Patients with juvenile idiopathic arthritis (JIA) are at increased risk for developing inflammatory bowel disease (IBD)1, although their association with systemic-onset JIAs is exceptional2–4, Mutations in genes causing immune deficiency due to immune dysregulation, common to both entities5, or the role of anti-IL1 treatments are hypothesised2,3.

We present the case of an 8-year old girl with fever of one month’s duration, anorexia, sweating, and weight loss. Physical examination revealed no abnormalities.

Laboratory tests revealed leukocytosis (25,000/mm3) with a shift to the left, thrombocytosis (528,000/mm3), CRP 163 mg/l, ESR 92 mm, ferritin 737 ng/mL, and hypergammaglobulinaemia (IgG 2,650 mg/dl). The rest of the biochemical study failed to exhibit any further alterations. The infectious screening, which included tuberculosis, was negative. ANA and rheumatoid factor were similarly negative. Faecal calprotectin (FC) was 58 ng/g. Chest X-ray, abdominal ultrasound, thoracoabdominal CT, echocardiography, MRI, and bone scan displayed no pathological findings. The fundus examination was normal and the bone marrow biopsy exhibited no abnormalities.

After one week she developed right axillary adenopathy, hypoechoic on ultrasound, and with pathological uptake on the PET-CT scan. After surgical excision, the anatomopathological and microbiological study showed no alterations. Subsequently, he developed swelling and limitation of the left knee, with joint effusion with no synovial hypertrophy.

With a diagnosis of JIA (ILAR 2001 criteria), oral prednisone (1.5 mg/ kg/ day) was started, to which subcutaneous anakinra (2 mg/ kg/ day) had to added after 3 weeks, due to relapse after decreasing the corticosteroids. One month later, it was substituted for subcutaneous canakinumab (4 mg/ kg/ 4weeks), making it possible to administer doses every 6 weeks, without incident, and she remained asymptomatic for 3 years.

After this, the patient once again began to have fever, arthralgias, and an increase in CRP. Given her incomplete clinical and analytical response to oral prednisone (1 mg/ kg/ day), intravenous tocilizumab (8 mg/ kg/ 2 weeks) was initiated, with partial improvement.

One month later she was readmitted for abdominal pain and haematochezia. Laboratory tests yielded haemoglobin values of 7.6 g/ dl and CF 3132 ug/ g. After ultrasound evidence of ileocolitis, upper and lower gastrointestinal endoscopy was performed and a diagnosis of Crohn’s disease (CD) was established according to the Oporto criteria6 and the Paris classification7. The genetic panel for autoinflammatory diseases resulted in a probably benign mutation (pR702W) and a mutation of uncertain significance (p1007finsC), both in heterozygosis, of the NOD2 gene.

Tocilizumab was discontinued, and subcutaneous adalimumab (160-80-40 mg/ 2 weeks) and oral prednisone (1.5 mg/ kg/ day) were started, requiring combo therapy with azathioprine. She is currently asymptomatic on adalimumab monotherapy, with CF < 15.6 ug/ g.

IBD very rarely develops as an extra-articular manifestation of JIA2–4. Although some authors propose that pharmacological blockade of IL-1 would favour the phenotypic change of JIA to IBD2,3, this evolution has also been detected in individuals with other treatments (Table 1)3,4 in whom this hypothesis is not supported.

The JIA-IBD overlap is an alternative to be considered, with good response to treatment based on anti-TNF drugs and even disappearance of systemic symptoms in patients with JIA who subsequently progress to CD3. The description of mutations of the LACC1 gene, associated with both entities, would support this possibility5.

The detection of mutations in the NOD2 gene associated with CD in our patient poses the alternative that the initial symptoms, compatible with JIAs, were the set of extraintestinal manifestations of the onset of an underlying CD, although the normal CF and imaging tests allowed this diagnosis to be ruled out. We do not know the role that anti-IL1 treatment played in lengthening the latency period until the development of classic CD symptoms8.

As this is a single case, and given the absence of genetic studies in the published series2–4, it is not possible to extrapolate conclusions, and further studies are needed to prove or disprove these hypotheses. Be that as it may, in diseases with a diagnosis of exclusion, such as JIA, other diagnostic alternatives or the association with other diseases due to immune dysregulation should be assessed in the event of the appearance of atypical manifestations or loss of effectiveness of the usual treatment.

This work has not received any funding of any kind.

None.