Adverse Cutaneous Reactions to Monoclonal Antibodies: Stevens-Johnson Syndrome, Toxic Epidermal Necrolysis, Erythema Multiforme, and Fixed Drug Eruption – A Systematic Review Open Access

The use of monoclonal antibodies (mAbs) for treatment has become an increasingly important field of medical research. Since the introduction of the first therapeutic mAb, Muromonab-CD3 or OKT3, in 1984, the landscape of mAb development has expanded dramatically. As of 2022, the FDA has approved 110 mAbs for the treatment of various health conditions, with hundreds more currently undergoing clinical trials [1‒3]. With mAbs becoming increasingly prevalent in the treatment of a variety of conditions, research regarding adverse events in the real-world setting is crucial. Adverse reactions to mAbs are varied, impacting single or multiple organ systems, and are often associated with hypersensitivity reactions of types I, II, III, and IV, encompassing conditions such as anaphylaxis, serum sickness, and vasculitis [2, 3].

Despite the broad documentation of adverse events, the adverse dermatological events (ADEs) of mAbs are less thoroughly characterized, particularly Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), fixed drug eruption (FDE), and erythema multiforme (EM). SJS/TEN are rare drug reactions characterized by splitting of the epidermal junction with subsequent development of dusky red/purpuric lesions, blisters, and mucosal sloughing [4]. SJS and TEN can be life-threatening conditions, with a reported mortality rate of around 5% and 15%, respectively [4, 5]. EM is classically present as macules or papules with 3 distinct color zones, referred to as target lesions; when mucosal involvement is present, as occurs in up to 60% of cases, a diagnosis of EM major is rendered [6]. While herpes simplex virus is the most common cause of EM, drugs have also been implicated as a potential trigger [7]. FDE can occur after the ingestion of certain drugs and is present localized to the same areas after repeated drug administration. These lesions can include patches, vesicles, and bullae that are prone to rupture and can involve the mucosa [8]. While EM and FDE are often self-limiting, they can be highly painful and greatly affect the quality of life in recurrent cases. In this review, we synthesize the current literature regarding cutaneous adverse reactions including SJS, TEN, EM, and FDE following mAb administration, highlighting the need for increased understanding of these potential reactions in the context of growing mAb use.

A comprehensive literature search was conducted on PubMed for English-language articles published from 1980 to January 2024, focusing on key terms including “monoclonal antibody,” “Stevens-Johnson Syndrome,” “Toxic Epidermal Necrolysis,” “Erythema Multiforme,” and “Fixed Drug Eruption.” The queried articles were uploaded to Covidence for screening by three team members who reviewed titles and abstracts, requiring at least two votes for an article to advance. Full articles were then assessed for data extraction suitability. Duplicates were removed during this process. Inclusion and exclusion criteria, detailed in the next section, guided both screening stages.

Eligible articles had to undergo peer review, involve human subjects, be published between 1980 and 2024, report original and primary data in English, explicitly mention mAb administration, and mention the occurrence of one of the four specified conditions (SJS, TEN, EM, and FDE) following mAb administration to be included in the study. Studies were excluded if they did not explicitly mention mAb use, did not report an association with SJS, TEN, EM, or FDE, did not undergo peer review, were written in a language other than English, or were conducted on animals or in vitro.

Following the screening process, eligible studies were extracted using a standardized collection sheet, with the data extraction divided equally among four team members. Data extraction included author information, year of publication, study design, number of cases, and demographic details (age, sex). It also included the underlying diseases, specifics of mAb treatment (antibody used, dosage, concurrent medications, time until rash onset), and rash characteristics (type, mucosal involvement, body area affected, treatment, and response). Other complications or notable findings were documented when present. These findings were systematically organized and are present in Tables 1-3 to facilitate analysis and interpretation.

Additionally, the review adhered to the PRISMA guidelines for the design and reporting of this systematic review (see PRISMA Checklist in the online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000545623). Risk of bias was assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Case Reports by four reviewers, with the included studies divided equally among them (see online suppl. Table for risk of bias assessments).

The initial search on PubMed yielded 2,002 results, with Covidence identifying two duplicates, leaving 2,000 studies for title and abstract screening. Of these, 1,942 were excluded based on the inclusion/exclusion criteria, and 58 underwent full-text review. This review led to the exclusion of 29 studies due to various reasons: not being peer-reviewed [1], wrong study design [7], absence of an mAb-related rash [6], and lack of mention of an mAb or rash [15]. Consequently, 29 articles met the criteria for data extraction. This selection process is depicted in Figure 1, with the extracted data presented in Tables 1, 2, and 3.

Among the 31 cases analyzed, 5 lacked patient sex data, and five omitted the time to onset of the rash. Melanoma was the most frequently treated disease [6 cases], followed by squamous cell carcinoma [5] and non-small cell lung carcinoma [5]. Systemic steroids were the most common medication used to treat these ADEs, having been utilized in 21 cases. Complete resolution or disappearance of the rash was reported in 12 cases, with significant improvement in another 11. Seven deaths were reported, all in patients with SJS or TEN due to varying etiologies; however, only two deaths were directly attributed to ADEs as a result of TEN.

Sixteen patients out of 31 cases were not on any other medications during mAb administration. For the remaining 15 patients on other concurrent medications, the most common were aspirin, antibiotics, beta-blockers, ACE inhibitors, antiepileptic drugs, and diabetes medications. All of these patients were taking their respective concurrent medications prior to the initiation of mAb treatment. Thirteen different mAbs were used, with pembrolizumab (7) and nivolumab (5) being the most prevalent. The oral mucosa was the most frequently affected mucosa, occurring in 16 of the 18 reported cases of mucosal involvement.

In this study, the term “SJS/TEN” encompasses the entire spectrum of SJS, TEN, and SJS-TEN overlap conditions. We determined the mean and median onset times for SJS/TEN rashes following anti-PD1 (pembrolizumab, nivolumab, and sintilimab) therapy to be 38.77 days and 28 days, respectively. Specific to the most common agent used among the PD-1 inhibitors, pembrolizumab’s mean and median time to onset of SJS/TEN was found to be 50.7 days and 35 days, respectively. Lastly, risk of bias, as assessed using the JBI Critical Appraisal Checklist for Case Reports, showed that all included studies were rated as having low risk of bias, except for two studies by Dika et al. [12] and Rupp et al. [26], which were rated as having moderate risk.

In our study, we examined the ADEs associated with mAb administration, which is becoming a central tenet of cancer therapy. We found that anti-PD1 checkpoint inhibitors, particularly pembrolizumab, nivolumab, and sintilimab, were the class of mAbs most frequently linked with causing severe dermatological reactions such as SJS/TEN and EM. Among these, pembrolizumab emerged as the mAb most commonly associated with ADEs, specifically SJS/TEN. However, high causality cannot be established from this study alone given the greater use of pembrolizumab as compared with other immunotherapies for the treatment of a variety of different types of cancer. However, this finding mirrors past reviews done on immune checkpoint inhibitors (ICIs) which reported ICIs being significantly associated with increased risk of SJS/TEN in patients [38].

SJS/TEN typically occurs within days to weeks after starting a new medication. One of the fastest reported onsets of action was observed in a 6-year retrospective study, which noted a mean time to onset of 4.1 days following anticonvulsant administration [39]. Similarly, another study reported that reactions began within just a few days of starting acetaminophen [40]. While anti-PD1 therapies are linked to acute and severe cases of true SJS/TEN, they are more commonly associated with causing SJS/TEN-like reactions of gradual onset and milder symptoms. To better understand and differentiate these potentially less severe cases, the concept of progressive immunotherapy-related mucocutaneous eruption (PIRME) has been proposed to classify these dermatologic reactions, which are distinctive from true SJS/TEN in their time to onset, appearance, mucosal involvement severity, and potential treatment [41]. Our study supports this, as the SJS/TEN reactions caused by anti-PD1 ICIs were typically subacute, with a slower onset of appearance (mean and median onset time of 38.77 days and 28 days, respectively) and no reported fatalities.

Among the three anti-PD1 mAbs studied, pembrolizumab exhibited the longest mean onset time for SJS/TEN at 50.7 days, with a median of 35 days. This prolonged timeframe likely reflects the gradual onset characteristic of PIRME and could be due to the extended half-life of these mAbs compared to other common causative agents. For example, pembrolizumab has a half-life of 27 days, whereas lamotrigine and acetaminophen have half-lives of approximately 24.1–35 h and 5.4 h, respectively [42‒44]. Similarly, the delayed onset of SJS/TEN was observed with non-PD1 inhibitors such as mogamulizumab and rituximab, which have half-lives of 16–18 days and 22 days, respectively [45, 46]. For each drug, the mean and median onset times were equal: 75 days for mogamulizumab and 26 days for rituximab. With prolonged half-lives, steady-state drug concentration is achieved much later than what would be expected with common causative agents. For example, rituximab takes 6–8 weeks to reach steady-state concentration when infused weekly [47]. The delay in SJS/TEN reactions may be attributed to the time required to achieve elevated serum drug concentrations sufficient to trigger these reactions, and similarly the long half-life of mAbs likely extends the duration of the reaction, necessitating longer courses of treatment such as corticosteroids to achieve full resolution.

Additionally, EM reactions, which typically occur about 3 days after drug ingestion, can be markedly delayed when induced by mAbs [48]. For example, our review found that the mean time to onset of EM was 18 months for the one reported case due to golimumab, 1 month for both infliximab and secukinumb, and 11 days for adalimumab. Taken together, our findings suggest that different classes of mAbs have the potential for delayed onset and atypical presentation of ADEs. This deviation is significant because it challenges the conventional diagnostic and management strategies that clinicians may employ based on their experiences with more immediate drug reactions. As such, this necessitates that clinicians maintain a vigilant and prolonged approach to patient monitoring, ensuring that adverse events, whether acute or subacute, are promptly identified and managed, even well after mAb therapy has begun.

Anti-TNF-alpha inhibitors were identified as the second most common class of mAbs leading to ADEs. Notably, adalimumab emerged as the mAb that most commonly caused FDE in our review (2 out of 3 cases), in one case with a delayed onset of reaction (12 days after initiation of drug) as would normally be expected with FDE. Although the exact cause of FDE is not fully understood, it is thought to be associated with the activation of CD8+ T cells and extended ICAM-1 expression in resident keratinocytes, which prompts the release of many immunological mediators [49]. Our findings suggest TNF-alpha blockage could be a key factor in this process.

The study also demonstrates that the patient’s sex may not influence the likelihood of developing an ADE. In the 26 cases where patient sex was reported, there was an equal distribution of ADEs between male and female patients (13 each). Additionally, we explored the potential impact of concurrent medication use on the prevalence of ADEs associated with mAbs. Interestingly, our findings demonstrated that nearly half the patients who developed a reaction were solely on mAb therapy (16 patients). Furthermore, patients were already on concurrent medications before starting mAb therapy, making it unlikely that these concomitant drugs were the primary cause of the observed ADEs and suggesting that the ADEs were due to the inherent risk from mAb treatment rather than other medications. Although most case reports included in this study did not specify the exact timing or duration of concurrent medication use, this temporal distinction reinforces the association between mAbs and the reported ADEs.

The study’s limitations stem from its reliance on retrospective data and variability in reporting standards, highlighting the need for prospective studies to validate and expand our findings. Future studies should aim to investigate the potential for genetic or biomarker predictors of susceptibility, anti-PD1 relationship to PIRME, and evaluate the effectiveness of different management strategies in real-world settings.

In conclusion, the clinical significance of our research extends into the realms of patient care optimization and healthcare policy, informing guidelines for the monitoring and management of the growing number of patients on mAb therapy. Clinicians must be aware that the risk of serious ADEs can persist well beyond the initial phases of mAb treatment. A more nuanced approach to patient care must be adopted, incorporating longer follow-up periods and tailored surveillance strategies to detect and manage these delayed reactions effectively. Educating patients on early symptoms and encouraging photographic documentation of lesions may aid in tracking disease progression and guiding clinical decisions. Additionally, patients should be counseled that discontinuation of the offending mAb may not lead to immediate resolution. For those with a history of severe reactions such as SJS/TEN, a medical alert bracelet may also be advisable to prevent unintentional reexposure. This approach not only ensures the safety and well-being of patients but also enhances the therapeutic potential of mAb treatments by managing their risks more effectively.

A statement of ethics is not applicable because the study is based exclusively on published literature.

The authors have no conflict of interest to declare.

This study was not supported by any sponsor or funder.

K.K.: formulation of project, data collection, data extraction, primary writing of journal, and editing. I.A. and A.B.: data collection, data extraction, secondary writing of journal, and editing. J.K. and I.R.: data collection, data extraction, and critical revision and editing. S.W.: formulation of project, secondary writing of journal, and critical revision and editing.

The data that support the findings of this study are available from articles indexed in PubMed. While many of the referenced articles are publicly accessible, some may require a subscription or purchase to access the full text. The articles can be found through PubMed at https://pubmed.ncbi.nlm.nih.gov. This systematic review was not registered.