Background Immune checkpoint inhibitors (specifically the anti-CTLA-4 antibody ipilimumab, and the anti-PD-1 antibodies nivolumab and pembrolizumab, have revolutionized the treatment of advanced malignancies. However, by virtue of their mechanism of action – that is, loss of T-cell inhibition and impaired self-tolerance, patients with underlying autoimmune and rheumatologic diseases were typically excluded from the clinical trials leading to the approval of these agents. Limited data, typically of single checkpoint inhibitors, exist for their safety in patients with autoimmune diseases. However, the risk for rheumatologic disease flare in patients exposed to checkpoint inhibitors is unknown.

Objectives To determine the risk of rheumatologic disease flare in patients receiving checkpoint inhibitor therapy.

Methods We retrospectively studied all patients who had received a checkpoint inhibitor (i.e. ipilimumab, nivolumab, pembrolizumab, or any combination thereof) for any malignancy on the Mayo Clinic Rochester, Minnesota campus between January 1st, 2011 and May 16th, 2016 (approximately 5,200 patients.) Of these patients, we identified those with preexisting rheumatologic disease according to specific diagnostic codes.

Results Of the 16 patients identified (13 [81%] female; median [range] age, 68.5 [34–86]y), 6 had inflammatory arthritis, 3 had polymyalgia rheumatica, 2 had lupus, 2 had Sjogren's syndrome, 1 had temporal arteritis, 1 had IBD-associated spondyloarthropathy and 1 had gout. Patients were treated for the following metastatic/advanced malignancies: melanoma (9 [56%]), lung (5 [31%]) and lymphoma (2 [12%]) with; ipilimumab (4 [25%]), nivolumab (7 [44%]), pembrolizumab (5 [31.2%]) and ipilimumab/nivolumab (1 [6%]). Notably, in all cases, checkpoint inhibitor therapy was offered after failure of numerous other chemotherapies. Ten (62.5%) patients were on immunosuppressive therapy (mainly low dose prednisone and methotrexate) at the time of cancer diagnosis, and the majority (10 [62.5%]) were well-controlled/in remission from the standpoint of their rheumatologic disease. Three (19%) patients had flares of their rheumatologic disease after treatment for the following cancers as noted: Patient 1 - temporal arteritis – nivolumab - non-small cell lung cancer, Patient 2 - Sjogren's (severe sicca)– pembrolizumab - melanoma, and Patient 3 - spondylitis-ipilimumab/pembrolizumab combination –melanoma. All flares responded to steroids or supportive therapy. Checkpoint inhibitor therapy was discontinued in all three patients for the following reasons: Patient 1 – flare of temporal arteritis, Patients 2 & 3 – cancer progression.

Conclusions To our knowledge, we have identified the largest cohort (16) of patients from a single academic with preexisting rheumatologic disease who had been exposed to checkpoint inhibitor therapy for advanced malignancy. Of these patients, only a minority experienced a flare of their disease during cancer treatment, and responded to standard therapies. With close monitoring and in the appropriate clinical context, checkpoint inhibitor therapies should be considered and made available to patients with preexisting rheumatologic disease who develop advanced malignancies.

Disclosure of Interest None declared