Immune checkpoint inhibitor-induced gastroenterocolitis (ICI-GEC) represents one of the common and severe side-effects of immunotherapy, impacting 7–30% of treated patients [1]. Current guidelines recommend systemic corticosteroids for moderate to severe cases of ICI-GEC, but whether these therapies have a dose- or time-dependent impact on patients’ survival outcomes remains unclear [2,3]. Murine data underscore the pleiotropic immunosuppressive effects of corticosteroids, but the clinical implications of these effects in the context of checkpoint blockade remain under-investigated [4]. Prior clinical investigations have been conflicting and limited, considering only selected cancer types and immune checkpoint inhibitor (ICI) regimens without integrative examination of corticosteroid exposures by magnitude or indication [1,3,5–7]. As ICI-GEC is one of the most common indications for corticosteroids, patients with this immunotoxicity are disproportionately impacted by the persistent lack of clinical consensus [1,3]. To enhance treatment approaches to ICI-GEC, we therefore sought to investigate the relationship between systemic corticosteroids for this indication and survival outcomes, while quantifying and accounting for all other corticosteroid exposures. We retrospectively reviewed the medical records of patients who underwent endoscopic evaluation for possible ICI-GEC at our institution between August 1, 2011 and April 31, 2019 (n = 111). Patients with celiac disease (n = 1) or microscopic colitis (n = 1) preceding immunotherapy initiation were excluded. Patients with pathology indicating non-ICI aetiology of gastroenterocolitis (GEC) (n Z 19) were also excluded, resulting in a final sample of ninety patients with biopsy-confirmed ICI-GEC. For this final sample, we abstracted demographics, oncologic history, ICI-GEC features, other immune-related adverse events, systemic corticosteroid exposures, immunosuppressive targeted therapy exposures, date of progression and date of death from the electronic medical record. Information documented before February 8, 2020 was considered. We categorised systemic corticosteroid exposures by indication (ICI-GEC, other immune-related adverse event, other medical reason) and established dose thresholds (‘high’, averaging 1 mg/kg prednisone equivalents/day for 1 week; ‘moderate’, averaging >7.5 mg prednisone equivalents/day for 2 months; ‘none or limited’, averaging 7.5 mg prednisone equivalents/day for 2 months) [2,6]. We then used Cox proportional hazards models to assess the relationship between systemic corticosteroids for initial episode of Table 1 Baseline demographic characteristics and survival outcomes (n Z 90). Characteristic No. (%)a Age, years e mean (SD)b 61 (12) Female sexb 34 (38) Cancer typeb Melanoma 68 (76) Non-small-cell lung cancer 10 (11) Otherc 12 (13) Cancer stage e median (IQR)b 4 (4e4) Brain metastasesb 16 (18) ECOG e median (IQR)b 0 (0e2) Pre-ICI treatments Radiation 29 (32) Traditional chemotherapy 16 (18) Targeted therapy 14 (16) Non-ICI immunotherapy 15 (17) ICI associated with ICI-GEC CTLA-4 39 (43) PD-1 or PD-L1 32 (36) CTLA-4 þ PD-L1/PD-1 19 (21) ICI-GEC features Distribution Macroscopic colitis enteritis, gastritis 63 (70) Microscopic colitis enteritis, gastritis 21 (23) Enteritis only gastritis 6 (7) CTCAE v5.0 enterocolitis severity grade e median (IQR) 2 (2e3) Symptom duration e days, median (IQR) 57.0 (34e80) Relapsed 20 (22) Number of non-GEC irAEs e median (IQR)g 1 (1e2) SCS for ICI-GEC e None or limited 19 (21) Moderate 30 (33) High 41 (46) Infliximab vedolizumabf 34 (38) SCS for non-GEC irAEse,h None or limited 66 (73) Moderate 16 (18) High 8 (9) SCS for other medical reasone,i,j None or limited 71 (79) Moderate 16 (18) High 3 (3) PFS e HR, 95% CI, P valuek,l Moderate SCS for ICI-GEC 1.28, 0.57e3.64, 0.59 High SCS for ICI-GEC 2.54, 1.11e5.80, 0.03 OS e HR, 95% CI, P valuek,l Moderate SCS for ICI-GEC 0.69, 0.22e2.07, 0.51 High SCS for ICI-GEC 1.62, 0.56e4.70, 0.37 95% CI, 95% confidence interval; AI, adrenal insufficiency; CTLA-4, cytotoxic T-lymphocyteeassociated protein 4; ECOG, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIGEC, immune checkpoint inhibitoreinduced gastroenterocolitis; CTCAE, Common Terminology Criteria for Adverse Events; IQR, interquartile range; irAE, immune-related adverse event; OS, overall survival; PD-1, programmed death 1; PDL-1, programmed death ligand 1; PFS, progression-free survival; SCS, systemic corticosteroid; ICI, immune checkpoint inhibitor. a Data are presented as the number (percentage) of patients unless otherwise indicated. b Age, sex, cancer stage, presence of brain metastases and Eastern Cooperative Oncology Group performance status were determined for the date of first ICI. c Including colorectal (3), non-Hodgkin lymphoma (2), ovarian (2), pancreatic (1), small-cell lung (1), renal-cell (1), breast (1) and nonmelanoma skin (1) cancers. d Relapse was defined as the return of diarrhoea or abdominal pain of grade I or more for the aforementioned patient, occurring >1 month after date of symptom resolution, meeting the following additional criteria: (1) requiring medical intervention with ICI-GEC directed therapy, (2) with negative infectious studies if performed and (3) supported by ongoing histologic evidence of ICI-GEC if the patient underwent second endoscopy. e SCS dose thresholds were ‘high’, on average 1 mg/kg/day for 1 week; ‘moderate’, on average > 7.5 mg/day for 2 months; ‘none or limited’, on average 7.5 mg/day for 2 months. f Patients received a median of 2 (IQR Z 1e3) doses of infliximab. Two patients receiving infliximab also received vedolizumab (range, 3e4 doses). g Common irAE types occurring after ICI start included dermatitis or other cutaneous irAEs (55), hepatotoxicity (15), hypopituitarism (7), pneumonitis (7), inflammatory arthritis (7) and thyroiditis (5). h Common indications for SCS for non-GEC irAEs included dermatitis or other cutaneous irAEs (7), hepatotoxicity (5), pneumonitis (5), arthritis (3), acute inflammatory phase of hypophysitis (3), nephritis (2), neurologic irAE (2) and hemophagocytic lymphohistiocytosis (1). Three patients had multiple irAEs independently meeting moderate or high-dose SCS thresholds. i Common medical reasons for SCS included brain metastases (7), management of adverse reaction to non-ICI drug (6) and pre-existing autoimmune disease (3). j 10 patients received SCS for AI; 4 at moderate doses. Two of these patients separately received SCS for another medical reason at moderate to high doses, 2 did not. k Cox regression model adjusted for age, sex, cancer type, cancer stagey, baseline brain metastases, Eastern Cooperative Oncology Group performance status, non-ICI immunotherapy before ICI-start, other irAE, SCS for non-EC irAE, SCS for other medical reason, lifetime doses of PD-1/PDL-1y and lifetime doses of CTLA-4. Covariates with an asterisk were statistically significant to p < 0.10 in analyses of progression-free survival, and those with a dagger symbol were statistically significant to this threshold in analyses of overall survival. l The index category for the Cox regressions was ‘none or limited’ dose of SCS for the first episode of ICI-GEC, considering progression-free survival (PFS) as a primary outcome and overall survival (OS) as a secondary outcome. Cox proportional hazard models were adjusted for age, sex, cancer type, cancer stage, baseline brain metastases, pre-ICI immunotherapy, lifetime ICI doses and systemic corticosteroid exposures from ICI start to death [6]. Further details regarding data collection and statistical analyses are provided elsewhere. The Partners Institutional Review Board approved this study (Protocol No. 2017P000501). Ninety patients (mean age = 61 years, 38% women) developed ICI-GEC corroborated by biopsy at our institution between August 1, 2011 and April 31, 2019 (Table 1). Patients most commonly had advanced melanoma (n Z 68, 76%) or non-small-cell lung cancer (n = 10, 11%) and were predominantly receiving cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) (n = 39, 43%) or PD-1/programmed death ligand 1 monotherapy (n = 32, 36%) at the time of ICI-GEC emergence. ICI-GEC symptoms appeared a median of 53 days (interquartile range [IQR], 30–98) after ICI initiation, peaking at a median Common Terminology Criteria for Adverse Events severity grade of II (IQR, 2–3) [3]. Nearly all (n = 71, 80%) patients required systemic corticosteroids for their initial episode of ICIGEC, many (n = 41, 46%) at doses exceeding the highest threshold. Steroid-refractory ICI-GEC was common (n = 48, 54%), necessitating escalation to infliximab or vedolizumab in more than a third (n = 34, 38%) of cases. Patients experienced ICI-GEC symptoms for nearly two months (median 57 days, IQR 34–80), with approximately a quarter (n = 20, 22%) experiencing symptom relapse after resolution of the first episode. ICI-GEC posed substantial disruption to planned treatment regimens, precipitating ICI suspension in most patients (n = 75, 83%), with subsequent discontinuation of all immunotherapy in two-thirds (n = 60, 67%) of individuals. Most patients (n = 69, 77%) experienced other immune-related adverse events in addition to ICI-GEC, with a quarter (n = 25, 36%) requiring systemic corticosteroids for this indication. Some patients (n = 19, 21%) also received steroids for other medical reasons, most commonly for treatment of brain metastases (n = 7), adverse drug reactions (n = 5), or pre-existing autoimmune disease (n = 3). In Cox proportional hazard models accounting for these corticosteroid exposures and other covariates, patients who received high-dose systemic corticosteroids for their first episode of ICI-GEC had reduced PFS (hazard ratio [HR] = 2.54, 95% confidence interval [CI]:1.11–5.80, P = 0.03) compared with those who did not (Table 1). OS was not affected by systemic corticosteroid therapy for ICI-GEC (HR = 1.62, 95% CI: 0.56–4.70, P = 0.37). In this study, we characterised relationships between systemic corticosteroids for ICI-GEC and survival outcomes, providing novel integration of other corticosteroid exposures by magnitude and indication. In contrast with several prior studies suggesting no relationship between systemic corticosteroids and survival, we found that receipt of high-dose steroids for the initial ICI-GEC episode was associated with decreased PFS when all steroid exposures were considered [3,5,6]. This divergence may be due to limited prior evaluation of corticosteroid exposures by indication, as available evidence suggests independent associations between some indications (i.e. brain metastases, certain immune-related adverse events) and PFS [3]. Observed differences could also reflect a dose- or duration-dependent impact of immunosuppression on the anti-tumour response, only identifiable through multidimensional analysis [4]. Collectively, our findings underscore nuanced relationships between systemic corticosteroids for ICI-GEC and survival and suggest potential merits of steroid-sparing approaches. Relationships between the full array of systemic corticosteroid exposures and survival outcomes warrant further investigation for other immune-related adverse events.
Table 1.
Baseline demographic characteristics and survival outcomes (n = 90)
| Characteristic | No. (%)a |
|---|---|
| Age, years - mean (SD)b | 61 (12) |
| Female sexb | 34 (38) |
| Cancer typeb | |
| Melanoma | 68 (76) |
| Non-small-cell lung cancer | 10 (11) |
| Otherc | 12 (13) |
| Cancer stage - median (IQR)b | 4 (4–4) |
| Brain metastasesb | 16 (18) |
| ECOG - median (IQR)b | 0 (0–2) |
| Pre-ICI treatments | |
| Radiation | 29 (32) |
| Traditional chemotherapy | 16 (18) |
| Targeted therapy | 14 (16) |
| Non-ICI immunotherapy | 15 (17) |
| ICI associated with ICI-GEC | |
| CTLA-4 | 39 (43) |
| PD-1 or PD-L1 | 32 (36) |
| CTLA-4 + PD-L1/PD-1 | 19 (21) |
| ICI-GEC features | |
| Distribution Macroscopic | |
| colitis ± enteritis, gastritis | 63 (70) |
| Microscopic colitis ± enteritis, gastritis | 21 (23) |
| Enteritis only ± gastritis | 6 (7) |
| CTCAE v5.0 enterocolitis severity grade - median (IQR) | 2 (2–3) |
| Symptom duration - days, median (IQR) | 57.0 (34–80) |
| Relapsed | 20 (22) |
| Number of non-GEC irAEs - median (IQR)g | 1 (1–2) |
| SCS for ICI-GECe | |
| None or limited | 19 (21) |
| Moderate | 30 (33) |
| High | 41 (46) |
| Infliximab ± vedolizumabf | 34 (38) |
| SCS for non-GEC irAEse,h | |
| None or limited | 66 (73) |
| Moderate | 16 (18) |
| High | 8 (9) |
| SCS for other medical reasone,i,j | |
| None or limited | 71 (79) |
| Moderate | 16 (18) |
| High | 3 (3) |
| PFS e HR, 95% CI, P valuek,l | |
| Moderate SCS for ICI-GEC | 1.28, 0.57–3.64, 0.59 |
| High SCS for ICI-GEC | 2.54, 1.11–5.80, 0.03 |
| OS e HR, 95% CI, P valuek,l | |
| Moderate SCS for ICI-GEC | 0.69, 0.22–2.07, 0.51 |
| High SCS for ICI-GEC | 1.62, 0.56–4.70, 0.37 |
95% CI, 95% confidence interval; AI, adrenal insufficiency; CTLA-4, cytotoxic T-lymphocyteeassociated protein 4; ECOG, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIGEC, immune checkpoint inhibitoreinduced gastroenterocolitis; CTCAE, Common Terminology Criteria for Adverse Events; IQR, interquartile range; irAE, immune-related adverse event; OS, overall survival; PD-1, programmed death 1; PDL-1, programmed death ligand 1; PFS, progression-free survival; SCS, systemic corticosteroid; ICI, immune checkpoint inhibitor.
Data are presented as the number (percentage) of patients unless otherwise indicated.
Age, sex, cancer stage, presence of brain metastases and Eastern Cooperative Oncology Group performance status were determined for the date of first ICI.
Including colorectal (3), non-Hodgkin lymphoma (2), ovarian (2), pancreatic (1), small-cell lung (1), renal-cell (1), breast (1) and nonmelanoma skin (1) cancers.
Relapse was defined as the return of diarrhoea or abdominal pain of grade I or more for the aforementioned patient, occurring >1 month after date of symptom resolution, meeting the following additional criteria: (1) requiring medical intervention with ICI-GEC directed therapy, (2) with negative infectious studies if performed and (3) supported by ongoing histologic evidence of ICI-GEC if the patient underwent second endoscopy.
SCS dose thresholds were ‘high’, on average 1 mg/kg/day for 1 week; ‘moderate’, on average > 7.5 mg/day for 2 months; ‘none or limited’, on average 7.5 mg/day for 2 months.
Patients received a median of 2 (IQR Z 1e3) doses of infliximab. Two patients receiving infliximab also received vedolizumab (range, 3e4 doses).
Common irAE types occurring after ICI start included dermatitis or other cutaneous irAEs (55), hepatotoxicity (15), hypopituitarism (7), pneumonitis (7), inflammatory arthritis (7) and thyroiditis (5).
Common indications for SCS for non-GEC irAEs included dermatitis or other cutaneous irAEs (7), hepatotoxicity (5), pneumonitis (5), arthritis (3), acute inflammatory phase of hypophysitis (3), nephritis (2), neurologic irAE (2) and hemophagocytic lymphohistiocytosis (1). Three patients had multiple irAEs independently meeting moderate or high-dose SCS thresholds.
Common medical reasons for SCS included brain metastases (7), management of adverse reaction to non-ICI drug (6) and pre-existing autoimmune disease (3).
10 patients received SCS for AI; 4 at moderate doses. Two of these patients separately received SCS for another medical reason at moderate to high doses, 2 did not.
Cox regression model adjusted for agey, sex, cancer type, cancer stagey, baseline brain metastases, Eastern Cooperative Oncology Group performance status, non-ICI immunotherapy before ICI-start, other irAE, SCS for non-EC irAE, SCS for other medical reason, lifetime doses of PD-1/PDL-1y and lifetime doses of CTLA-4. Covariates with an asterisk were statistically significant to p < 0.10 in analyses of progression-free survival, and those with a dagger symbol were statistically significant to this threshold in analyses of overall survival.
The index category for the Cox regressions was ‘none or limited’ dose of SCS for the first episode of ICI-GEC.
References
- [1].Collins M, Soularue E, Marthey L, Carbonnel F. Management of patients with immune checkpoint inhibitor-induced enterocolitis: a systematic review. Clin Gastroenterol Hepatol 2020. 10.1016/J.CGH.2020.01.033. Published online January 31. [DOI] [PubMed] [Google Scholar]
- [2].Faje AT, Lawrence D, Flaherty K, Freedman C, Fadden R, Rubin 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):3706e14. 10.1002/cncr.31629. [DOI] [PubMed] [Google Scholar]
- [3].Thompson JA, Schneider BJ, Brahmer J, Andrews S, Armand P, Bhatia S, et al. Management of immunotherapy-related toxicities, version 1.2019, NCCN clinical practice guidelines in Oncology. J Natl Compr Canc Netw 2019;17(3):255e89. 10.6004/jnccn.2019.0013. [DOI] [PubMed] [Google Scholar]
- [4].Cain DW, Cidlowski JA. Immune regulation by glucocorticoids. Nat Rev Immunol 2017;17(4):233e47. 10.1038/nri.2017.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Das S, Johnson DB. Immune-related adverse events and antitumor efficacy of immune checkpoint inhibitors. J Immunother Cancer 2019;7(1):306. 10.1186/s40425-019-0805-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Horvat TZ, Adel NG, Dang T-O, Momtaz P, Postow MA, Callahan MK, 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):3193e8. 10.1200/JCO.2015.60.8448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Wang Y, Abu-Sbeih H, Mao E, Ali N, Shaukat Ali F, Qiao W, et al. Immune-checkpoint inhibitor-induced diarrhea and colitis in patients with advanced malignancies: retrospective review at MD Anderson. J Immunother Cancer 2018;6(1):37. 10.1186/s40425-018-0346-6. [DOI] [PMC free article] [PubMed] [Google Scholar]