Dermatologic Immune-Related Adverse Events From Immune Checkpoint Inhibition: Diagnostic and Therapeutic Challenges, Prognostic Implications, and Value of Inpatient Dermatology Consultation

Abstract

Background and Objectives: As immune checkpoint inhibitors (ICIs) continue to revolutionize the treatment of countless advanced malignancies, the immune-related adverse events (irAEs) they provoke are likely to increase in prevalence and complexity. Dermatologic irAEs – though among the most common of these unique, multisystem toxicities – nevertheless remain among the most poorly understood. Particularly notable knowledge gaps exist regarding the ways that dermatologic irAEs differ in presentation and therapeutic responsiveness from their classic autoimmune counterparts, the incidence and implications of co-occurring complications from ICI use, and the role and responsibility of dermatology consultation in the management of severe skin toxicities. The underlying goal of the studies presented within this thesis is to build a foundational understanding and provide an evidence-based framework that can begin to fill these knowledge gaps. Methods and Results: To appraise clinically relevant differences between dermatologic irAEs and their idiopathic counterparts, we focused on bullous pemphigoid (BP) and Stevens-Johnson syndrome (SJS) as independent case studies that may offer insight into the broader question. After matching ICI-induced BP (ICI-BP) cases to idiopathic BP (iBP) controls, we found that ICI-BP presents unique diagnostic challenges – prolonged pruritic prodrome (medians, 28 vs7 days, P < .001) and non-bullous presentation – compared to classic iBP that may lead to diagnostic delays as well as a disproportional need for immunosuppression to manage ICI-BP compared to iBP (odds ratio [OR], 9.4; P = .04). Reviewing our multicenter cohort of a severe, generalized bullous eruptions mimicking SJS, we proposed a new terminology for this atypical toxicity pattern - progressive immunotherapy-related mucocutaneous eruption (PIRME). We found that PIRME is notably distinct from SJS in its delayed onset (median, 63 days from ICI initiation), mild initial presentation, rare ocular involvement, benign clinical course, and favorable treatment response. To better understand co-occurring ICI toxicities, we undertook a retrospective cohort study of ICI-induced colitis – the next most common irAE – patients with or without dermatologic irAE. We found that co-occurring ICI-induced rash and colitis is common, may indicate increased risk of developing additional irAEs (OR, 18.5; P = .001), and could be among the earliest clinical markers of improved outcomes, compared to patients with colitis alone, in progression-free survival (hazard ratio [HR], 0.32; P = .001) and overall survival (HR, 0.32; P = .018). Lastly, to quantify the value of involving an inpatient dermatology service, we identified all patients admitted to our institution with dermatologic irAEs. We found that patients who received dermatology consultation were significantly less likely to experience disruptions in their oncologic management (exposure to systemic immunosuppression or need for ICI discontinuation) due to dermatologic irAE, compared to patients without dermatology consultation (OR, 0.03; P = .015). Conclusions: Our research indicates that while dermatologic irAEs may represent uniquely challenging entities that are distinct from their classic autoimmune counterparts, multidisciplinary collaboration and involvement of expert subspecialists remain crucial to the prompt recognition, adequate management, and responsible counseling of these medically complex and increasingly prevalent patients.