A checkup on immune checkpoint inhibitor-induced subacute cutaneous lupus

By Warren R. Heymann, MD, FAAD
March 15, 2023
Vol. 5, No. 11

Keeping up with the expanding indications and adverse reactions to immune checkpoint inhibitors (ICIs) can be a full-time job. These drugs are anti-neoplastic because of cytotoxic T-cell activation. ICIs include the anti-cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4) agent ipilimumab, anti- programmed cell death protein-1 (PD-1) antibodies including pembrolizumab, nivolumab, and cemiplimab, and anti- programmed death-ligand 1 (PD- L1) inhibitors such as atezolizumab, avelumab, and durvalumab. Unsurprisingly, altering the immune response can lead to immune intolerance and immune-related adverse events. Indeed, > 60% of patients develop adverse effects, potentially involving any organ, resulting in diverse disorders such as thyroiditis, hepatitis, pneumonitis, hypophysitis, uveitis, polyneuritis, pancreatitis, colitis, myocarditis, and cutaneous eruptions. (1)

Cutaneous side effects (CSE) in patients treated with ICIs include morbilliform, follicular, pruritic, pustular, vesicular, acneiform, and exfoliative eruptions. CSE occur in up to 45% of patients treated with ipilimumab and 34% of patients treated with nivolumab and pembrolizumab. (2) In their retrospective review of 352 consecutive oncology patients treated with ICIs for cancer, 46 patients (13.1%) experienced CSE. The incidence of CSE was less with nivolumab (n=16; 9.5%) and pembrolizumab monotherapy (n=9; 9.6%) as compared to ipilimumab (n=10; 23.3%) and combination therapy (n=11; 23.9%); P<0.05. Skin toxicity demonstrated rash/eczema (n=28; 60.9%), autoimmune (n=8; 17.4%, vitiligo n=5, lichen sclerosus n=2, guttate psoriasis n=1), lichenoid reaction (n=5;10.9%), pruritus (n=4; 8.7%), and a miscellaneous group (n=3; 6.5%). The limited severity grades of CSE caused immunotherapy disruption in only three (0.9%) cases. (2) Other cutaneous disorders reported with ICIs include bullous pemphigoid, Grover’s disease, vasculitis, Steven-Johnson syndrome/Toxic epidermal necrolysis, and alopecia areata. Less common adverse reactions include DRESS, sarcoidosis, dermatomyositis, pyoderma gangrenosum, and sclerodermoid reactions. (1)

More than 100 drugs from more than 10 drug classes are suspected to cause drug-induced lupus erythematosus. It has been estimated that up to 30% of all SCLE cases are drug-induced. (3) Drugs most likely to trigger SCLE include hydrochlorothiazide, calcium channel blockers and angiotensin-converting enzyme inhibitors. Drugs such as proton-pump inhibitors (PPIs), terbinafine, immunomodulators (leflunomide, TNF-α inhibitors), and chemotherapeutic agents, can also induce SCLE. (4) This commentary will focus on the increasing number of cases of subacute cutaneous lupus erythematosus (SCLE) in patients on ICIs.

Liu et al observed the development of biopsy and serologically-proven SCLE in a patient with metastatic non-small cell lung cancer for which she had been started on nivolumab 5 months earlier. After a brief discontinuation of nivolumab, her SCLE was controlled with topical betamethasone, oral prednisolone, and hydroxychloroquine, thereby allowing nivolumab to be successfully readministered. (5) Subsequently, Zitouni et al reported 2 cases of nivolumab-induced SCLE — a 72 year-old woman with melanoma and a 43 year-old man with metastatic non-small-cell lung cancer. Both patients tested positive for antinuclear antibodies and anti-SSA antibodies. Lesions regressed with topical corticosteroids and hydroxychloroquine for the first patient and oral prednisone for the second patient. No systemic involvement was observed. (6) Kosche et al reported a patient with metastatic serous ovarian carcinoma undergoing treatment with combination ipilimumab and nivolumab who developed subacute cutaneous lupus erythematosus (SCLE), who was able to successfully restart immunotherapy after a course of oral corticosteroids and maintenance oral hydroxychloroquine and topical corticosteroid therapy. (7) Diego et al reported 2 additional cases of nivolumab-induced SCLE; of note is that one case of SCLE, in a 75-year-old man with non-small-cell lung cancer, continued to flare 6 months after nivolumab discontinuation. (8) Marano et al reported 2 more cases, induced by nivolumab and pembrolizumab, respectively. Interestingly, the nivolumab-treated patient’s SCLE cleared with hydroxychloroquine and steroids (oral and topical), but he developed dermatomyositis with reintroduction of nivolumab. (8)

Bui et al presented 5 patients who developed SCLE while on PD-1 or PDL-1 inhibitors, manifested by photodistributed, papulosquamous eruptions. All patients were on long-standing omeprazole (a well-known cause drug-induced SCLE), yet did not develop SCLE until the administration of the ICIs. The authors proposed a “multihit” hypothesis in which PD-1/PDL1 blockade may have unmasked an immune response to a previously tolerated medication, resulting in the SCLE phenotype. Alternatively, the ICIs may have directly stimulated B-cell mediated humoral immunity with the development of autoantibodies directed against cutaneous targets. (9) Albeit rare, it must also be remembered that paraneoplastic SCLE is a distinct entity that was reported prior to the advent of ICIs. (10)

Theoretically, it is fascinating to ponder why some patients develop SCLE and others different immunologic adverse reactions. In a retrospective study of 4,487 patients receiving ICIs, 11 patients had immunotherapy-associated cutaneous connective tissue disease, for a frequency of 0.025%. Eight cases were SCLE (72.7%), 1 SLE (9.1%), 1 eosinophilic fasciitis (9.1%), and 1 case of dermatomyositis (9.1%). (11) On a practical level, most patients may continue PD-1/PD-L1 inhibitor therapy (or after a brief respite, reinstitute therapy), while successfully managing patients with steroids (oral or topical) with or without hydroxychloroquine. As we gain more experience with these medications, more refined management guidelines will be forthcoming.

Point to Remember: Subacute cutaneous lupus erythematosus has been associated with, or induced by, immune checkpoint inhibitors. Clinicians should be familiar with this information as the number of such cases will likely increase along as new immune checkpoint inhibitors are developed for diverse malignant tumors.

Our experts’ viewpoint

Emily Y. Chu, MD, PhD, FAAD
Associate Professor of Dermatology & Pathology and Laboratory Medicine
Perelman School of Medicine at the University of Pennsylvania
Philadelphia, PA

Since the advent of ICIs, we have learned a great deal about the spectrum of their associated cutaneous adverse reactions, especially given the increasing number of indications for their use. That such a diverse array of cutaneous effects are seen in association with ICIs is fascinating from many perspectives, not the least of which is the potential for dermatologists to gain more insight into the etiology of non-ICI related “ordinary” counterparts of ICI reactions, including lichenoid dermatitides, vitiligo, dermatomyositis, lupus erythematosus and sarcoidosis, to name a few. There is frequent discussion about whether ICI cutaneous reactions represent bona fide disease or rather “like” eruptions, i.e. “lupus-like.” In some ICI skin reactions, the clinical and histopathologic features show many features of, but not do completely recapitulate, their non-ICI counterparts, for instance the lupus-like findings in a pembrolizumab treated patient described in Shao et al. (12) In other cases, such as the ones described by Liu et al. and Zitouni et al. discussed above, the serological as well as dermatologic findings make a convincing case for bona fide cutaneous lupus. (5,6)

Another intriguing aspect of ICI reactions is their timing after starting immunotherapy, which are frequently late, defined as occurring greater than 3 months after initiation of therapy. (13) Not infrequently, ICI adverse cutaneous reactions may even occur following discontinuation of ICI therapy, as seen in one of the cases presented by Zitouni et al. (13, 6) ICIs are capable of producing highly durable tumor responses, persisting often even after the therapy has been discontinued. (14) As such it is not surprising that their effects in the skin may similarly be prolonged. For dermatologists, being aware of delayed and prolonged effects of ICIs is important in identifying the correct drug culprit when a medication reaction is suspected.

Joseph F. Merola, MD, MMSc, FAAD
Vice Chair of Clinical Trials and Innovation
Director, Center for Skin and Related Musculoskeletal Diseases (SARM)
Director, Clinical Unit for Research Innovation and Trials (CUReIT)
Associate Program Director, Combined Medicine-Dermatology Residency Program
Dept of Dermatology and Dept of Medicine, Division of Rheumatology
Brigham and Women's Hospital, Boston, MA

There is an ever-increasing list of agents associated with drug-induced SCLE. It therefore remains important to consider the possibility of a drug-induced cause in every patient presenting with SCLE — inquiring about not only prescription but also commonly used over-the-counter medications (such as proton-pump-inhibitors). A contextual factor that might alter pre-test probability of drug-induced vs. idiopathic SCLE, include the presence of underlying systemic connective tissue disease. Clinical features that might suggest a drug-induced SCLE over idiopathic SCLE include a more wide-spread distribution over the body (including in non-photo exposed areas); more aggressive features including erythema multiforme-like, bullous or concurrent vasculitic lesions. (15) Interestingly, the presence of tissue eosinophilia does not appear to be an indicator of drug-induced disease, nor is it distinguished by presence of mucin or other histopathologic feature. (16)

Another important mention is the distinction of drug-induced lupus syndrome from drug-induced SCLE. Classical drug-induced lupus, of the variety taught in medical school, is driven by specific drugs, namely procainamide and hydralazine, considered high risk drugs and leading to systemic features such as fever, arthritis, serositis and associated with anti-histone antibodies. Drug-induced SCLE is generally a skin-limited process and is distinct from this drug-induced lupus syndrome, only infrequently having any associated arthralgia or other such systemic manifestations.

It may come as no surprise that immune-checkpoint inhibitors are capable of producing SCLE among myriad other drug-induced / drug-exacerbated autoimmune conditions which this class of agent seems to be capable of producing. To many oncologists, the appearance of these inflammatory skin eruptions are lumped into the category of ‘hypersensitivity or cutaneous toxicity,’ where nuance may be lost. It will be up to the dermatologist to uncover the eruption of SCLE in these cases, and to maintain a high index of suspicion for a drug-induced etiology. We should also be aware that treating through more mild-moderate drug-induced SCLE without systemic features may be reasonable in order to maintain necessary oncologic therapy. This may be accomplished with a variety of agents including topical and/or systemic corticosteroids as needed, but especially with the use of anti-malarial and other CLE-active therapies as patients navigate the balance between their underlying malignancy treatment and the toxicities inherent to many of these cancer therapies.

1. Muntyanu A, Netchiporouk E, Gerstein W, Gniadecki R, Litvinov IV. Cutaneous Immune-Related Adverse Events (irAEs) to Immune Checkpoint Inhibitors: A Dermatology Perspective on Management [Formula: see text]. J Cutan Med Surg. 2021 Jan-Feb;25(1):59-76. doi: 10.1177/1203475420943260.

2. Rovers JFJ, Bovenschen HJ. Dermatological side effects rarely interfere with the continuation of checkpoint inhibitor immunotherapy for cancer. Int J Dermatol. 2020 Dec;59(12):1485-1490. doi: 10.1111/ijd.15163. Epub 2020 Sep 8. PMID: 32895923.

3. Haugaard JH, Kofoed K, Gislason G, Dreyer L, Egeberg A. Association Between Drug Use and Subsequent Diagnosis of Lupus Erythematosus. JAMA Dermatol. 2020 Sep 2;156(11):1–9. doi: 10.1001/jamadermatol.2020.2786. Epub ahead of print. PMID: 32876672; PMCID: PMC7489398.

4. He Y, Sawalha AH. Drug-induced lupus erythematosus: an update on drugs and mechanisms. Curr Opin Rheumatol. 2018 Sep;30(5):490-497. doi: 10.1097/BOR.0000000000000522. PMID: 29870500; PMCID: PMC7299070.

5. Liu RC, Sebaratnam DF, Jackett L, Kao S, Lowe PM. Subacute cutaneous lupus erythematosus induced by nivolumab. Australas J Dermatol. 2018 May;59(2):e152-e154. doi: 10.1111/ajd.12681. Epub 2017 Jul 20. PMID: 28726325.

6. Zitouni NB, Arnault JP, Dadban A, Attencourt C, Lok CC, Chaby G. Subacute cutaneous lupus erythematosus induced by nivolumab: two case reports and a literature review. Melanoma Res. 2019 Apr;29(2):212-215. doi: 10.1097/CMR.0000000000000536. PMID: 30489484.

7. Kosche C, Owen JL, Choi JN. Widespread subacute cutaneous lupus erythematosus in a patient receiving checkpoint inhibitor immunotherapy with ipilimumab and nivolumab. Dermatol Online J. 2019 Oct 15;25(10):13030/qt4md713j8. PMID: 31735010.

8. Marano AL, Clarke JM, Morse MA, Shah A, Barrow W, Selim MA, Hall RP 3rd, Cardones AR. Subacute cutaneous lupus erythematosus and dermatomyositis associated with anti-programmed cell death 1 therapy. Br J Dermatol. 2019 Sep;181(3):580-583. doi: 10.1111/bjd.17245. Epub 2018 Dec 10. PMID: 30244487.

9. Bui AN, Hirner J, Singer SB, Eberly-Puleo A, Larocca C, Lian CG, LeBoeuf NR. De novo subacute cutaneous lupus erythematosus-like eruptions in the setting of programmed death-1 or programmed death ligand-1 inhibitor therapy: clinicopathological correlation. Clin Exp Dermatol. 2021 Mar;46(2):328-337. doi: 10.1111/ced.14449. Epub 2020 Oct 14. PMID: 32939795.

10. Evans KG, Heymann WR. Paraneoplastic subacute cutaneous lupus erythematosus: an underrecognized entity. Cutis. 2013 Jan;91(1):25-9. PMID: 23461055.

11. Bui AN, Singer S, Hirner J, Cunningham-Bussel AC, Larocca C, Merola JF, Lian CG, LeBoeuf NR. De novo cutaneous connective tissue disease temporally associated with immune checkpoint inhibitor therapy: A retrospective analysis. J Am Acad Dermatol. 2021 Mar;84(3):864-869. doi: 10.1016/j.jaad.2020.10.054. Epub 2020 Oct 24. PMID: 33323344.

12. Shao K, McGettigan S, Elenitsas R, Chu EY. Lupus-like cutaneous reaction following pembrolizumab: An immune-related adverse event associated with anti-PD-1 therapy. J Cutan Pathol. 2018 Jan;45(1):74-77. doi: 10.1111/cup.13059. PMID: 29028121

13. Wang LL, Patel G, Chiesa-Fuxench ZC, McGettigan S, Schuchter L, Mitchell TC, Ming ME, Chu EY. Timing of Onset of Adverse Cutaneous Reactions Associated With Programmed Cell Death Protein 1 Inhibitor Therapy. JAMA Dermatol. 2018 Sep 1;154(9):1057-1061. doi: 10.1001/jamadermatol.2018.1912. PMID: 30027278

14. Robert C, Ribas A, Hamid O, Daud A, Wolchok JD, Joshua AM, Hwu WJ, Weber JS, Gangadhar TC, Joseph RW, Dronca R, Patnaik A, Zarour H, Kefford R, Hersey P, Zhang J, Anderson J, Diede SJ, Ebbinghaus S, Hodi FS. Durable Complete Response After Discontinuation of Pembrolizumab in Patients With Metastatic Melanoma. J Clin Oncol. 2018 Jun 10;36(17):1668-1674. doi: 10.1200/JCO.2017.75.6270. PMID: 29283791

15. Marzano AV, Lazzari R, Polloni I, Crosti C, Fabbri P, Cugno M. Drug-induced subacute cutaneous lupus erythematosus: evidence for differences from its idiopathic counterpart. Br J Dermatol. 2011 Aug;165(2):335-41. doi: 10.1111/j.1365-2133.2011.10397.x. Epub 2011 Jul 11. PMID: 21564069.

16. Hillesheim PB, Bahrami S, Jeffy BG, Callen JP. Tissue eosinophilia: not an indicator of drug-induced subacute cutaneous lupus erythematosus. Arch Dermatol. 2012 Feb;148(2):190-3. doi: 10.1001/archdermatol.2011.290. Epub 2011 Oct 17. PMID: 22004879.

---