Abstract
This cohort study investigates the association of systemic corticosteroid exposures with infectious complications and survival outcomes among patients with a cutaneous immune-related adverse event.
Cutaneous immune-related adverse events (cirAEs) may disrupt immune checkpoint inhibitor (ICI) therapy.1 Current guidelines recommend systemic corticosteroids (SCS) for treatment of moderate to severe cirAEs, but SCS-associated complications and association with survival remain poorly understood.1 We investigated the association of SCS exposures with infectious complications and survival outcomes among patients with a cirAE.
Methods
A total of 2459 patients who initiated anti–programmed cell death 1/ligand 1 and/or anti–cytotoxic T-lymphocyte–associated antigen 4 ICI therapy at Massachusetts General Hospital between January 1, 2016, and March 8, 2019, with possible cirAEs (n = 910) were identified using International Classification of Diseases and Related Health Problems, Tenth Revision codes, with subsequent confirmation of cirAE status through manual records review (IRB No. 2017P000501).1,2 This study was approved by the Massachusetts General Brigham Institutional Review Board. Informed consent was waived given the minimal risk posed to participants under the Common Rule (45 CFR 46). Eligible cirAEs included reactions emerging after ICI initiation, consistent with established morphologic categories, and attributed to ICI by the evaluating oncologist or dermatologist.1,2 For each patient with a cirAE (n = 358), oncologic history, clinical features, SCS exposures, infection rates, and survival outcomes were abstracted.2,3,4,5 Exposures to SCS were categorized by indication (cirAE, other immune-related adverse event, other medical reason) and dosage in prednisone equivalents (low, ≤7.5 mg/d for ≥2 months; moderate, >7.5 mg/d for ≥2 months; high, ≥1 mg/kg/d for ≥1 week).4,5 Infection rates were compared among patients treated with SCS for an initial cirAE and those with no steroid exposures for any indication. Cox proportional hazards models were used to assess associations between SCS for first cirAE episode, progression-free survival (primary outcome), and overall survival (secondary outcome). Cox proportional hazards models were adjusted for age, sex, and covariates with P < .05. Further methodological details may be found in the eMethods in the Supplement.2,5 All statistical analyses were conducted using Stata/IC, version 16.0 (StataCorp LLC).
Results
Of 2459 patients assessed, 358 (14.6%) developed a cirAE (median [interquartile range] age, 64 [55-73] years; 145 [40.5%] were women; 150 [41.9%] had melanoma) (Table 1). A total of 50 (14.0%) patients received SCS for an initial cirAE, 192 (53.6%) received SCS for another indication, and 116 (32.4%) had no SCS exposures. Patients who received SCS for an initial cirAE had higher median rash severity (median [interquartile range] Common Terminology Criteria for Adverse Events grade 3 [2-3] vs grade 1 [1-2]; P < .001) and were more likely to be hospitalized for cirAE management (n = 10 [20.0%] vs n = 6 [1.9%]; P < .001) than those who did not receive SCS for this indication (n = 42 [84.0%]). Delivery of SCS for an initial cirAE was predominantly at low doses (n = 42 [84.0%]). Common non-SCS treatments for an initial cirAE included topical corticosteroids (n = 190 [53.1%]) and oral antipruritics (n = 153 [42.7%]).
Table 1. Demographic and Clinical Features.
| Characteristic | No. (%) | P valuec | ||
|---|---|---|---|---|
| SCS for initial cirAE (n = 50)a | SCS for any other indication (n = 192)b | No SCS for any indication (n = 116) | ||
| Age, median (IQR), y | 67 (57-78) | 63 (53-71) | 66 (57-73) | .003 |
| Female sex | 19 (38.0) | 85 (44.3) | 41 (35.3) | .29 |
| Cancer type | ||||
| Melanoma | 13 (26.0) | 99 (51.6) | 38 (32.8) | .001 |
| NSCLC | 13 (26.0) | 31 (16.2) | 15 (12.9) | |
| GI | 4 (8.0) | 13 (6.8) | 17 (14.7) | |
| Head or neck | 7 (14.0) | 9 (4.7) | 16 (13.8) | |
| Otherd | 13 (26.0) | 40 (20.8) | 30 (25.9) | |
| Cancer stage, median (IQR) | 4 (3-4) | 4 (3-4) | 4 (3-4) | .79 |
| ECOG, median (IQR) | 1 (0-1) | 1 (0-1) | 1 (0-1) | .08 |
| Pre-ICI treatments, median (IQR) | 1 (0-2) | 1 (0-2) | 1 (0-2) | .21 |
| ICI precipitating initial cirAE | ||||
| CTLA-4 | 2 (4.0) | 11 (5.7) | 1 (0.86) | .009 |
| PD-1 or PD-L1 | 42 (84.0) | 134 (69.8) | 99 (85.3) | |
| CTLA-4 plus PD-L1/PD-1 | 6 (12.0) | 47 (24.5) | 16 (13.8) | |
| No. of extracutaneous irAEs, median (IQR) | 0 (0-1) | 1 (0-2) | 0 (0-1) | <.001 |
| cirAE features | ||||
| Time to cirAE, days, median (IQR) | 64 (21-190) | 42 (15-111) | 63 (21-188) | .05 |
| Duration, d, median (IQR) | 136 (32-380) | 55 (21-198) | 103 (29-329) | .04 |
| CTCAE v5.0 severity, median (IQR) | 3 (2-3) | 1 (1-1) | 1 (1-2) | <.001 |
| Required hospitalization | 10 (20.0) | 5 (2.6) | 1 (0.9) | <.001 |
| Referred to dermatology department | 30 (60.0) | 43 (22.4) | 35 (30.2) | <.001 |
| Morphologic type | ||||
| Maculopapular | 21 (42.0) | 102 (53.1) | 57 (49.1) | .048 |
| Isolated pruritus | 4 (8.0) | 39 (20.3) | 23 (19.8) | |
| Eczematous | 3 (6.0) | 11 (5.7) | 10 (8.6) | |
| Othere | 22 (44.0) | 40 (20.8) | 26 (22.4) | |
| ICI disrupted by cirAE | 28 (56.0) | 7 (3.6) | 8 (6.9) | <.001 |
| Non-SCS treatments for cirAE | ||||
| None | 0 | 67 (34.9) | 29 (25.0) | <.001 |
| Topical corticosteroid | 42 (84.0) | 83 (43.2) | 65 (56.0) | |
| Other topical therapyf | 3 (6.0) | 10 (5.2) | 7 (6.0) | |
| Phototherapy | 1 (2.0) | 1 (0.5) | 0 | |
| Oral antibiotic | 3 (6.0) | 1 (0.5) | 1 (0.9) | |
| Oral antipruritic | 30 (60.0) | 79 (41.2) | 44 (37.9) | |
| Targeted therapyg | 5 (10.0) | 0 | 0 | |
Abbreviations: cirAE, cutaneous immune-related adverse event; CTCAE v5.0, Common Terminology Criteria for Adverse Events, version 5.0; CTLA-4, cytotoxic T-lymphocyte–associated antigen 4; ECOG, Eastern Cooperative Oncology Group; GI, gastrointestinal; ICI, immune checkpoint inhibitor; irAE, immune-related adverse event; IQR, interquartile range; NSCLC, non–small cell lung cancer; PD-1, programmed cell death 1; PD-L1, programmed death ligand 1; SCS, systemic corticosteroids.
The SCS dosages for initial cirAE included low dose (n = 42 [84.0%]), moderate dose (n = 6 [12.0%]), and high dose (n = 2 [4.0%]).
Common noncutaneous irAEs included hepatitis/transaminitis, gastroenterocolitis, hypophysitis, pneumonitis, nephritis, and arthritis. The SCS dosages for noncutaneous irAEs included low dose (n = 31), moderate dose (n = 53), and high dose (n = 30). Common other medical reasons for SCS included polymyalgia rheumatica, rheumatoid arthritis, adrenal insufficiency, fevers, clinical decompensation, and brain metastases. The SCS dosages for other medical reasons included low dose (n = 82), moderate dose (n = 39), and high dose (n = 12).
P values reflect Fisher exact or Kruskal-Wallis tests.
Other cancer types included (overall No. [%]): genitourinary (n = 27 [7.6]), gynecologic or breast (n = 20 [5.6]), nonmelanoma skin cancer (n = 13 [3.6]), endocrine or neuroendocrine (n = 9 [2.5]), hematologic (n = 8 [2.2]), neurologic (n = 5 [1.4]), and sarcoma (n = 1 [0.3]).
Other cirAE types included (overall No. [%]): lichenoid (n = 14 [3.9]); psoriasiform (n = 12 [3.4]); vitiligo (n = 9 [2.5]); mucositis (n = 9 [2.5]); bullous (n = 7 [2.0]); acneiform (n = 5 [1.4]); erythema multiforme (n = 4 [1.1]); urticarial (n = 4 [1.1]); follicular (n = 3 [0.8]); panniculitis (n = 2 [0.6]); Stevens-Johnson syndrome (n = 1 [0.3]), Sweet syndrome (n = 1 [0.3]).
Other topical therapies included emollients, topical keratolytics, calcipotriene, topical antipruritics, topical antifungals, topical antibiotics, salt-water rinses (mucositis), and “magic mouthwash” (magnesium-aluminum-simethicone-diphenhydramine-lidocaine-nystatin compound, for mucositis).
Targeted therapies included methotrexate, mycophenolate mofetil, and oral calcineurin inhibitors.
Infection rates were higher in patients who received SCS for an initial cirAE compared with those with no SCS exposures for any indication (34.0% vs 19.8%).2,6 Most infections in both groups required systemic therapy (88.2% vs 95.7%), with a substantial proportion necessitating hospitalization (29.4% vs 43.5%) (Table 2). In multivariate models adjusted for age, sex, and SCS exposures by indication and dosage, patients who received SCS for an initial cirAE and those who did not had similar progression-free survival (hazard ratio, 0.7; 95% CI, 0.4-1.3; P = .29) and overall survival (hazard ratio, 3.0; 95% CI, 0.3-35.3; P = .38).
Table 2. Systemic Corticosteroid Infectious Complications.
| Infectious eventsa | No. (%) | P valuec | |
|---|---|---|---|
| SCS for initial cirAE (n = 50) | No SCS for any indication (n = 116)b | ||
| Overall infectionsd | 17 (34.0) | 23 (19.8) | .07 |
| Requiring treatment | 15 (30.0) | 22 (19.0) | .15 |
| Requiring hospitalization | 5 (10.0) | 10 (8.6) | .77 |
Abbreviations: cirAE, cutaneous immune-related adverse event; SCS, systemic corticosteroids.
For patients who received SCS for initial cirAE, infectious events were defined as any infections requiring medical intervention between the date of SCS initiation to 1 month after SCS completion. For patients who did not receive SCS for any indication, infectious events were defined as any infections requiring medical intervention from date of initial cirAE to 3 months after.
To isolate the association of SCS for initial cirAE with infection rates, infection rates among those with SCS for initial cirAE were compared solely with those with no SCS exposures for any indication.
P values reflect Fisher exact or Kruskal-Wallis tests.
Common infectious complications occurring during initial cirAE included (overall No.; SCS for initial cirAE No.; no SCS exposures for any indication No.): urinary (n = 11; n = 6; n = 6), upper respiratory tract (n = 9; n = 4; n = 5), gastrointestinal (n = 3; n = 0; n = 3), skin or mucosal (n = 15; n = 7; n = 9), and blood (n = 1; n = 0; n = 1).
Discussion
In this retrospective cohort study, we characterized SCS exposures in patients with a cirAE and SCS-associated complications for cirAE and examined the association between SCS for treatment of an initial cirAE and survival. Patients received SCS for a range of indications, most commonly receiving low-dose SCS for an initial cirAE. We observed higher rates of infection than previously reported among both patients who did and did not receive SCS for an initial cirAE.2,6 Despite the theoretical risk of SCS impeding the antitumor response, we found no association between SCS for an initial cirAE and progression-free survival or overall survival, corroborating prior findings.4 Collectively, these findings suggest that with appropriate management, low-dose SCS may be safely administered for cirAE without a significant association with survival outcomes.
Study limitations include a small sample size, retrospective design, and an analysis focused solely on use of SCS for an initial cirAE. These elements constrained our ability to examine the timeline of SCS delivery and may limit the generalizability of the findings.
eMethods. Study Design
References
- 1.Phillips GS, Wu J, Hellmann MD, et al. Treatment outcomes of immune-related cutaneous adverse events. J Clin Oncol. 2019;37(30):2746-2758. doi: 10.1200/JCO.18.02141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Thompson LL, Krasnow NA, Chang MS, et al. Patterns of cutaneous and noncutaneous immune-related adverse events among patients with advanced cancer. JAMA Dermatol. Published online March 24, 2021. doi: 10.1001/jamadermatol.2021.0326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Abu-Sbeih H, Tang T, Ali FS, et al. The impact of immune checkpoint inhibitor-related adverse events and their immunosuppressive treatment on patients’ outcomes. J Immunother Precis Oncol. 2018;1(1):7. doi: 10.4103/JIPO.JIPO_12_18 [DOI] [Google Scholar]
- 4.Horvat TZ, Adel NG, Dang T-O, et al. Immune-related adverse events, need for systemic immunosuppression, and effects on survival and time to treatment failure in patients with melanoma treated with ipilimumab at Memorial Sloan Kettering Cancer Center. J Clin Oncol. 2015;33(28):3193-3198. doi: 10.1200/JCO.2015.60.8448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Faje AT, Lawrence D, Flaherty K, et al. High-dose glucocorticoids for the treatment of ipilimumab-induced hypophysitis is associated with reduced survival in patients with melanoma. Cancer. 2018;124(18):3706-3714. doi: 10.1002/cncr.31629 [DOI] [PubMed] [Google Scholar]
- 6.Del Castillo M, Romero FA, Argüello E, Kyi C, Postow MA, Redelman-Sidi G. The spectrum of serious infections among patients receiving immune checkpoint blockade for the treatment of melanoma. Clin Infect Dis. 2016;63(11):1490-1493. doi: 10.1093/cid/ciw539 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Study Design