Postmarketing adverse events of efgartigimod alfa: a real-world pharmacovigilance study based on the FAERS database

Published 19 hours ago• 19 minute read

Efgartigimod alfa is an important novel drug for the treatment of myasthenia gravis. However, postmarketing safety data for this drug is limited, underscoring the need for comprehensive safety evaluations in real-world populations.

This study aims to identify adverse event (AE) signals associated with efgartigimod alfa using the Food and Drug Administration Adverse Event Reporting System (FAERS) database, with a focus on evaluating unexpected AEs not previously observed in clinical trials. AE reports with efgartigimod alfa as the primary suspect from the first quarter of 2022 to the fourth quarter of 2023 were extracted.

Signal strength was assessed using Reporting Odds Ratio, Proportional Reporting Ratio, Empirical Bayes Geometric Mean and Bayesian Confidence Propagation Neural Network methods at the Preferred Term level.

A total of 1403 valid cases were retrieved. Urinary tract infection was the most reported AE, while procedural headache demonstrated the strongest signal across all four algorithms. Sepsis, atrial fibrillation and transient ischaemic attack were significant unexpected AEs. The median onset time for AEs was 57.00 days, with cumulative incidence of AEs reaching 37.31% at 30 days and 64.25% at 100 days post-treatment initiation.

Our analysis of real-world data from the FAERS database revealed that most significant AEs were consistent with clinical trials, but some unexpected AEs were additionally identified, warranting further investigation.

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request. The original data source is publicly available from the US Food and Drug Administration (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html).

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

When the immune system fails to distinguish between self-antigens and non-self-antigens, it may attack normal components of the body, resulting in tissue damage. Myasthenia gravis (MG) is a rare, chronic autoimmune disorder characterised by muscle weakness and fatigue, primarily due to impaired transmission at key neuromuscular junctions, especially after prolonged physical activity.1 Generalised MG can affect ocular movement, swallowing, speech, mobility and respiratory function, leading to generalised debilitation and potentially life-threatening complications.2 Traditional therapies for MG include symptomatic treatment (eg, acetylcholinesterase inhibitors), short-term rescue immunotherapy (eg, plasma exchange (PLEX) and intravenous immunoglobulins (IVIgs)), long-term immunosuppressive therapy (eg, corticosteroids and non-steroidal immunosuppressants) and thymectomy for patients with MG with acetylcholine receptor antibody (AChR-Ab+).3 The therapeutic window for broad-spectrum non-specific immunosuppressants is narrow, and these treatments may take weeks to months to show efficacy. They are also associated with a range of adverse effects, such as weight gain, hypertension, osteoporosis, gastrointestinal symptoms, bradycardia and renal impairment.4 Many patients fail to achieve complete or stable remission, with 10%–20% of patients being refractory to or intolerant of these medications.5 Other treatments like PLEX, IVIg and immunoadsorption can rapidly reduce autoantibody titres and show satisfactory results in various IgG-driven autoimmune diseases. However, they also present adverse effects and face challenges, such as high production costs, labour-intensive procedures and dependence on voluntary plasma donations.6

Compared with traditional immunosuppressants, newer targeted therapies offer a more rapid onset of action, improved tolerability and better alignment with the therapeutic needs of individual patients. One approach is to induce the clearance of pathogenic IgG autoantibodies by inhibiting the neonatal Fc receptor (FcRn), which is considered an effective strategy and a non-traditional alternative therapy for certain autoimmune diseases, including MG.3 Intravenous efgartigimod alfa (also known as efgartigimod alfa-fcab, where ‘fcab’ stands for Fc fragment with antigen-binding, in the USA; Vyvgart) is the first FcRn antagonist approved in multiple countries, including the USA and the European Union, for the treatment of generalised MG in patients with AChR-Ab+. In Japan, it is approved for generalised MG regardless of antibody status.7 Like mainstream therapies, such as PLEX, efgartigimod alfa also reduces serum IgG levels but without the logistical issues associated with intravenous routes, potentially enabling patients with MG to maintain better disease control and an enhanced health-related quality of life (HRQoL).8 Early clinical trials have reported that efgartigimod alfa may cause various adverse events (AEs), primarily respiratory infections, headaches and urinary tract infections.9 However, long-term experimental studies on the AEs of this drug are still limited, necessitating the analysis of substantial real-world data.

In real-life scenarios, MG treatment must address several issues, including drug-related AEs and treatment resistance, with the goal of improving patients’ QoL. Given its suitability for individualised treatment based on clinical responses, insights gained from follow-up experiences are valuable for optimising the use of efgartigimod alfa. The US Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) is an anonymised database encompassing a large number of drug-related AE reports from healthcare professionals, consumers and manufacturers, among others, and is publicly accessible. It currently contains nearly 30 million reports, making it one of the largest pharmacovigilance databases globally.10

To date, various statistical methods are available for detecting postmarketing drug safety signals.11 12 This study applied four algorithms—Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Multi-item Gamma Poisson Shrinker (MGPS) and Bayesian Confidence Propagation Neural Network (BCPNN)—to analyse AE reports related to efgartigimod alfa in the FAERS database, with the goal of providing novel insights into its safety profile based on real-world data.

The FAERS database has publicly released original data since 2004 on its official website (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html). Since efgartigimod alfa was first approved in December 2022, this study included data from Q1 2022 to Q4 2023, covering a total of eight quarters. Four FAERS subdatabases were used for this study, each providing specific information: the REAC subdatabase provided AE terms and identification codes; the DRUG subdatabase provided drug names, roles in causing AEs, dosages and concomitant medications; the DEMO subdatabase provided demographic information such as gender, age, event date, reporting country and reporter type and the THER subdatabase provided treatment start and end dates.

We conducted data preprocessing to improve the reliability of subsequent analyses (figure 1a). First, efgartigimod alfa was set as the target drug. Considering that the FAERS database allows AE reports to be submitted using either generic or brand names, both the generic name (efgartigimod alfa) and brand names (Vyvgart and Vyvgart Hytrulo) were used to retrieve data to avoid statistical errors. Next, only data where the drug was listed as the primary suspect for AEs were included in the study. We then followed FDA-recommended rules for removing duplicate entries (online supplemental table 1). Finally, we established the Preferred Term (PT) level for this study based on the Medical Dictionary for Regulatory Activities (MedDRA). MedDRA is a continuously updated medical terminology database that provides a standardised system of terms commonly used in drug regulation during both premarket and postmarket phases.13 PT refers to the PT for specific symptoms or diagnoses (eg, dizziness and atrial fibrillation) (figure 1b).

Figure 1

Processing design and research levels of the study. (a) Study flow diagram and (b) the five levels of Medical Dictionary for Regulatory Activities. FAERS, Food and Drug Administration Adverse Event Reporting System.

We reviewed the data obtained through the aforementioned process and summarised the demographic characteristics, including total case count, gender, age group, reporter type, reporting country, reporting quarter and indications.

In this analysis, we ranked AEs at the PT level based on their reporting frequency and generated a list of the top 70 most frequently reported events. Four algorithms were used to detect safety signals: ROR, PRR, MGPS and BCPNN. Only when a specific AE met the criteria of all four algorithms was it considered a significant signal (online supplemental table 2). Unexpected significant signals were defined as those not observed in prior clinical trials at the PT level. It is important to note that PTs related to outcomes such as death, disease, hospitalisation or symptoms of MG, including muscular weakness, muscle fatigue, physical deconditioning, asthenia and eyelid ptosis, were excluded from this analysis.

Time-to-onset data were obtained by calculating the interval between the AE date (EVENT_DT) from the DEMO subdatabase and the treatment start date (START_DT) from the THER subdatabase. Data were excluded if they met any of the following criteria: inaccurate dates, missing dates or if the treatment start date occurred after the AE date. We then calculated the median and IQR and conducted a Weibull Shape Parameter (WSP) test. The WSP test can determine the rate of change in AE occurrence, and the shape parameter β evaluates the risk without a reference population. The classification is as follows: if β<1 and the 95% CI is <1, the drug-related AE risk decreases over time (early failure-type profile); if β is equal to or near 1 and the 95% CI includes 1, the risk persists over time (random failure-type profile) and if β>1 and the 95% CI is >1, the risk increases over time (wear-out failure-type profile).14 Additionally, we used the Kaplan-Meier method to plot the cumulative incidence of efgartigimod alfa-induced AEs.

ROR, PRR, Empirical Bayes Geometric Mean (EBGM and BCPNN values were calculated by comparing them with data from other drugs in the FAERS database for the same period. Data processing and statistical analyses were performed using WPS Office and R software V.4.3.2. Tables were created with WPS Office, and figures were generated using R software V.4.3.2. All data in this study were anonymised and publicly available, so ethical approval was not required.

None.

During the period from Q1 2022 to Q4 2023, a total of 1403 cases of AEs were reported in the FAERS database where efgartigimod alfa was identified as the primary suspect. Among these cases, 99.6% were for the indication of MG, while 0.4% had unknown indications. The quarter with the fewest cases was Q1 2022 (n=4), and the quarter with the highest number of cases was Q2 2023 (n=512) (figure 2). Among these cases, the proportion of females (17.5%) was similar to that of males (15.7%), with 81.7% of cases having unknown gender. The proportion of individuals aged ≥65 years (9.7%) was similar to that of those aged 18–65 years (8.5%), with age being unknown in 81.7% of cases. Regarding the type of reporter, the majority of reports came from consumers (75.3%) and physicians (12.9%), with 0.6% of cases having an unknown reporter. Most reports originated from the USA (86.5%), followed by Japan (10.6%) (table 1).

Table 1

Demographic characteristics of efgartigimod alfa-related AEs in the FAERS database

At the PT level, the reported frequency of the 70 most common AEs ranged from 10 to 123. The most frequently reported AE was urinary tract infection, while procedural headache had the highest signal strength (online supplemental table 3). A total of 14 AEs were identified as significant signals by all four algorithms, including urinary tract infection (ROR=5.51), upper respiratory tract infection (ROR=7.93), procedural headache (ROR=911.25), herpes zoster (ROR=3.49), nephrolithiasis (ROR=3.61), diverticulitis (ROR=5.81), cellulitis (ROR=3.23), poor venous access (ROR=6.28), respiratory tract infection (ROR=4.75), respiratory syncytial virus infection (ROR=6.95), staphylococcal infection (ROR=3.54), sepsis (ROR=3.66), atrial fibrillation (ROR=3.75) and transient ischaemic attack (ROR=4.47) (figure 3). As observed in clinical studies conducted in several countries, including China15 and Japan,16 a small proportion of patients receiving efgartigimod alfa reported mild to moderate AEs, primarily upper respiratory tract infections, which resolved with appropriate antimicrobial treatment.17 A rash was rarely reported, and no long-term disability, MG progression, severe infections or cardiovascular or cerebrovascular events were observed.18 19 Therefore, the occurrences of sepsis, atrial fibrillation and transient ischaemic attack represent unexpected and significant AE signals.

Figure 3

Number of adverse events identified as significant signals by different algorithms. BCPNN, Bayesian Confidence Propagation Neural Network; EBGM, Empirical Bayes Geometric Mean; PRR, Proportional Reporting Ratio; ROR, Reporting Odds Ratio.

Infusion-related or injection site reactions, which are typically mild and self-limiting, may occur with intravenous biological therapies, especially at higher doses.20 21 Previous reports have linked headaches following therapeutic antibody infusion to peak IgG concentrations,22 suggesting that the potential risks for procedural headache (ROR₀₂₅=608.72) and poor venous access (ROR₀₂₅=3.84) observed after efgartigimod alfa treatment may also be dose-related. Although many headaches reported in patients receiving efgartigimod alfa did not meet the typical definition of procedural headache—which is usually related to cerebrospinal fluid leakage following spinal procedures like lumbar puncture—they may still be classified as immune responses triggered by fluid balance or haemodynamic changes during drug infusion or by the drug itself, including those related to infusion.23 24 The high signal strength of procedural headaches across four algorithms can partly be explained by disproportionality analysis. Since this result was derived from a comparison with the overall FAERS database, procedural headaches appear more frequently in reports involving injectable drugs than in those involving oral medications, which are more commonly associated with general headaches. This enhances the signal strength. In an ongoing open-label trial (ADAPT+, NCT03770403) designed to assess the safety and efficacy of efgartigimod alfa, the most common AEs in efgartigimod alfa-treated subjects were infections, headaches, infusion-related reactions and diarrhoea.25 It has been reported that patients with IgG levels consistently below 100 mg/dL are at an increased risk for recurrent infectious diseases, which can sometimes be life-threatening.26 The generation rate or FcRn interactions of IgG4 are considered to differ from those of other IgG subclasses.27 Most AChR and lipoprotein receptor-related protein 4 (LRP4) antibodies belong to the IgG1 subclass, which can activate complement, whereas the IgG4 subclass, including muscle-specific receptor tyrosine kinase antibody (MuSK-Ab), does not.28 Additionally, some efgartigimod alfa-treated subjects developed antidrug antibodies, including neutralising antibodies.24

In the ADAPT trial, the most common AEs associated with efgartigimod alfa were headache, upper respiratory tract infection and urinary tract infection, with most infection-related AEs being mild to moderate, except for three severe cases—two cases of influenza and pharyngitis with efgartigimod alfa and one case of upper respiratory tract infection with placebo. We observed higher risks not only for general infections such as urinary tract infection (ROR₀₂₅=4.61), upper respiratory tract infection (ROR₀₂₅=5.89) and respiratory tract infection (ROR₀₂₅=2.91) but also for specific pathogens such as respiratory syncytial virus infection (ROR₀₂₅=3.84), staphylococcal infection (ROR₀₂₅=1.90) and herpes zoster (ROR₀₂₅=2.38), as well as more severe infections like diverticulitis (ROR₀₂₅=3.70), cellulitis (ROR₀₂₅=2.03) and even sepsis (ROR=2.70). This may be related to the early use of immunosuppressants. The impact of efgartigimod alfa on the MG course may be related to the severity of infections prior to treatment, as patients with milder symptoms may no longer require hospitalisation or additional short-term immunotherapy, such as PLEX, which may lead to more extensive immunosuppression and higher infection risk, a trend also observed with long-term immunosuppressive therapy.29 30 The sustained therapeutic response to efgartigimod alfa depends on accurate monitoring and timely retreatment during severe disease exacerbations. More frequent dosing regimens enable sustained disease control.31 This study also identified unexpected significant signals for atrial fibrillation (ROR₀₂₅=2.76) and transient ischaemic attack (ROR₀₂₅=2.54). It is worth noting that efgartigimod alfa treatment does not necessitate additional vaccination, and patients with MG who have been vaccinated do not appear to experience adverse effects, making it a more convenient option for patients and healthcare providers.32

After screening, time-to-onset data were extracted from 386 valid cases. The median time to onset for efgartigimod alfa-related AEs was 57.00 days. The cumulative incidence of AEs was 37.31% by 30 days after treatment initiation and 64.25% by 100 days (figure 4a,b). The results of the WSP test indicated that the time to onset of AEs for efgartigimod alfa followed an early failure-type profile (table 2). Through the WSP test, we found that during the eight-quarter period analysed, the characteristics of AEs reported for efgartigimod alfa were consistent with an early failure-type profile (with the majority of AEs reported within the first 30 days), indicating that the risk of AEs decreases over time. Consistent with previous findings, more patients with MG responded in the early weeks of efgartigimod alfa treatment—particularly during the first treatment cycle—while those who did not respond early were generally more difficult to treat.33 It should be emphasised that a key limitation of the FAERS database is that time-to-onset data for specific drugs are only recorded to the nearest day, making it impossible to analyse the crucial window within the first 24 hours after administration. This inherent limitation may have introduced bias into the WSP test results. In fact, no AE, such as upper respiratory tract infection or procedural headache, was reported within 24 hours following a single infusion of efgartigimod alfa. Therefore, in addition to comparing groups with similar baseline event rates, it is necessary to design larger-scale studies with refined observation protocols, including assessments within the first 24 hours.

Table 2

Time-to-onset characteristics of AEs reported for efgartigimod alfa in the FAERS database

The pathogenesis of MG primarily involves B cells and autoantibodies that target proteins on the postsynaptic muscle membrane. Currently, four autoantibodies have been identified, targeting AChR, MuSK, low-density LRP4 and agrin.34 Generalised MG is classified into three subtypes: generalised early-onset MG, which typically occurs before the age of 50 years and is often associated with thymic hyperplasia and AChR-Ab; generalised late-onset MG (g-LOMG), occurring after the age of 50 years, typically without thymic abnormalities but with more severe symptoms and generalised seronegative MG, characterised by the absence of typical AChR-Ab and MuSK-Ab, making diagnosis and treatment more complex and underscoring the need for personalised treatment strategies.1 The FcRn antagonist efgartigimod alfa is a human IgG1-derived Fc fragment designed using ABDEG (AntiBody that enhances IgG DEGradation) technology, which retains pH-dependent interactions with FcRn and selectively reduces endogenous IgG levels without affecting other Ig isotypes such as IgA, IgD, IgE and IgM.35 Phase II36 and Phase III ADAPT trials24 have demonstrated the rapid efficacy and safety of efgartigimod alfa in treating MG, though its AE should not be overlooked. The FAERS database is a key source for evaluating drug safety postmarketing, providing insights into unanticipated or rare drug-related AEs, particularly those not observed in earlier clinical trials. To our knowledge, this study is the first to use the FAERS database to evaluate efgartigimod alfa-related AEs in a real-world setting, offering valuable insights for clinical practice.

The decline in HRQoL in patients with MG is driven by several factors, including female gender, advanced age, lower educational levels, limited social support and poor disease acceptance.37 38 Special attention should also be given to patients with comorbidities or severe disease.39 Studies indicate that anxiety and depression are more common in patients with MG, with approximately 50% experiencing mood disorders, which are correlated with increased disease severity.40 41 Among the sample of efgartigimod alfa-related AE cases included in our study, the population aged over 65 years was larger than those aged under 18 years and between 18 and 65 years, and women outnumbered men. Considering some unknown cases, larger-scale epidemiological studies are still needed. Since 2023, there has been a sharp increase in reports of efgartigimod alfa-related AEs compared to 2022, which is likely due to the launch of its new formulation, Vyvgart Hytrulo. This subcutaneous formulation offers the convenience of at-home administration, making it particularly suitable for patients who may be hesitant about intravenous therapy. This ease of use may have contributed to an increase in patient uptake, thereby leading to a higher number of AE submissions. Additionally, the expansion of market coverage and sales channels has facilitated the drug’s global distribution, potentially resulting in greater sales volume and more patients receiving treatment, which could further increase the likelihood of AE reports.

This study has several limitations. First, determining the safe range of IgG levels is crucial for minimising AEs associated with efgartigimod alfa. In MG treatment, the severity of AEs from efgartigimod alfa may depend on the duration and dosage of treatment. It remains unclear to what extent the reduction of autoantibodies is sufficient to alleviate MG symptoms without compromising general protective immunity. Clinicians often need to compare the benefits of continuous versus cyclical therapy when developing treatment plans, and a fair approach would involve identifying fixed time points within the treatment cycle of efgartigimod alfa. At present, no standardised method exists for this purpose. One of the main challenges in diagnosing and evaluating MG is the comparison of quantitative MG scores or Myasthenia Gravis Foundation of America classifications across comparable patient populations. This database relies on self-reporting, and information on these classifications or scores is lacking. Under-reporting may occur when AEs from efgartigimod alfa are too mild or complex. Second, we need to consider the impact of age on efgartigimod alfa treatment and prognosis. Older patients, compared with the average onset age of MG, may present confounding comorbidities that could influence the results. Generally, ageing is associated with the loss of skeletal muscle mass in specific limb and trunk regions. For instance, patients with g-LOMG may require more intensive treatment due to age-related skeletal muscle loss (also known as ‘sarcopenia’), which could result in more severe low IgG depletion and associated AEs, such as increased susceptibility to infections.42 Finally, the high cost of efgartigimod alfa suggests that patients receiving this treatment may have better financial conditions. This study mainly included samples from the USA and Japan, meaning that these patients may have access to better healthcare resources, thereby potentially reducing the occurrence of AEs and mitigating severe outcomes. We recommend further studies with more comprehensive cohorts and matching by propensity score for different ethnicities.

This study aims to evaluate the real-world impact of efgartigimod alfa on AEs in patients with MG, providing actionable insights for early risk identification during treatment. The findings will help optimise efgartigimod alfa treatment strategies and enhance clinical care for patients with MG, contributing to a broader understanding of disease management. This research will also support the development of personalised treatment strategies for patients with MG receiving efgartigimod alfa.

Not applicable.

We thank the US Food and Drug Administration for making raw data from the FAERS database publicly available.