Abstract
Objective To evaluate tocilizumab (TCZ) efficacy in patients with refractory Behçet disease (BD).
Methods This is a multicenter study of 30 patients fulfilling the international criteria for BD and treated with TCZ at different European referral centers. The clinical response was evaluated at 6 months from TCZ initiation.
Results Ninety percent of patients with BD were refractory or intolerant to anti–tumor necrosis factor (anti-TNF) agents. Overall, TCZ was effective in 25 (83%) patients with BD of whom 18 (60%) and 7 (23%) were complete and partial responders, respectively. The complete response was 67%, 60%, and 42% in patients with uveitis (18/30), neurological manifestations (5/30), and mucocutaneous and/or articular (7/30) manifestations, respectively. TCZ had a significant steroid-sparing effect allowing patients to decrease their median daily prednisone dose from 20 (IQR 10-40) mg/day to 9 (IQR 5-13) mg at 6 months (P < 0.001). The number of patients with BD needing concomitant disease-modifying antirheumatic drug therapy fell from 7 (23%) to 4 (13%) at 6 months. Mild to moderate side effects were observed in 6 (20%) patients, and 3 (10%) presented with serious adverse events (pneumonia, intestinal perforation, and septicemia) requiring therapy discontinuation in 2 cases.
Conclusion TCZ seems to be an effective alternative to anti-TNF agents in treating BD-related uveitis and neurological manifestations.
Behçet disease (BD) is a systemic vasculitis of unknown origin affecting vessels of variable size.1,2 Its classical clinical manifestations include recurrent oral and genital ulcers, pseudofolliculitis, erythema nodosum, and uveitis, and in the more severe forms, gastrointestinal, articular, vascular, and neurological manifestations.3,4 Even if clinical manifestations are variable, different major clusters of the disease have been described: mucocutaneous and articular, extraparenchymal neurological and peripheral vascular, and parenchymal neurological and ocular.5 Treatments for BD range from colchicine, low-dose glucocorticoids (GCs), topical GCs, and nonsteroidal antiinflammatory drugs for mucocutaneous and articular involvement to immunosuppressive therapies for ocular, vascular, neurological, and gastrointestinal symptoms. Prompt initiation of immunosuppressants such as GCs, conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), interferon-α, cyclophosphamide (CYC), and biologic disease-modifying antirheumatic drugs (bDMARDs), in particular anti–tumor necrosis factor (anti-TNF) agents,6-8 is mandatory in cases of life-threatening manifestations.9
Anti-TNF agents have proven effective in treating most BD clinical manifestations and their use is recommended as first-line therapy in patients with severe ocular, vascular, gastrointestinal, and central nervous system involvement.7,8 Despite their efficacy, there is still a need for alternative therapies, as up to 35% of patients are refractory, intolerant, or present with contraindications to anti-TNF agents.10 Therefore, unmet therapeutic needs in BD have drawn attention to biologic agents targeting cytokines other than TNF.10-12 Interleukin (IL)-6 seems to play an instrumental role in BD. High levels of IL-6 were found in sera of patients with BD, correlating with disease activity and arthritic manifestations.13 In addition, high levels of IL-6 have also been detected in the cerebrospinal fluid of patients with neuro-BD and in the vitreal fluid of patients affected by autoimmune uveitis, contributing to ocular inflammation.14,15 IL-6 stimulates the differentiation of T lymphocytes to Th17 lymphocytes, which act as proinflammatory mediators, with a concomitant reduction in regulatory T cells.16 Tocilizumab (TCZ), a humanized antibody targeting the membrane IL-6 receptor, has been used thus far as an off-label therapy in a case series of patients with BD not responding to the approved treatments.17-23 To date, the TCZ experience is limited in BD, and even if TCZ seems effective in most ophthalmological, neurological, vascular, and gastroenterological case series, conflicting results were obtained in terms of its effect on mucocutaneous and/or articular disease manifestations.
Herein, we aimed to evaluate TCZ efficacy in 3 different BD phenotypes (mucocutaneous and/or articular disease, ophthalmological, and neurological BD) in a multicenter cohort of 30 patients with refractory BD.
METHODS
Patients. We conducted a multicenter retrospective study in referral hospitals from France, Italy, Spain, and Turkey between December 2021 and June 2022. All 30 enrolled patients met the criteria of the international study group for BD.24 All patients had either mucocutaneous and/or articular manifestations, and/or uveitis, and/or neurological BD manifestations that were refractory to colchicine, csDMARDs, and/or bDMARDs. The study was performed according to the Declaration of Helsinki. According to our national policy, patients systematically received information on the electronic storage of their data for administrative and research purposes. They can exercise their right of opposition. No institutional review board approval is necessary.
TCZ regimen. TCZ was administered intravenously (IV) at 8 mg/kg every 4 weeks in 77% of patients or subcutaneously 162 mg once weekly in 23% of patients. Concomitant therapy included a stable dose of GC in 27 (90%) patients, colchicine in 12 (of 27, 44%), and csDMARDs in 7 (of 27, 26%).
Data collection. Demographic features and past medical history of BD were recorded. Data regarding BD manifestations, including oral and genital ulcers, skin manifestations, ophthalmological, vascular, and neurological involvement, were collected. Joint involvement was assessed using tender and swollen joint count. TCZ indication and route of administration, concomitant treatments, and previous failed therapies were also of special interest. Clinical variables, safety assessment, daily GC use, and laboratory findings were collected before TCZ therapy, at the time of TCZ first administration, after 3 months, 6 months (M6), and at the date of the last follow-up visit.
End points. The primary efficacy end point was the proportion of patients reaching a clinical response (complete or partial) at M6. Complete response (CR) was considered as the remission of the affected organs involved at baseline. Response to treatment was evaluated for each organ representing the indication for TCZ treatment and assessed as CR, partial response (PR), or nonresponse (NR). For the uveitis group, CR was defined as a complete resolution of uveitic macular edema (ME; central foveal thickness ≤ 300 mm with resolution of intraretinal cystic spaces) with a GC daily dosage of ≤ 10 mg at M6, without intraocular inflammation (grade 0 for anterior chamber cells and vitreous haze).25 PR was defined as an improvement of ME without complete resolution, an improvement of intraocular inflammation, and a reduction of the initial GC dosage at M6. Patients showing CR of uveitic ME with a GC dosage > 10 mg/day at M6 were also considered to be partial responders. The remaining patients were considered nonresponders. For neurological BD, CR was defined as a complete clinical remission and imaging normalization (evaluated by magnetic resonance imaging) in the absence of neurological sequelae (defined as a Rankin score ≤ 1) at M6. PR was defined as an improvement, without imaging normalization. The remaining patients were considered to be nonresponders. For mucocutaneous and/or articular disease, CR was defined as the absence of oral and genital aphthae, skin lesions, and swollen joints at M6. PR consisted of the reduction of ≥ 50% in the number of oral and genital aphthae, skin lesions, and swollen joints at M6. The remaining patients were considered to be nonresponders.
Secondary end points included the proportion of patients with a CR, PR, and NR at M6; the disease relapse rate in the course of treatment; the TCZ steroid-sparing effect between baseline and M6; the TCZ retention rate; and the safety profile of TCZ.
Statistical analysis. Data are presented as a mean (SD) or median (IQR) for continuous variables, and as a percentage for qualitative variables. Wilcoxon signed-rank test was used to compare continuous variables and Fisher exact test to compare categorical variables. P values < 0.05 were considered significant. Statistical analyses were performed using GraphPad Prism 6.0 (Dotmatics Inc.).
RESULTS
Characteristics of patients with BD. We included 30 patients (17 women) with a median age at BD diagnosis of 30 (IQR 24-33) years. Baseline characteristics and outcomes are summarized in Tables 1, 2, 3, and 4.
Demographic and clinical features of patients with BD.
Efficacy and safety of TCZ in patients with BD.
Outcomes of patients with BD treated with TCZ for uveitis.
Outcomes of patients with BD treated with TCZ for neurological or mucocutaneous and/or articular manifestations.
Indications for TCZ were refractory uveitis in 18 (60%) patients, mucocutaneous and/or articular in 7 (23%), and neurological manifestations in 5 (17%; Table 2). One of the 7 patients with mucocutaneous and/or articular disease also presented with renal AA amyloidosis. HLA-B51 was positive in 14/19 (74%) subjects tested (Table 1).
Before TCZ, all patients had already received colchicine and 27 (90%) GCs. Patients received a median of 3 (IQR 2-4) courses of csDMARDs (azathioprine [48%], methotrexate [30%], CYC [30%], cyclosporine [19%], mycophenolate mofetil [7%], and tacrolimus [4%]) before TCZ treatment. Twenty-seven (90%) patients also received anti-TNF agents (11 [37%] and 3 [10%] patients received 2 or 3 anti-TNF agents, respectively), 5 (17%) anakinra, 2 (7%) ustekinumab, and 1 (3%) canakinumab prior to TCZ (Table 1).
Efficacy. TCZ was effective in 25 (83%) patients with BD at M6. Eighteen (60%) patients reached a CR, 7 (23%) PR, and 5 (17%) NR (Table 2).
Among the 18 patients treated for uveitis, TCZ was effective in 15 (83%) with a CR and PR in 12 (67%) and 3 (17%) patients, respectively (Table 3). Among the 3 nonresponders, 2 patients discontinued TCZ at month 2 and month 4, respectively, because of refractory retinal vasculitis. TCZ was effective for uveitic ME in 88% (CR 75%, PR 13%) and retinal vasculitis in 84% of patients (CR 67%, PR 17%). The mean visual acuity increased from 5.5 and 4.8 out of 10 (left and right eye, respectively) at baseline to 8.3 and 7.8, respectively, at M6.
TCZ was effective in all patients with neurological manifestations (CR 60%, PR 40%; Table 4).
Among the 7 patients treated for mucocutaneous and/or articular disease, a clinical response was obtained in 5 (71%) patients (CR 42%, PR 29%; Table 4). Two patients presenting with pyoderma gangrenosum as a cutaneous manifestation had a CR, as well as a patient with renal AA amyloidosis (Table 4).
TCZ demonstrated a GC-sparing effect. The median daily prednisone dose dropped from 20 (IQR 10-40) mg/day to 9 (IQR 5-13) mg at M6 (P < 0.001) and 5 (IQR 0-9) mg at the last visit of follow-up. Three of the 7 patients receiving concomitant csDMARDs were able to withdraw these therapies at M6.
TCZ retention rate. After a median follow-up of 31 (IQR 21-56) months, 13 (43%) patients were still receiving TCZ and maintained remission. The 17 (57%) remaining patients discontinued TCZ as a result of side effects, failure, and/or relapse (n = 12) or after achieving remission (n = 5; Tables 2, 3, and 4).
Safety. Under TCZ therapy, mild to moderate side effects were observed in 6 (20%) patients and included skin rash (n = 2), worsening of mucocutaneous disease manifestations (n = 1), recurrent urinary tract infections (n = 1), neutropenia (n = 1), and dyslipidemia (n = 1). Three (10%) patients presented serious adverse events (pneumonia, intestinal perforation, and septicemia) requiring therapy discontinuation in 2 cases (Table 2).
DISCUSSION
Despite the efficacy of anti-TNF agents, there is still a need for alternative therapies, as up to 35% of patients are refractory, intolerant, or present contraindications to these agents.10 Therefore, unmet therapeutic needs in BD have drawn attention to biologic agents targeting cytokines other than TNF.10 IL-6 seems to play a pivotal role in BD, and TCZ represents a possible new therapeutic strategy.11 Currently, the TCZ experience is limited in BD, and even if TCZ seems effective in case series, conflicting results were obtained in terms of its effect on mucocutaneous and/or articular disease manifestations.
Herein, we report the largest experience of TCZ that we know of in patients with refractory BD. Ninety percent of our patients with BD were refractory or intolerant to anti-TNF agents. TCZ was administered IV in most of our patients. We could evaluate TCZ efficacy in 3 main BD phenotypes such as uveitis, neurological, and mucocutaneous and articular manifestations. The main conclusions drawn by this study are: (1) TCZ seems to be an effective alternative to anti-TNF agents in patients with BD with refractory uveitis and neurological manifestations, and (2) the efficacy of TCZ seems less clear in the mucocutaneous and articular phenotype.
In a systematic literature review, Akiyama et al analyzed the outcomes of TCZ in 47 patients with refractory BD, concluding that anti–IL-6 treatment could be a valid alternative for refractory ocular, neurological, and vascular BD, as well as for secondary AA amyloidosis, but not for mucocutaneous and articular forms.26 Previous case reports have indeed highlighted the diversity of response rates to TCZ among the different disease clusters, confirming good efficacy in ophthalmological, neurological, and vascular disease with conflicting results for the articular and mucocutaneous phenotype15-21,25-29 (Table 5).
Case series of patients with BD treated with TCZ.
In our series, ophthalmological manifestations were well-controlled by TCZ, with a recovery in 84% of patients and a beneficial effect on uveitic ME in 88% of cases and on retinal vasculitis in 84% of subjects. This is in agreement with previous studies of TCZ in BD uveitis21,27 (Table 5). Atienza-Mateo et al highlighted the efficacy of TCZ (CR 63%, PR 19%) in 16 patients with BD uveitis who did not respond to conventional and anti-TNF agents (Table 5). Eser Ozturk et al reported 5 patients with sight-threatening BD uveitis who were refractory to interferon-α and conventional and anti-TNF agents; all cases achieved CR with TCZ treatment (Table 5). Leclercq et al compared TCZ and anti-TNF agents in refractory uveitic ME, showing TCZ superiority.28 Many other single case reports or limited case series showed the efficacy of TCZ in ocular BD manifestations.18,19,29,30
For neurological involvement, TCZ was also able to induce remission in all our patients. These results are consistent with previous reports. Liu et al treated 11 patients with BD with refractory neurological involvement achieving a CR in 20% and PR in 80% of cases31 (Table 5). Atienza-Mateo et al reported efficacy of TCZ in 5 patients with refractory neurological BD (3 CR and 1 stabilization21; Table 5). Many other single case reports or limited case series confirm this trend.17,32-34
We could not evaluate the efficacy of TCZ in vascular BD because none of our patients were treated for vascular involvement. TCZ showed good results in vascular BD in a previous Chinese study, in which 9 of the 10 patients with vascular BD obtained a clinical response (CR 50%, PR 40%35; Table 5). Ding et al reported a cohort of 7 patients with vascular BD refractory to GCs and csDMARDs, achieving a clinical response with TCZ (CR 42%, PR 42%, and 1 patient nonevaluable because of premature discontinuation due to financial issues20; Table 5).
Reports on the effects of TCZ on mucocutaneous and articular lesions are contradictory.13-15,17-19,30-32 Among our mucocutaneous and articular cluster, we obtained a lower remission rate as compared to other clinical phenotypes. In the literature, TCZ had few effects on oral ulcerations in some case series.21,22,36 Worsening of mucocutaneous lesions after TCZ has also been reported, sometimes requiring drug discontinuation.17,23,37 IL-6, in fact, is an important key factor in wound healing, so its reduction may impair the cutaneous and mucosal healing process.38
For joint manifestations of BD, literature results on TCZ are also conflicting. In the case series by Atienza-Mateo et al, 4 of 7 patients improved, with a CR in 2 cases.21 In contrast, many other reports pointed out the failure of TCZ for articular symptoms.19,23 In line with these previous experiences, our data show poor efficacy of TCZ on joint manifestations in BD. One patient in our study with mucocutaneous and articular disease, and who also presented with renal AA amyloidosis, obtained a CR with TCZ. Two cases of BD-related secondary renal AA amyloidosis treated with TCZ have been reported in the literature, and in both cases a CR was obtained.39,40 Finally, TCZ was effective in sparing GCs in our series and in published case reports20,31,35 (Table 5).
We do not report new safety signals of TCZ. Common side effects of TCZ such as neutropenia, thrombocytopenia, dyslipidemia, increased transaminase level, and upper respiratory tract infections were observed in 30% of our patients with BD, in line with previous reports.41 We observed 2 cases of severe sepsis requiring treatment interruption, including a patient presenting with an intestinal perforation—a well-known drug-related side effect.42,43
This study presents some limitations. Its retrospective nature could not allow for the evaluation of the exact number of oral and genital ulcers every month, as well as the exact number of swollen joints. Another limitation is the small number of subjects affected by neurological and mucocutaneous/articular disease forms.
In conclusion, our study provides the results of the largest cohort of patients, that we know of, with refractory BD treated with TCZ. Ninety percent of patients with BD were refractory or intolerant to anti-TNF agents. We highlight that TCZ seems to be an effective alternative to anti-TNF agents in patients with BD with refractory uveitis and neurological manifestations. Further prospective studies are warranted to confirm these results.
Footnotes
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication January 25, 2023.
- Copyright © 2023 by the Journal of Rheumatology
