High proportion of drug hypersensitivity reactions to sulfasalazine following its use in anti-PD-1-associated inflammatory arthritis

Sulfasalazine use in anti-PD-1-associated inflammatory arthritis may be associated with more frequent hypersensitivity reactions. Sir, Checkpoint blockade has revolutionized the management of several cancers by targeting the co-inhibitory pathways that down-modulate T cell receptor–mediated T cell activation. However, the resulting enhanced T cell activation also gives rise to immune-related adverse events (IrAEs). Arthritis is an increasingly recognized IrAE following treatment with the licensed antibodies to programmed cell death protein-1 (PD-1) and may persist following discontinuation of anti-PD-1 [1]. Treatment often consists of corticosteroids, with escalation if required to DMARDs. However, uncertainty exists about the possible impact of potent DMARD therapy on the durability of cancer response, with some oncologists reluctant to consider the use of MTX [2]. SSZ combines salicylic acid with the antimicrobial sulfapyridine and has been suggested as one possible therapy for IrAEs [1, 2]. SSZ can cause a drug hypersensitivity reaction that rarely may be systemic and potentially life-threatening [3]. Importantly, such drug hypersensitivity reactions are T cell mediated [4, 5]. We reviewed all cases of anti-PD-1-associated inflammatory arthritis treated with SSZ at our centre. Patient characteristics for the four patients are presented in Table 1. All had metastatic melanoma treated with the anti-PD-1 antibody pembrolizumab and three had prior checkpoint blockade with the anti-CTLA4 antibody ipilimumab. Patient characteristics 5 11 9 21 (concomitant steroid) Fever, cough, elevated CRP Fever, elevated CRP Fever, non-productive cough, elevated CRP Nausea, diarrhoea, elevated ALT and CRP Discontinuation Discontinuation Discontinuation and oral corticosteroid Discontinuation and oral corticosteroid 5 11 9 21 (concomitant steroid) Fever, cough, elevated CRP Fever, elevated CRP Fever, non-productive cough, elevated CRP Nausea, diarrhoea, elevated ALT and CRP Discontinuation Discontinuation Discontinuation and oral corticosteroid Discontinuation and oral corticosteroid All cases were anti-CCP antibody and RF negative. aPredated introduction of SSZ. ALT: alanine aminotransferase. Patient characteristics 5 11 9 21 (concomitant steroid) Fever, cough, elevated CRP Fever, elevated CRP Fever, non-productive cough, elevated CRP Nausea, diarrhoea, elevated ALT and CRP Discontinuation Discontinuation Discontinuation and oral corticosteroid Discontinuation and oral corticosteroid 5 11 9 21 (concomitant steroid) Fever, cough, elevated CRP Fever, elevated CRP Fever, non-productive cough, elevated CRP Nausea, diarrhoea, elevated ALT and CRP Discontinuation Discontinuation Discontinuation and oral corticosteroid Discontinuation and oral corticosteroid All cases were anti-CCP antibody and RF negative. aPredated introduction of SSZ. ALT: alanine aminotransferase. In patient 1, the initial management of knee and then ankle monoarthritis was with intra-articular steroid, but SSZ was introduced after progression to involve both knees and ankles. Ten days later the patient was admitted with pyrexia (37.7°C) and a widespread erythematous maculopapular rash. Blood tests revealed an acute increase in alanine aminotransferase (66 U/l) and CRP (189 mg/l) and new lymphopenia (0.8 × 109/l). He improved with 3 days of i.v. methylprednisolone followed by a reducing course of oral prednisolone. SSZ and pembrolizumab were both discontinued. In patient 2, SSZ was commenced for initial monoarthritis of the knee following only transient improvement with intra-articular steroid. Five days after starting SSZ, the patient was admitted with fever (38.5°C) and cough with an unremarkable chest radiograph and an elevated CRP (78 mg/l). He was treated with antibiotics for a presumed infection of unknown origin. SSZ was suspended for 1 week and future pembrolizumab infusions were deferred. Eleven days after restarting SSZ he was readmitted with fever and again treated for presumed infection of unknown origin and SSZ was stopped. One month later, SSZ was restarted, but 9 days later he was admitted with low-grade fever, non-productive cough and elevated CRP (170 mg/l). No source of infection was identified and he was treated with oral prednisolone and SSZ was stopped. Five weeks later, an increasing dose of SSZ was reintroduced, but upon a reduction in prednisolone he developed nausea, diarrhoea, deranged liver function tests (alanine aminotransferase 288 U/l) and elevated CRP. SSZ was stopped and his symptoms resolved. Patient 3 had rash, facial swelling and fever but did not seek medical attention, as her symptoms settled on stopping SSZ. Pembrolizumab was suspended in patient 4 due to the development of a symmetrical polyarthritis involving the shoulders, knees and hands. Four months after the final pembrolizumab infusion, an increasing dose of SSZ was introduced and 5 weeks later he was admitted with a 5 day history of profound loose stool successfully treated with high-dose steroids. SSZ was discontinued. We observed a high proportion of adverse reactions with SSZ in the context of anti-PD-1-associated inflammatory arthritis and SSZ was discontinued in all cases. Drug hypersensitivity reactions are mediated by drug-reactive T lymphocytes [5], and the risk may increase with higher drug levels or with an alteration in T cell function, as exemplified by the association of amoxicillin–clavulanate hypersensitivity with the PTPN22 locus [6], which encodes a regulator of T cell function. It is therefore plausible that inhibition of PD-1 might also increase risk and this might also apply to other drugs strongly associated with T cell–mediated drug hypersensitivity syndromes, such as carbamazapine and allopurinol. In RA clinical trials, the withdrawal rate for SSZ was as high as 25%, but this was mostly due to the common side effects of nausea, headache and dizziness, with a drug rash occurring in only 4–5% and other hypersensitivity manifestations being much rarer [3]. Cutaneous and other IrAEs are common with checkpoint blockade and could give rise to diagnostic uncertainty, as illustrated in patient 2. However, the temporal association, not observed in our experience with MTX, the high prevalence and similarity to the recognized SSZ hypersensitivity syndrome and the reoccurrence upon rechallenge all point towards a genuine drug association. Patient 4 was different and characterized by colitis, but in this case pembrolizumab was discontinued 4 months before SSZ was started. Colitis is a common IrAE following checkpoint blockade and is rare following SSZ, suggesting this is unrelated. However, recent data indicate that the gut microbiome can influence the emergence of IrAEs following checkpoint blockade [7], and since SSZ has an antibiotic component, we cannot exclude an indirect association through alteration of faecal flora [8]. In conclusion, we observed a high proportion of hypersensitivity reactions following the use of SSZ and suggest caution in its use to manage anti-PD-1-related inflammatory arthritis. B.A.F. has received support from the National Institute for Health Research (NIHR) Birmingham Biomedical Research Centre and the NIHR/Wellcome Trust Birmingham Clinical Research Facility. The views expressed in this publication are those of the authors and not necessarily those of the National Health Service, the NIHR or the Department of Health. A.F. is supported by the Medical Research Council–Arthritis Research UK Centre for Musculoskeletal Ageing Research and the Kennedy Trust–funded Arthritis Therapy Acceleration Programme. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article. Disclosure statement: B.A.F. has received consultancy fees from Novartis, Roche and Bristol-Myers Squibb. A.F. has received consultancy fees from Novartis and Chugai. All other authors have declared no conflicts of interest.