PWE-048 The nature of checkpoint inhibitor-associated lower gastrointestinal toxicity: a single cancer centre experience. (8th June 2018)

• Record Type: Journal Article
• Authors: Samani, Amit; Tippu, Zayd; Zhang, Shuai; Merrick, Sophie; Ibraheim, Hajir; Powell, Nicholas; Josephs, Debra; Papa, Sophie
• Is Part Of: Gut. Volume 67(2018)Supplement 1
• Page Start: A91
• Page End: A91
• Publication Date: 2018-06-08
• Subjects: Gastroenterology -- Periodicals; 616.33
• DOI: 10.1136/gutjnl-2018-BSGAbstracts.180
• Languages: English
• ISSNs: 0017-5749

Abstract

Abstract : Introduction: Immune checkpoint inhibitors (CPi) are a standard of care for various cancers however can cause unpredictable toxicity, notably of the lower gastrointestinal (GI) tract. As continued expansion of their use is anticipated, we sought to investigate the 'real world' frequency and characteristics of such toxicity. Methods: We conducted a retrospective case-note review of patients treated with ≥one dose of either nivolumab monotherapy (N), pembrolizumab monotherapy (P), sequential pembrolizumab/ipilimumab (P-I) or combination nivolumab +ipilimumab (n+I) between 01/03/2016 – 28/02/2017. Toxicity (tox) was graded using NCI Common Terminology Criteria for Adverse Events v4.03. Statistical comparisons used Student's T-test or Chi Squared tests with Bonferroni corrections. Results: 141 patients were treated (49 female, 34.8%), mean age of 64.6 years. 71, 33 and 37 (50.3%, 23.4% and 26.2%) had melanoma, lung and renal cancers respectively. 83 received P (58.9%) while 39 (27.7%) received N with only 11 (7.8%) and 8 (5.7%) receiving n+I or P-I respectively. Of these, 29 had any grade GI tox. Their mean age (65.7 years), female proportion (34.4%) and proportion with melanoma, lung or renal cancers (55.2%, 20.1% and 24.1% respectively) did not significantly differ from the 112 patients with no GI tox. However, 6 (20.6%) vs 2 (1.8%) received P-I and 5 (17.2%) vs 6 (5.4%) received n+I in the GI and no-GI tox cohorts respectively which was significantly different (p=0.001 and 0.036 respectively). Only 2 (7.1%) in the GI tox cohort had initial progressive disease compared to 47 (46.1%, p<0.001). In the GI tox cohort, 17 had grade 1 (G1) tox, 5 G2 and 7 G3/4. When comparing the G3/4 to G1 cohort there was a trend towards younger age (54.3 vs 67.7 years), treatment with P-I or n+I (71.5 vs 17.8%) and higher CRP (100 vs 62 mg/L). The rate of G3/4 toxicity was similar to the reported literature. For n+I, 5/11 (45.5%) had any grade GI tox with 3/11 (27.3%) having grade 3/4. Finally, of those with G3/4 tox, 4 had documented fever at presentation, only 1 had haematochezia and all had diarrhoea of ≥5 x/day. All required IV methylprednisolone but 3 received alternate routes initially (2 oral, 1 rectal). Three patients required infliximab with only 1 requiring ≥1 dose. Conclusions: A significant proportion of patients develop lower GI tox with CPi. They were more likely to have been treated with ipilimumab containing regimens and to have a favourable response to therapy although this may have been confounded by the higher use of n+I in this cohort. The rate of G3/4 toxicity was similar to the literature. Some patients are initially treated with oral/rectal steroid; they may benefit from earlier IV therapy. Infliximab as a single dose is usually effective.