Rheumatology Advances in Practice | 2019

8.\u2003An autoinflammatory syndrome resembling Castleman’s disease with excellent response to IL-6 blockade

 
 

Abstract


Abstract Introduction Autoinflammatory diseases are an emerging group, characterised by recurrent inflammatory episodes due to dysregulated innate immunity. Common features include fevers, rash, arthralgia, lymphadenopathy and systemic symptoms. Castleman’s disease is a rare lymphoproliferative disease, associated with the overproduction of interleukin-6 (IL-6). Its two main variants: unicentric and multicentric differ in aetiology and clinical outcomes. The cytokine storm driven by IL-6 can mimic autoinflammatory disease. We present the case of an acquired autoinflammatory syndrome in a 33-year-old male with the clinical phenotype of Castleman’s but no culprit lymph node detected radiologically. Symptoms dramatically improved with yocilizumab, an IL-6 blocker. Case description A 33-year-old Caucasian man was referred to rheumatology at his local District General Hospital with flitting joint pains, fevers, night sweats and weight loss for eighteen months. He had associated fatigue, myalgia, sore throat and an intermittent maculopapular rash. There were no specific symptoms of infection, malignancy or underlying connective tissue disease. No risk factors for blood-borne viruses were identified. Past medical history included hearing difficulties in childhood, and maternal family history of Crohn’s disease. Currently he was unemployed, having previously worked in a concrete factory. There was a 30 pack-year smoking history with moderate alcohol intake. Physical examination revealed a faint maculopapular rash over his right forearm but was otherwise normal. Full blood count showed an improved microcytic anaemia with recent haemoglobin 132\u2009g/L, raised white cell count up to 33 x 109/L (predominant neutrophilia) and mild thrombocytosis up to 480 x 109/L. Inflammatory were persistently elevated with CRP 124\u2009mg/L and ESR 67\u2009mm/hr. Renal, liver and thyroid functions were all normal as well as creatine kinase. Iron studies suggested iron-deficiency with negative anti-endomysial antibodies. Serum ferritin peaked at 1028 during µg/L during flares, with normal triglycerides. A full autoimmune screen was negative. Immunoglobulins showed a polyclonal rise only. HIV and Hepatitis screens were negative. CT chest, abdomen and pelvis and subsequent PET-CT scan were unremarkable. A bone marrow biopsy showed reactive changes only. A trial of low-dose prednisolone provided dramatic symptomatic improvement but symptoms flared on weaning to 10mg daily. Both steroid-sparing agents azathioprine and methotrexate were not tolerated. After further investigations by the National Amyloidosis Centre, he was commenced on weekly tocilizumab 162mg subcutaneous injections after a successful individual funding request. This provided an excellent clinical response which has been sustained for over two years. Discussion This case was difficult given the wide differential diagnoses. It was important to rule out infection, malignancy and autoimmune disease which were commoner causes of recurrent fevers and systemic symptoms. The long duration of symptoms, negative blood cultures and unremarkable CT imaging were against deep-seated infection. He was low risk for tuberculosis, zoonosis and tropical infections. No solid tumours or lymph nodes were seen on imaging but the PET-CT noted non-specific bone marrow changes. Bone marrow biopsy showed increased granulopoiesis without features of malignancy, and JAK-2 mutation was negative. Lactate dehydrogenase was normal with negative haemolysis screen. Upper and lower gastrointestinal endoscopies to investigate his iron-deficiency anaemia were normal. A full autoimmune screen was normal including anti-nuclear antibody, extractible nuclear antigen, rheumatoid factor, anti-cyclic citrullinated peptide antibody, complement C3 and C4 and anti-double-stranded DNA antibody. As no malignancy was found, prednisolone 40mg daily was trialled with fortnightly tapering. This produced a marked improvement in symptoms and inflammatory markers. However there were frequent flares on tapering the dose. He was therefore referred to the National Amyloidosis Centre at the Royal Free Hospital in London for an expert opinion. A genetic screen was negative for NLPR3 (CAPS gene), LRP12, TRAPS gene and the mevalonate kinase gene. Serum amyloid A (SAA) was very high 591 m/l (<10) with CRP 120\u2009mg/L. The clinical picture suggested an acquired autoinflammatory disease, most consistent with Castleman’s disease of the solitary plasma cell type. Adult-onset Still’s disease was considered but ferritin levels were not typical. A culprit lymph node is usually seen on imaging but occasionally can be too small to identify. Castleman’s responds very well to IL-6 blockade and SAA and CRP normalised with four doses of tocilizumab. Duration of treatment is unclear. Interval imaging was planned in case a resectable lymph node developed. Key learning points Autoinflammatory diseases are rare but treatable causes of fever syndromes. Extensive investigations are needed to exclude mimics such as infection, malignancy (especially haematological) and autoimmune conditions. Genetic testing can reveal the diagnosis for monogenic types such as familial Mediterranean fever (FMF), cryopyrin-associated periodic syndrome (CAPS) and tumour necrosis factor receptor-associated periodic syndrome (TRAPS). Castleman’s disease can be caused by a single lymph node (unicentric) or diffuse lymph nodes (multicentric). The unicentric type is less associated with systemic symptoms compared to multicentric, except for its rarer plasmacytosis variant. Consider HIV and human herpes virus-8 infection in the multicentric type. IL-6 blockade is extremely effective in Castleman’s but optimum duration of therapy remains unclear. Surgical resection of the solitary lymph node in unicentric Castleman’s has a good prognosis. Serum amyloid A can be a useful marker of disease activity in autoinflammatory disease compared with CRP. Conflict of interest The authors declare no conflicts of interest.

Volume 3
Pages None
DOI 10.1093/rap/rkz023.002
Language English
Journal Rheumatology Advances in Practice

Full Text