Certolizumab induced sarcoidosis in a patient with axial spondyloarthritis

Abstract

Introduction: Anti-TNF-alpha monoclonal antibody therapy (TNF-i) is an effective treatment for axial spondyloarthritis (AxSpA) with many patients achieving successful remission. Like other immunosuppression, the main risk to patients is infection but sarcoidosis is a rare complication associated with TNF-i. This report describes a patient with well controlled AxSpA on certolizumab who developed sarcoidosis with lymphadenopathy and CNS involvement. The mechanism is not fully understood, especially as TNF-i is an effective treatment for sarcoidosis. A number of pathways are likely involved. This case highlights the importance of identifying this paradoxical reaction, as well as the challenges in ongoing treatment decisions. Case description: We describe a case of a 37-year-old female patient with axial spondyloarthritis and associated recurrent uveitis, diagnosed in 2019. She also has a history of audio-spatial migraines. Her disease was in remission on certolizumab (ASDAS-CRP 1.7) when she presented to hospital with dysphasia, confusion, nausea, vomiting and bilateral frontal headache. Examination revealed bilateral 6th nerve palsies. Fundoscopy showed bilateral papilledema with small peripapillary haemorrhages. She was initially treated for suspected CNS infection; lumbar puncture was attempted several times but was unsuccessful. ANA, dsDNA, PR3, MPO, NMDA-R, CASPR-2, LGI-1, AMPA-1, AMPA-2, GABA-B receptor, hepatitis B and C serology, HIV, syphilis and ACE were all normal or negative. CT venogram and MRI head with contrast were unremarkable, however MRI orbits demonstrated mild flattening of the posterior sclerae and minor prominence of the subarachnoid optic nerves, consistent with idiopathic intracranial hypertension (IIH). EEG was consistent with mild non-specific focal cerebral dysfunction. CT chest, abdomen and pelvis showed lymphadenopathy above and below the diaphragm, confirmed on CT PET. An iliac lymph node biopsy revealed non-necrotising granulomatous lymphadenitis consistent with sarcoidosis. Specialist neurology opinion felt that CNS sarcoid was probable based on the clinical and radiological findings, +/- a diagnosis of IIH. She was treated with acetazolamide and certolizumab was held. When seen in clinic, her neurological symptoms had resolved but she had a flare of AxSpA (ASDAS-CRP 3). She was still family planning so it was felt that a trial of a different TNF-i would be the safest option (rather than an IL-17 inhibitor or a JAKi), and she was initiated on intravenous infliximab. She was reviewed after three infliximab infusions with some improvement (ASDAS-CRP 1.9), fatigue being the predominant symptom but no recurrence of confusion or IIH symptoms. She continues on infliximab and has been referred to the neuro-inflammatory clinic. Discussion(s): Whether this patient had co-existing sarcoidosis and AXSpA, or TNF-i induced sarcoidosis is not definite. Importantly, TNF-i have been showed to be effective treatment for both AxSpA and sarcoidosis, owing to TNF involvement in granuloma formation. Although also rare, there have been more reported cases of TNF-i induced sarcoidosis (a so-called paradoxical reaction) than of co-existent AxSpA and sarcoidosis. The mechanism is poorly understood. A possible explanation is that the blocking of TNF-alpha associated immune pathways leads to up-regulation of other pathways causing granuloma formation. Withdrawal of treatment, as was done in this case, leads to improvement of sarcoid related symptoms. This case presented a diagnostic dilemma in the acute presentation, further complicated by the absence of CNS sampling. Tissue diagnosis was made from lymph node biopsy, but CNS involvement remains speculative, although the patient did improve with cessation of TNF-i. Isolated neurosarcoidosis is rare and so a whole body CT PET scan was pertinent in the work up and indeed enabled the diagnosis. This case is of further interest as the patient was restarted on a different TNF-i subsequently as this was felt to be the safest and most effective treatment for her. So far, there has not been any clinical recurrence of sarcoid-like disease. The risk of original occurrence and of recurrence for this patient is unknown, and not likely related to the length of exposure to TNF-i. This patient had been on certolizumab for three years before her acute presentation with sarcoidosis. In a case series describing RA patients who developed TNF-i induced sarcoidosis, time of exposure to TNF-i before the development of sarcoidosis ranged from nine months to ten years. This uncertainty certainly contributes to the complexity of counselling for these patients, whose disease is well controlled on TNF-i and where other options may be limited. Key learning points: Our patient cohort can present with various complications related to their underlying disease or their immunosuppressive treatment. Presentations with neurological symptoms on immunosuppression do warrant a high-index of suspicion for CNS infection (as was the case in this patient). However, the paradoxical reaction of TNF-i induced sarcoidosis is a key learning point from this case, as patients can present with a constellation of symptoms affecting different organ systems, and diagnosis can be challenging. Etanercept has been found to be most associated with the phenomenon followed by infliximab and adalimumab. Certolizumab has been reported as a rarer culprit, and mostly with single organ involvement. There are no proven mechanisms, but etanercept induced sarcoidosis could be explained by its limited inhibition of TNF-alpha (only targeting soluble TNF-alpha), which leads to redistribution in low concentrated regions such as the lungs. Another explanation relates to its role in activated interferon gamma production, contributing to granuloma formation. The mechanism in certolizumab may be due to the absence of the Fc portion, limiting its immunogenic effect. Another theory proposed is that neutralizing antibodies to certolizumab may contribute to granuloma formation via insufficient blockage of TNF-alpha. In the literature, these paradoxical reactions often exhibit pulmonary involvement - where patients are exposed to exogenous antigens - and it is important to consider that patients on TNF-i can also develop sarcoid-like disease triggered by infections. Making a diagnosis of CNS sarcoidosis can be difficult even with a full complement of investigations and often requires specialist neurology input. This case also involved a diagnosis of possible IIH, a condition which can be associated with neurosarcoidosis. Treatment involves withdrawal of TNF-i and management of flares with methotrexate and corticosteroids. Decisions about biologic switch (including trial of a different TNF-i) require careful consideration and counselling, taking into account disease control and risk of recurrence.