was conducted on the VigiBase safety reports related to evolocumab focusing specifically on signals that were significant within the main result (Fig. 2). Utilizing both the ROR and IC, a total of 16 significant signals were identified within the existing set of 19 signals. Significant signals that were excluded in comparison to the main results included coronavirus infection, dermatitis infected, diverticulitis intestinal haemorrhagic, kidney infection, pharyngitis, and prostate infection. The significant infection-related AEs were bronchitis (ROR 2.27, 95% CI 2.04–2.53), influenza (ROR 2.83, 95% CI 2.67–3.01), nasopharyngitis (ROR 4.67, 95% CI 4.48–4.87), sinusitis (ROR 2.79, 95% CI 2.56–3.04), diverticulitis (ROR 1.72, 95% CI 1.39–2.12), gastric infection (ROR 2.69, 95% CI 1.72–4.21), and urinary tract infection (ROR 1.28, 95% CI 1.17–1.41). Moreover, our main results showed a higher risk of infection-related adverse events compared with alirocumab in the FOURIER trial (Supplementary Table 3, see ESM).

3.5.3 Sensitivity Analysis for the Risk of Infection-Related AEs with Statins

The results of the sensitivity analysis are consistent with the main results. Compared with statins, PCSK9 inhibitors exhibited a higher risk of ADRs belonging to the ‘infections and infestations’ SOC in ICSRs. The significant infection-related AEs were bronchitis (ROR 4.13, 95% CI 3.95–4.31), coronavirus infection (ROR 6.59, 95% CI 5.81–7.47), nasopharyngitis (ROR 7.81, 95% CI 7.70–7.92), diverticulitis (ROR 2.03. 95% CI 1.82–2.27), gastric infection (ROR 5.34, 95% CI 4.46–6.39), and urinary tract infection (ROR 2.46, 95% CI 2.35–2.58) (Supplementary Table 4, see ESM).

4 Discussion

We observed a broad range of infection safety profiles for alirocumab and evolocumab against gastrointestinal, renal, skin, and respiratory infections. This study found that musculoskeletal, respiratory, and gastrointestinal disorders were the most frequently reported SOCs associated with infection-related AEs. The most significant ADR signals occurred within approximately 5 months of treatment initiation. The main findings of this study demonstrated consistency across sensitivity analyses using different comparators. Given that most patients in these reports were older adults, our findings suggest that monitoring for infectious disorders should be performed during the early stages of PCSK9 inhibitor treatment in this population.

Randomized controlled trials (RCTs) assessing the clinical effects of PCSK9 inhibitors have shown a detrimental safety profile for infectious AEs [22,23,24,25,26,27]. Our findings are consistent with those of previous studies, implying that the risk of upper respiratory tract infection increases with PCSK9 inhibitors compared with other drugs. According to the labels of PCSK9 inhibitors, the most commonly reported AEs (>3% of patients) were upper respiratory tract infections, primarily nasopharyngitis, and influenza [37, 38]. In particular, the FOURIER and ODYSSEY studies, which investigated the infection-related incidence of PCSK9 inhibitors, indicated an elevated risk of nasopharyngitis in patients treated with PCSK9 inhibitors. The FOURIER-OLE study evaluated the effects of evolocumab in patients with established atherosclerotic cardiovascular disease and reported unexpected infection-related AEs. The incidence of influenza was slightly higher in patients treated with evolocumab during the integrated interim extension studies year 1 SoC (standard-of-care)-controlled period (45 cases/1489 patients, 3.0%) than in those who received the SoC (108 cases/2976 patients, 3.6%) [39]. Similarly, in the ODYSSEY FH II study, which evaluated long-term (78 weeks) efficacy and safety with hypercholesterolemia and high CVD risk, higher rates of influenza as infection-related AEs were more frequently reported in patients treated with alirocumab (24 cases/167 patients, 14.4%) than in those who received the placebo (7 cases/81 patients, 8.6%) [40]. Moreover, a study by Gürgöze et al. indicated that upper respiratory tract infections, especially nasopharyngitis, which is the most commonly reported AE in both the VigiLyze and Lareb databases, were reported in 16.2% of patients receiving PCSK9 inhibitors in the Erasmus Medical Centre hospital registry [41]. Hence, our findings were consistent with those of several studies and labels approved by both the US Food and Drug Administration (FDA) and European Medicines Agency (EMA), which reflect the commonly reported AEs based on RCT analyses. We observed additional outcomes, including higher rates of gastrointestinal, kidney, and skin infections, following the administration of PCSK9 inhibitors compared with other drugs. However, in the study conducted by Gürgöze et al., most reported infectious AEs resolved during the follow-up period [41]. We found that most respiratory tract infection-related AEs were reported within an average of 5 months. Additionally, Zhou et al. conducted a meta-analysis that included 20 RCTs, specifically examining severe infections (sepsis) and the association between sepsis and the use of PCSK9 inhibitors. This analysis demonstrated that PCSK9 inhibitor use was not linked to severe infection-related AEs, including sepsis, when compared with placebo (sepsis: RR 0.85, 95% CI 0.67–1.07; any severe infection: RR 0.96, 95% CI 0.89–1.03), and younger people are less likely to experience such events [30]. The majority of patients included in our study analysis who were using PCSK9i were elderly. Therefore, there is a critical need for additional long-term safety assessments of PCSK9i in the real-world setting, utilizing more comprehensive data. Therefore, our findings reveal that infection-related AEs associated with PCSK9 inhibitors may play an important role in the short-term prognosis of patients, especially older adults, and close monitoring is warranted to identify infectious conditions.

Current evidence suggests that AEs associated with infectious diseases are mechanically plausible. PCSK9 inhibitors are strictly linked to the metabolism of lipid and cholesterol via LDL-R regulation [42]. LDL-Rs are involved in cholesterol efflux capacity and macrophage functioning, which decreases cholesterol uptake [43,44,45]. Overall, decreased activity of PCSK9 inhibitors during infections, especially sepsis, may exhibit a harmful effect on the pathway of cholesterol metabolism and stimulate the immune system by possibly deteriorating lipid raft composition [28]. Specifically, the virus enters hepatic cells via LDL-R, which also plays a critical role in viral replication [46, 47]. Syed et al. observed that individuals infected with the virus had higher LDL-R levels in their hepatocytes compared with the control group [48]. Furthermore, the results of an in vitro study suggested that PCSK9 inhibitors may stimulate the upregulation of LDL-R with the participation of CD81 particles, enabling viral entry into the cells [29, 49]. Thus, the above evidence indicates that PCSK9 inhibitors may lead to increased susceptibility to bacterial and viral infection.

Our analysis identified nine SOCs associated with PCSK9 inhibitors. Our study exhibited high consistency in the rate of co-reported SOC incidence with previous studies of ADRs associated with PCSK9 inhibitors, including US FDA and EMA labels and clinical studies [23, 27, 37, 38, 50,51,52,53]. ‘Musculoskeletal and connective tissue disorders’ (31.1%), ‘respiratory, thoracic, and mediastinal disorders’ (29.4%), and ‘gastrointestinal disorders’ (8.2%) are some of the most common SOCs following PCSK9 inhibitor administration, according to both our study and the labels. This was consistent with the ranking of the most frequently reported combinations of SOCs and infection-related AEs (697, 599, and 292 cases, respectively). Given that coronavirus infection, identified in our study, can cause gastrointestinal symptoms such as diarrhea and nausea via pathophysiological mechanisms, the co-reported gastrointestinal ADRs may be additional AEs caused by infection-related AEs [54, 55].

To enhance the robustness of our findings, we conducted a sensitivity analysis by changing the comparator group from all other drugs to statins. Statins are commonly used as a first-line therapy to significantly reduce LDL-C levels, and concomitant administration of PCSK9 inhibitors is often considered in statin-intolerant cases [56]. We detected signals for infection-related AEs associated with PCSK9 inhibitors compared with statins, which was consistent with our primary findings that showed a higher risk of organ infections with PCSK9 inhibitors compared with other drugs. Furthermore, the sensitivity analysis was conducted to mitigate confounding by the indication that may arise from the analysis of PCSK9i versus all other drugs. The results of this analysis demonstrate a concerning association of infection-related AEs with both alirocumab and evolocumab. Therefore, consistent sensitivity analysis results with our main analysis indicate the robustness of our study findings.

5 Strengths and Limitations

To our knowledge, this is the first study to analyze the safety profile of PCSK9 inhibitors in terms of infectious AEs worldwide. This study has several strengths; it provides a comprehensive evaluation of ICSRs associated with PCSK9 inhibitors using the WHO’s VigiBase. Considering the lack of safety data in real-world settings, our study reflected the broad characteristics of the source data, including vulnerable patients (usually older adults) who were excluded from clinical trials. This study identified infection-related AEs across a comprehensive range of PCSK9 inhibitors. Evidence obtained from spontaneous reports will help improve the safety profiles of PCSK9 inhibitors.

Despite its strengths, this study has several limitations. This study was affected by the inherent limitations of spontaneous reporting databases, which cannot be used to deduce causal associations owing to underreporting and a potential lack of comparability. Therefore, reports are prone to selection bias owing to the reporter or reporting systems in different countries. Additionally, we could not determine whether the infections were solely attributable to the use of PCSK9 inhibitors. We also did not consider confounding factors such as health examinations and previous or current medical conditions. Further studies are required to confirm our findings and provide a more definitive assessment of the safety profile of PCSK9 inhibitors in terms of infectious AEs.

6 Conclusion

In addition to the labelled respiratory infections, we identified six infection-related symptoms affecting the gastrointestinal, urinary, and renal organs using the WHO’s VigiBase. Our findings align with those of a landmark RCTs of PCSK9 inhibitors, but we observed a broader spectrum of signals for infection-related AEs. Our study provides important insights into the safety profiles of PCSK9 inhibitors with respect to infection-related AEs and emphasizes the need for continued surveillance and monitoring. However, the clinical implications of our findings should be carefully considered because these signals were derived from spontaneous reporting systems. Further studies are required to establish a potential association between infection and PCSK9 inhibitor use.