Combination anti-PD-1 and ipilimumab (ipi) therapy in patients with advanced melanoma and pre-existing autoimmune disorders (AD).

10026 Background: Clinical trials of immunotherapy exclude patients (pts) with pre-existing AD. While retrospective data exist regarding the efficacy and safety of single agent ipi and anti-PD1 antibodies (PD1) in pts with AD, no data are available regarding the safety and efficacy of combination therapy in pts with AD, which has a higher toxicity risk. Methods: Pts with melanoma and pre-existing AD treated with combination ipi/PD1 were retrospectively identified from 10 international centres. Data regarding AD, treatment, toxicity and outcomes were examined. Results: Fifty-five pts were included, 46 were treated with ipi/nivolumab and 9 with ipi/pembrolizumab. 40 had an ipi dose of 3mg/kg while 15 had a lower dose regimen. 9 pts received prior PD1 therapy; 3 suffered moderate immune-related adverse events (irAE) with no flares of AD on single agent PD1. Pre-existing AD included inflammatory bowel disease (IBD), thyroiditis, rheumatoid arthritis (RA), multiple sclerosis and psoriasis. 10 pts had active symptoms of AD and 13 were immunosuppressed at commencement of ipi/PD1. Eighteen pts (33%) experienced a flare of their AD including 4/7 with RA, 3/6 with psoriasis, 5/9 with IBD, 3/18 with thyroiditis, 1/1 with Sjogren’s syndrome, 1/1 with polymyalgia, 1/1 with Behcet’s syndrome. Median time to flare was 19 days (range 4 – 167). 13 pts were managed with steroids, 5 required additional immunosuppressants. 7 pts were hospitalised for management of flare (5 with IBD, 2 with RA). 2 pts required intensive care and vasopressors for severe IBD flare, quiescent prior to ipi/PD1. One for diarrhoea and shock and one for duodenal perforation. 8 pts ceased treatment due to flare (3 with IBD, 2 with RA, 1 with Behcet’s, 1 with Sjogren’s). Thirty-seven pts (67%), experienced an irAE unrelated to their AD, 38% G3 or G4. The most frequent irAEs were colitis (n = 16), hepatitis (n = 12), endocrinopathies (n = 12), with 13 pts experiencing an irAE in ≥ 2 organs. 9 pts experienced both AD flare and an irAE. 20 pts (36%) ceased immunotherapy due to irAEs. ORR was 55% (54% in PD1 naive pts), at a median follow up of 14 months, 77% of responses ongoing. ORR in pts who had a flare of their AD was 44% and in pts on immunosuppression was 46%. Median PFS was shorter in pts who had a flare of AD compared with those who did not (2.6 vs 9 months; P-value 0.047). Conclusions: Combination ipi/PD1 shows efficacy comparable to clinical trial populations in pts with pre-existing AD and advanced melanoma. Whilst there was a substantial risk of flare of AD, no increased frequency of irAE’s was observed.