To the Editor:
Cryopyrin-associated periodic fever syndrome (CAPS) is a group of genetic diseases consisting of familial cold autoinflammatory syndrome (FCAS) and presenting with urticaria triggered by cold, Muckle-Wells syndrome (MWS) with fever, hearing loss, rash and joint pain, and neonatal onset multisystem inflammatory disease (NOMID), a severe neonatal disease. CAPS is associated with mutations in the NLRP3 gene1,2. NLRP3 associates with the proteins ASC and pro-caspase 1 to form the NLRP3 inflammasome, which is important for activation of pro–inter-leukin 1β (IL-1β) to mature IL-1β. Consequently, CAPS can be treated with anti-IL-1β therapy. Some patients with more diffuse inflammatory symptoms together with NLRP3 mutations have been classified as “atypical” CAPS3. We describe a patient with inflammatory symptoms fulfilling criteria of adult-onset Still disease (AOSD) but with a genotype typical of CAPS.
A 32-year-old man with previously known thrombophlebitis related to activated protein C resistance presented in May 2009 with sore throat, relapsing fever above 39°C especially in the afternoon, cervical adenitis, and transient maculopapular exanthema. C-reactive protein (CRP) was 192 mg/l, leukocyte count 11.3 × 109/l (ref < 8.8), neutrophils 8.2 × 109/l, aspartate aminotransferase 1.6 μkat/l (ref < 0.76), alanine aminotransferase 2.5 μkat/l (ref < 1.2), and lactate dehydrogenase 9.7 μkat/l (ref < 3.5). IgG antibodies against Epstein-Barr virus (EBV) were present, but EBV, cytomegalovirus, and other infections could not be detected. Three weeks after presentation, CRP was 54 mg/l; the patient felt better, but relapsing exanthema, tiredness, and low-grade fever continued.
In September 2010, sore throat, fever above 39°C, cervical adenitis, and exanthema relapsed, and in addition the patient presented with arthralgia, gonarthritis, and tenosynovitis of the right palmar hand. A faint red maculopapular rash appeared transiently, but it was neither urticaria-like nor typically salmon pink (Figure 1). No symptoms were induced by cold, whereas physical activity or psychological stress seemed to trigger flares. CRP was 214 mg/l, serum ferritin 5118 μg/l (ref 34–275), and leukocyte count 13.6 × 109/l (neutrophils 10.6 × 109/dl). Liver enzymes were normal at this occasion. Rheumatoid factor and antinuclear antibodies were negative. Serum amyloid A levels were within the normal range at 2 occasions. Urine albumin was never detected on dipstick examinations. Computed tomography examinations of thorax and abdomen were normal, and malignancy or infections could not be detected. AOSD was diagnosed according to the classification criteria of Yamaguchi, et al4. Table 1 shows different findings compatible with AOSD and/or CAPS. The clinical picture was not specific for FCAS, MWS, or NOMID, and the family history was negative. Audiogram was normal in 2012. No previous episodes of hearing loss or eye symptoms were reported.
The patient was given prednisolone 40 mg daily, with no relief of symptoms. Subcutaneous injections of anakinra 100 mg/day were started and 6 days later the fever ceased, CRP was < 10 mg/dl, and serum ferritin decreased to 1519 μg/dl. However, anakinra injections caused severe local skin reactions with fever, increasing levels of CRP (80 mg/dl), and serum ferritin (4818 μg/dl). Treatment with anakinra was stopped in October 2010 and trials to reintroduce anakinra were unsuccessful. With single therapy of prednisolone 40 mg/day, arthralgia and exanthema continued intermittently, although CRP normalized and serum ferritin decreased to 800 μg/l.
From November 2010 to April 2011 etanercept was used, and prednisolone was tapered from 30 mg to 5 mg/day. CRP and ferritin levels were in the normal range, but symptoms from joints and skin continued. In March 2011 the symptoms deteriorated, with sore throat, migrating arthralgia, and exanthema, and serum ferritin rose to 1211 μg/l. Prednisolone was increased to 40 mg/day, and in April 2011 intravenous tocilizumab 800 mg every second or third week was introduced. The infusions were complicated by thrombophlebitis and the clinical effect was only partial. In December 2011, the original symptoms relapsed in spite of increased prednisolone dose from minimum 10 to 30 mg/day, and eventually tocilizumab was discontinued.
In January 2012, subcutaneous canakinumab (an antibody against IL-1β) 150 mg every eighth week was started. The patient immediately felt better and all joint and skin symptoms disappeared. CRP, ferritin, and leukocyte levels returned to normal and have been normal since then. Prednisolone was stopped in July 2012 and he was able to resume long-distance running.
Gene sequencing of exon 3 of the NLRP3 gene revealed a heterozygous mutation (c.778C>T) in position R260W, typically found in CAPS2. He was also heterozygous for c.2107C>A (p. Q703K, rs35829419), a known gene polymorphism with an allele frequency of 6.5% in a Swedish population5. The importance of the Q703K polymorphism is difficult to judge because it has been reported in healthy individuals and patients with fever5. However, functional tests have revealed high IL-1β production from THP-1 cells with the R260W mutation, and moderately increased levels for the Q703K polymorphism compared to wild-type6.
The patient did not tolerate the IL-1 receptor antagonist anakinra, whereas the effect of the anti-IL-1 antibody canakinumab was remarkable, supporting the pathogenic importance of IL-1β in this case. Anti-IL-1 therapy is useful in CAPS but also in AOSD7, although patients with AOSD usually do not present mutations in the NLRP3 gene8,9. The symptoms of AOSD and CAPS partly overlap (Table 1), so different sets of diagnostic criteria may be challenging to use. Our patient fulfilled the Yamaguchi criteria, responded partially to higher doses of steroids, and had high ferritin levels, supporting phenotypic AOSD. The rash was indicative of AOSD, but rash as well as fever and response to IL-1 inhibition may also be seen in CAPS. Many patients with CAPS do not have a family history of CAPS or have symptoms triggered by cold. To our knowledge this is the first report of a patient with a phenotype compatible with AOSD but a genotype typical of CAPS.