Clinical manifestations and anti-TNF alpha therapy of juvenile Behçet’s disease in Taiwan

Behçet’s disease (BD) is a recurrent, multisystem, inflammatory disorder that affects vessels of all sizes. Its variety of clinical manifestations include recurrent oral and genital aphthous ulcers, skin lesions, arthritis, uveitis, thrombophlebitis, and gastrointestinal and central nervous system involvements [1, 2]. Juvenile BD (JBD) or pediatric BD refers to those patients who was diagnosed with Behçet’s disease at or before the age of 16 [3], and it accounts for 4–25% of all patients with BD [4]. The BD incidence varies according to geographical location, with the highest prevalence in Turkey (420 per 100,000 people), followed by Iran, northern China, and Korea [2]. The estimated BD incidence in Taiwan was 0.9 per 100,000 person-years from 2005 to 2009 [5]. Although the age of peak onset is between the third and fourth decades of life, awareness about the presence of BD during childhood has increased gradually [3, 6, 7]. Given its rarity and variety of clinical presentations, diagnosing JBD is challenging and often delayed.

The pathogenesis of BD remains unclear. While BD is largely considered a type-1 T helper (Th1) and type-17 T helper (Th17)-mediated inflammatory disease, characterized by elevated Th1 and Th17 cytokine levels [2, 8‐10], recent evidence suggests that the autoinflammatory nature of BD plays a role in its pathogenesis [11]. Given the complex mechanisms underlying BD, therapy depends largely on the affected site and disease severity. While the European League Against Rheumatism (EULAR) established treatment recommendations for the management of adult BD, no controlled studies have evaluated the treatment of JBD [8]. Immunosuppressive therapy has improved patients’ outcomes, but some BD patients have refractory disease courses. Anti-TNF-alpha therapy is useful for patients with refractory, severe BD, and especially for those with ocular [10, 12], central nervous system [10], and gastrointestinal [9, 10, 13] involvements. The use of anti-TNF-alpha therapy in children has been described in the context of small patient series or case reports only [14‐16]. In this study, we evaluated the clinical characteristics of patients with JBD and their treatments, with a particular focus on the efficacy and safety of anti-TNF-alpha therapy.

In this single-center retrospective study, we analyzed patients who were diagnosed with JBD and their treatment. We especially reported the favorable response and safety of using anti-TNF alpha in those patients with refractory symptoms after conventional treatments. Not only the disease severity scores were significantly reduced, but also the dosage of corticosteroid and immunosuppressive drugs were markedly reduced.

The variety of and lack of exact biomarkers for BD has led to the development of several sets of diagnostic criteria for BD. The International Behçet’s Study Group criteria are the most widely used diagnostic criteria for BD, but their sensitivity is suboptimal, especially for JBD [19‐21]. Other BD classification criteria have been developed for more robust diagnoses. The ICBD have been validated in JBD patients with sensitivities that range from 70 to 80%, but specificity data are absent [22, 23]. The PEDBD criteria were developed for children, and they have a sensitivity of 73.5% and a specificity of 97.7% [4, 21]. To replicate an actual population of JBD patients, we enrolled JBD patients based on the ICBD, because of their high level of sensitivity for these patients. We also applied the PEDBD criteria to these patients, because they are specific to JBD. In addition to the 50 patients who met either the ICBD or PEDBD criteria, five patients were identified as possible JBD in this study. All of these patients had recurrent oral aphthous ulcers, and they were diagnosed based on their associated symptoms or pathological evidence. They received colchicine or prednisolone, and the symptoms improved in four of these patients. The remaining patient who presented with fever, recurrent oral ulcers, and severe enteritis, experienced persistent symptoms while receiving conventional treatment, and she was eventually administered anti-TNF-alpha therapy. Our study identified pediatric patients who did not fulfill any of the clinical BD criteria, but required therapy like patients with confirmed JBD.

The median age at disease onset and time of diagnosis were 11 and 13 years old respectively in our JBD patients. The mean onset age of JBD reported from other studies [4, 22, 23] ranged from 4.87 to 7.4 years old, which was younger than that of our patients. The reported age at JBD diagnosis was similar amongst those studies (11.29–13.87 years old) and our data. The differences may be explained by the recall bias but it may also be an indication that our physicians were not as alarming as those in other studies in recognizing BD symptoms which appeared in early childhood. There were patients who were diagnosed as BD at an extremely young age in our study group. Among 55 patients, three of them had disease onset at the age less than one-year-old, which were 0.1, 0.25, 0.25 years old respectively. The patient diagnosed as BD at 0.1 years old has been reported as a case report [24]. The case presented with oral ulcer, genital ulcer, and associated GI symptoms, and no evidence of infection were identified after serial studies. She had high serum level of CRP (20 mg/dl) and ESR (90 mm/hr) initially. The symptoms resolved after treatment with short-course of prednisolone and the baby was clinically asymptomatic after 2 months old. There were also other case reports of neonatal BD from Turkey, UK, and Canada [25‐28]. Hence despite being rare, diagnosis of BD should be considered in a neonate even if the full diagnostic criteria are not met. The other two cases had symptoms at infancy but fulfilled BD diagnosis at 2 and 3 years old respectively. Both of them had persistent symptoms when they grew up and eventually received anti-TNF alpha therapy.

Specific treatment guidelines for JBD are not available. The EULAR recommends BD treatment should be individualized according to age and gender, the type and severity of organ involvement, and a patient’s preferences [8]. Our data showed that the most frequently used treatments were prednisolone and colchicine, and these comprised first-line therapy for JBD. Compared with the findings from previous studies conducted in other countries, prednisolone was used more frequently in our patients [22, 23, 29]. Among the patients who received oral prednisolone, 24.3% had uveitis and 26.8% presented with gastrointestinal involvements, both of which are associated with poor prognoses [8, 16], and more aggressive treatment was required. Given the rapid responses to systemic corticosteroids and the limited data describing other medications administered to patients with JBD, corticosteroids remain a good choice for short-term therapy.

We found that the patients who received anti-TNF-alpha therapy had used corticosteroids for longer durations and they had used more types of DMARDs and immunosuppressants before they received anti-TNF-alpha therapy. The BDCAF scores were 7 or 8 out of a total score of 20 before they received anti-TNF alpha therapy. Hence, these patients may not have a very severe disease entity but had a refractory disease course even after multiple conventional treatments. Our analysis showed that the patients who received anti-TNF-alpha therapy were younger at disease onset, and two patients were diagnosed with JBD in infancy. The influence of age on JBD has not been described. In addition, the JBD patients who received anti-TNF-alpha therapy had higher serum CRP levels at the time of disease onset and higher rates of gastrointestinal symptoms and arthritis/arthralgia. Intestinal involvement is more frequent in juvenile BD, and its severity can range from mild to hemorrhage or perforation [3]. The disease in patients with intestinal BD sometimes becomes refractory, and other forms of therapeutic intervention and biologics are reported in addition to conventional therapy [14, 15, 30].

TNF-alpha inhibitors treat several rheumatic diseases effectively, even those in children [31, 32]. The findings from several studies that have evaluated the effectiveness and safety of anti-TNF-alpha therapy in adult BD patients with refractory uveitis, have shown symptom improvements in 43.5–67.7% of the patients [10, 12]. Inoue et al reported the treatment response of adalimumab in adult patients with refractory intestinal BD. Sixty percent of the patients showed marked improvements and 20% of them attained complete remission [13]. Among the 6 patients receiving anti-TNF-alpha therapy, we used adalimumab for the two patients with uveitis based on our experience in the treatment of uveitis in juvenile idiopathic arthritis [33, 34]. Furthermore, adalimumab could be reimbursed by the National Health Insurance in Taiwan for patients with uveitis. However, for BD patients with GI tract involvement, adalimumab is an out of pocket expense. Infliximab was not available in Taiwan during the period when our patients had the disease. The other four patients chose etanercept as it was a cheaper choice of anti-TNF alpha agent for these patients. Our results indicated the effectiveness and safety of anti-TNF-alpha therapy in JBD patients whose symptoms were refractory to conventional treatments. Four patients experienced complete symptom remissions and steroid sparing within 6 months. The remaining two patients had partial responses, and their conditions remained stable without disease progression. The reduced BDCAF scores, the changes in the steroid doses and the immunosuppressant load scores after anti-TNF-alpha therapy provided indirect evidence of the patients’ treatment responses, especially after 6–12 months of treatment.

Though several large-scale clinical trials of anti-TNF-alpha therapy in adult BD were conducted, there was a lack of similar trials in pediatric patients. We searched the Medline database published from January 2000 till December 2018 via PubMed. The searching keywords included “Behçet’s disease”, either “child” or “pediatric”, and either “anti-tumor necrosis factor” or “infliximab.” The references which had been cited in other related case report were also reviewed. Literature without full text or in a non-English language was excluded. We found seven case reports and series reported anti-TNF-alpha therapy for refractory JBD [14‐16, 35‐38]. The details of the reported cases were listed in Table 5. Combining our cases with the previous seven reports of JBD with anti-TNF alpha therapy in literature, there were a total of 18 JBD patients receiving anti-TNF alpha agents. There were more female patients (11/18). The median disease onset age was 10 years old with a range of 2 to 15 years old. The most commonly involved organs in these patients were cutaneous lesions (6/18) and intestinal involvement (6/18), followed by ocular lesions (4/18). Use of adalimumab, infliximab and etanercept have been reported respectively. Ten patients were treated with etanercept, six patients received infliximab, and three of them used adalimumab. Among the 18 cases, fifteen of them had favorable outcomes after treatment, especially in intestinal and ocular lesions, but three patients failed to show any significant improvement, including two cases with Budd-Chiari syndrome. There were some side effects which were mainly non-serious infection. Combining these previous studies and our studies together, using anti-TNF-alpha therapy may be beneficial and safe for refractory JBD patients.

Nevertheless, the anti-TNF-alpha therapy did not cure the JBD patients optimally, as we expected. Two patients in our study experienced disease flares after we reduced the frequency of the anti-TNF-alpha therapy. A previous study reported disease relapses occurred in 58.6% of the adult BD patients 1 year after the anti-TNF-alpha treatments were discontinued [39]. Further investigations are required to determine the timing of treatment cessation. Our group also included a patient who did not reach complete remission until he received tocilizumab, which is a humanized anti-IL-6 receptor monoclonal antibody, 1.4 years after using etanercept. This patient does not use any drugs other than tocilizumab currently. Some publications describe refractory BD patients who received tocilizumab [40‐42], but the effectiveness of the treatment varied, and limited numbers of BD patients are described. As evidence of the pathogenesis underlying BD continues to emerge, more therapeutic options become available for severe BD. Large-scale trials that investigate the treatment of BD patients, and especially JBD patients, with biologics are necessary to enable the development of more effective and safe treatments.

This study has some limitations. First, this was a retrospective study, and recall and selection biases existed. Second, a relatively small number of patients received anti-TNF-alpha therapy. Third, not all enrolled cases were evaluated by disease severity score. Nevertheless, this was a relatively large retrospective study of anti-TNF-alpha therapy in Asian JBD patients. Although further larger prospective studies are needed to determine the efficacy and safety of anti-TNF-alpha therapy in JBD patients, it seems that anti-TNF therapy might play an important role for refractory JBD.

Most of the JBD patients who received corticosteroid- or colchicine-based treatments had favorable outcomes; however, a small proportion of these patients had refractory disease courses even after receiving conventional medication. We found that patients who were younger at disease onset, had higher serum CRP levels at baseline, and presented with gastrointestinal symptoms tended to have poor response to our traditional standard treatments. Anti-TNF-alpha therapy was effective and safe in these patients and it showed corticosteroid- and immunosuppressant-sparing effects.