Abstract
Introduction
Auto-immune and auto-inflammatory disorders are believed to cause approximately 20% of cases of pyrexia of unknown origin. Rheumatological opinion is often sought when an infectious source has not been detected. Assessment of recurrent fever is challenging for fear of initiating immunosuppression in the presence of undetected infection. This challenge is even greater in patients with a previous history of auto-immune or infectious disorders.
Here, we discuss the investigation and management of a challenging case of recurrent fever, ultimately diagnosed as adult-onset Still’s disease, complicated by the previous occurrence of pulmonary tuberculosis and myasthenia gravis.
Case description
Our 34-year-old female patient, originally from India, had lived in the UK for 8 years. Her background included previous treatment for pulmonary TB in 2011 and myasthenia gravis diagnosed in 2015 with subsequent thoracoscopic thymectomy in 2016.
She was admitted to the infectious diseases unit in October 2016 with 3 weeks of recurrent fever and an itchy rash which had commenced 1-week after holidaying in Spain. Pyrexiae were quotidian, occurring nocturnally and would usually last 1 hour with associated malaise and tachycardia. The rash affected the upper arms, buttocks and face but was not consistent in appearance; initially urticarial and later described as maculopapular. Polyarthralgia of the joints of the hands was reported.
Extensive infection screening including blood cultures and serology was negative. A CT-CAP revealed changes of old TB and borderline axillary lymphadenopathy. Immunology revealed a negative ANA and ENA screen along with normal levels of anti-PR3, MPO, DsDNA and CCP antibodies. Further investigations included a CRP of 213, ESR of 75 and serum ferritin of 450mcg/l (15-200). A provisional diagnosis of a periodic fever syndrome was made and the patient agreed to a trial of anti-interleukin 1 therapy (Anakinra 100mg SC OD) while awaiting genetic testing. Immediate defervescence of fever occurred with an improvement in the rash and dramatic reduction in inflammatory indices.
Two months later the patient represented with malaise, tachycardia, periorbital odema, widespread rash, diarrhoea and a marked peripheral eosinophilia (5.69 x109/L). Skin biopsy demonstrated vacuolar inflammation and prominent eosinophils. A diagnosis of DRESS (Drug Reaction with Eosinophilia and Systemic Symptoms) syndrome was made and anakinra was withdrawn. However, the patient’s condition deteriorated with a rebound elevation in inflammatory markers, pyrexia and development of synovitis. Repeat serum ferritin during this period was recorded at > 40,000mcg/ml and a diagnosis of adult-onset Still’s disease was made.
Discussion
This case was challenging on two particular fronts. Firstly, the arrival of a confident and definitive diagnosis was difficult. On initial review and again on deterioration 2 months later, the patient’s previous history of TB, recent travel history and presence of lymphadenopathy led to significant concern of an alternative primary diagnosis. A large number of investigations for occult infection (e.g. TB recurrence) and malignancy (e.g. lymphoma) were conducted. Likewise, the absence of a history of a sore throat, the atypical skin rash, a very modest elevation in serum ferritin, and the absence of synovitis made adult-onset Still’s disease less likely. However, the rapidity and magnitude of the response to Interleukin-1 inhibition with anakinra supported our suspicion of an auto-inflammatory syndrome.
The second hurdle in this case was the problematic pharmaceutical management once adult-onset Still’s disease was diagnosed. The occurrence of DRESS syndrome secondary to anakinra is not something we had previously experienced nor does it appear to have been reported in the literature before despite a relatively high incidence of anaphylaxis and localized skin reactions with anakinra.
Subsequent interleukin-6 blockade with tocilizumab was partially effective in improving symptoms and clinical parameters but did lead to significant derangement in liver function tests and treatment was stopped. There was an incomplete response to TNF-α inhibition with weekly subcutaneous etanercept injections in combination with high dose oral corticosteroids. Subsequently there was a good response to the monoclonal antibody canakinumab (anti-IL-1β). Our patient remains on 10mg of oral prednisolone but has largely remained in remission for 18 months in combination with canakinumab. Remaining concerns relate to the long-term efficacy of canakinumab for this patient and the limited therapeutic options if recurrent relapses occur. Furthermore, the risk of TB re-activation remains an unavoidable risk with a high degree of clinical suspicion required.
Key learning points
This was a challenging case complicated by the patient’s past medical history of TB, myasthenia gravis and thymoma, in addition to the occurrence of DRESS syndrome which led to a period of diagnostic uncertainty. While the input of many specialties (i.e., respiratory, infectious diseases, haematology, and dermatology) were required and critical to the overall management of the patient, extensive and prolonged investigation can lead to significant delays in treatment. In such situations, an open discussion with the patient as to the risks and benefits of delaying treatment versus pursuing further investigation is advised. Furthermore, we have learnt from this case to appreciate that focused repetition of some investigations, in the setting of diagnostic uncertainty, can be beneficial. In this case repetition of serum ferritin levels, skin biopsy and cross-sectional CT imaging all led to important diagnostic conclusions and decisions that ultimately resulted in the correct diagnosis and successful management of this patient.
Conflict of interest
The authors declare no conflicts of interest.
Incidental papillary thyroid cancer diagnosis in patient with adult-onset Still’s disease-like manifestations
