Rituximab therapy for intractable pemphigus: A multicenter, open‐label, single‐arm, prospective study of 20 Japanese patients

Abstract This was a multicenter clinical trial of rituximab, a chimeric monoclonal IgG antibody directed against CD20, for the treatment of refractory pemphigus vulgaris and pemphigus foliaceus. In total, 20 patients were treated with two doses of rituximab (1000 mg; 2 weeks apart) on days 0 and 14. The primary end point was the proportion of patients who achieved complete or partial remission on day 168 following the first rituximab dose. Of the 20 enrolled patients, 11 (55%) and four (20%) achieved complete and partial remission, respectively; therefore, remission was achieved in a total of 15 patients (75.0% [95% confidence interval, 50.9%–91.3%]). It was demonstrated that the remission rate was greater than the prespecified threshold (5%). In addition, a significant improvement in clinical score (Pemphigus Disease Area Index) and decrease in serum anti‐desmoglein antibody level were observed over time. Four serious adverse events (heart failure, pneumonia, radial fracture, and osteonecrosis) were recorded in two patients, of which only pneumonia was considered causally related with rituximab. The level of peripheral blood CD19‐positive B lymphocytes was decreased on day 28 after rituximab treatment and remained low throughout the study period until day 168. Our results confirm the efficacy and safety of rituximab therapy for refractory pemphigus in Japanese patients.

remission on day 168 following the first rituximab dose. Of the 20 enrolled patients, 11 (55%) and four (20%) achieved complete and partial remission, respectively; therefore, remission was achieved in a total of 15 patients (75.0% [95% confidence interval, 50.9%-91.3%]). It was demonstrated that the remission rate was greater than the prespecified threshold (5%). In addition, a significant improvement in clinical score (Pemphigus Disease Area Index) and decrease in serum anti-desmoglein antibody level were observed over time. Four serious adverse events (heart failure, pneumonia, radial fracture, and osteonecrosis) were recorded in two patients, of which only pneumonia was considered causally related with rituximab. The level of peripheral blood CD19-positive B lymphocytes was decreased on day 28 after rituximab treatment and remained low throughout the study period until day 168. Our results confirm the efficacy and safety of rituximab therapy for refractory pemphigus in Japanese patients.

K E Y W O R D S
anti-desmoglein antibody, corticosteroid tapering, refractory pemphigus, rituximab

| INTRODUC TI ON
Pemphigus is an autoimmune bullous disease characterized by skin and mucous membrane blisters and erosions with intraepidermal blister formation caused by IgG autoantibodies directed against adhesion proteins, desmoglein 1 (Dsg1) and desmoglein 3 (Dsg3), located in the desmosomes between epidermal keratinocytes. 1,2 For patients with moderate to severe pemphigus, guidelines recommend high-dose systemic corticosteroids, such as prednisone (1 mg/kg per day) as the first-line treatment. [3][4][5] The goal of treatment is remission, defined as symptom freedom on ≤10 mg/day of prednisone or equivalent dose and minimal adjuvant therapy (e.g., immunosuppressive agents). 6 Japanese guidelines recommend that pemphigus treatment be divided into two phases: consolidation and maintenance. 3 During the consolidation phase (initial therapy), treatment is adjusted until the disease is controlled, including tapering of the corticosteroid dose. During the maintenance phase (maintenance therapy), treatment is maintained and the corticosteroid dose is tapered. Additional therapies, such as plasma exchange and high-dose intravenous immunoglobulin, should be administered if initial therapy is not sufficient in controlling disease activity. A recent study reported that >90% of patients with pemphigus achieved remission within the first 2 years of guideline-based treatment, suggesting the effectiveness of such a treatment plan. 7 However, there are still unmet needs in pemphigus treatment, including the inability to achieve remission in ≈10% of patients and the high incidence (≈80%) of treatmentrelated side effects, such as diabetes, osteoporosis, and serious infections, which may be fatal. 8,9 Therefore, targeted therapies against autoantibody production are required to improve the treatment of pemphigus.
Rituximab is a chimeric human-mouse IgG monoclonal antibody that binds to the transmembrane antigen CD20 expressed from the pre-B-cell stage to the preplasma cell stage. 10 The binding of rituximab to CD20 leads to B-cell depletion through various mechanisms, including complement-dependent cytotoxicity, antibody-dependent cell-mediated cytotoxicity, and apoptosis. 11 Rituximab is approved for the treatment of B-cell lymphoma and autoimmune diseases, including rheumatoid arthritis and anti-neutrophil cytoplasmic antibody-associated vasculitis, although the indications vary between regions. Since 2001, several case reports and case series have reported remarkable therapeutic effects of rituximab in pemphigus patients. [12][13][14][15] Rituximab has already been administered to >500 patients and is recommended as the standard treatment for refractory cases of pemphigus by the European guidelines. 16 In Japan, we conducted an exploratory study of the efficacy and safety of rituximab in patients with autoimmune bullous diseases refractory to corticosteroid therapy since 2009 and found that rituximab reduced the disease activity and decreased the corticosteroid dose requirement. 19 Based on previous reports, we performed an investigator-initiated clinical trial to validate the efficacy and safety of rituximab for the treatment of refractory pemphigus in patients who did not achieve remission with other treatments.

| Patients
Patients diagnosed with pemphigus vulgaris (PV) or pemphigus foliaceus (PF) in accordance with the Japanese guidelines for the management of pemphigus and disease relapse during corticosteroid taper were included in the study. 3 Patients with a Pemphigus Disease Area Index (PDAI) score of 1 to 50 at the time of obtaining consent, daily corticosteroid dose of 15 to 30 mg/day (prednisone equivalent), no change in dose during 14 days of the screening period before study registration, and unchanged or worsened PDAI score during the screening period were included. 20 We excluded patients receiving consolidation therapy, such as increasing doses of immunosuppressive agents, intravenous immunoglobulin, steroid pulse therapy, or plasma exchange in the prior 8 weeks. For safety reasons, we also excluded patients who were pregnant, who had undergone surgery within the prior 4 weeks, who had taken antibiotics within the prior 8 weeks, and/or those with a history of allergic reactions against chimeric human-mouse IgG monoclonal antibody, severe organ disorders (such as chronic obstructive pulmonary disease, asthma, heart disease, and hypertension), active or chronic infection, deep-seated infection (such as abscess, fasciitis, and osteomyelitis) within the past 1 year, or malignant tumors.

| End Points
The primary end point was the rate of complete remission (CR) + partial remission (PR) at day 168 (i.e., 24 weeks after the first rituximab infusion). CR was defined as the absence of new or established lesions for at least 8 weeks while the patient was receiving minimal therapy (prednisone ≤ 10 mg/day alone or with minimal adjuvant immunosuppressive therapy), whereas PR was defined as only transient lesions that resolved within 1 week, with no treatment or only topical corticosteroids without increasing the systemic prednisone dose. 6 The secondary end points were PDAI change, autoantibody titers against Dsg1 and Dsg3, and peripheral blood B-and T-cell counts. Safety events were recorded using MedDRA/J version 22.0, and their severity was graded according to the common terminology criteria for adverse events version 4.0. Other end points included pharmacokinetic analysis, expression rate of antidrug antibody (ADA; treatment-induced anti-rituximab antibody), and immunoglobulin levels.

| Statistical analysis
Based on previous studies, we expected a 50% remission rate at day 168 following rituximab administration and a spontaneous remission rate of 5% (threshold). 19 Under these assumptions, eight patients were needed to have ≥80% power to demonstrate that the true remission rate is greater than the threshold. In the protocol, the target number of patients for enrollment was 10, and it was stated that the enrollment would be continued after achieving the target until the end of the accrual period. All enrolled patients were included in the primary efficacy analysis population (full analysis set). The 95% confident intervals (CIs) for proportions were calculated using the Clopper-Pearson method. If the 95% lower confidence limit was greater than the prespecified threshold (5%), the null hypothesis (H0: remission rate = 5%) was rejected. Patients who withdraw from the study before day 168 (i.e., week 24) were considered as nonresponders. The significance level for tests was set at two-tailed 5%.
All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc). This study was registered with the UMIN Center Trials registry, number UMIN000024265.

| Patient characteristics
Between 2016 and 2018, a total of 20 patients (11 with PV, eight with PF, and one with PV/PF overlap) were enrolled in the study.
The clinical characteristics of the patients are presented in Table 1.
The PDAI score improved significantly at all study points after day 7 and remained at the lowest level (<2) between day 112 and 168 ( Figure 2b). In addition, serum anti-Dsg autoantibodies decreased over time. The anti-Dsg1 autoantibody titer decreased significantly from 309. 16 (Table S1). In addition, there were no differences in baseline characteristics between patients who did (n = 15) and did not (n = 5) achieve remission ( Figure S1). ADA was detected in serum samples collected on day 168 in four patients (20%). Of these patients, three achieved CR with low levels of peripheral blood B cells and anti-Dsg autoantibody levels throughout the study. In one patient, the serum rituximab level was below the detection limit on day 56, and the serum anti-Dsg autoantibody titers gradually increased although peripheral blood B cells were not detected until the end of the study.
Although the PDAI score decreased from 27 to 17 by day 168, the patient did not achieve remission. Among the 70 AEs, nine were possibly related to rituximab treatment (adverse drug reaction) and developed in five patients (25%).

| Adverse events
Eight of nine AEs were grade 1 or 2 and included palpitations (two events in one patient), pyrexia, hypogammaglobulinaemia, infusionrelated reaction (two events in one patient), elevated gammaglutamyl transferase level, and decreased immunoglobulin levels.
All AEs resolved without treatment, except for one patient who required IgG supplementation. Of the four SAEs, pneumonia was the only event considered to have a causal relationship with rituximab.

| DISCUSS ION
The present study was conducted in Japan to evaluate the efficacy and safety of rituximab in patients with refractory pemphigus who failed to achieve remission with conventional therapies. Overall, 75% of the 20 patients achieved remission with minimal therapy (55% CR and 20% PR). In addition, the PDAI score improved, serum anti-Dsg autoantibody levels decreased, and corticosteroid dose was successfully reduced from baseline in all cases. Although 85% of patients experienced AEs, most were not serious and four SAEs were appropriately handled, indicating no significant risk from rituximab   compared with the MMF group (15% and 6%). 24 There were no deaths in the rituximab group during the study, whereas one patient in the MMF group died of lung cancer. Based on recent reports, including the present study, rituximab appears to be safe for the treatment of pemphigus. In addition to infections, an important AE of rituximab therapy in patients with pemphigus is infusion-related reactions, which had an incidence of 5% (one of 20 patients) in the present study and was not observed in a prior study conducted in Japan in 2019. 19 On the other hand, infusion-related reactions occurred in 22% of patients (15 of 67) in the PEMPHIX trial. 24 The discrepancy between the results of the present and previous studies may be attributable to differences in the diagnostic criteria for infusion-related reactions.
However, a possible role of other factors, such as race, cannot be excluded and needs to be examined in future studies.
Similar to previous studies, rituximab treatment was effective for most patients with pemphigus in the present study. Because some patients did not achieve remission, it would be helpful to identify predictive factors for response to rituximab treatment.
In the present study, there were no differences in the baseline characteristics between patients who did and did not achieve remission, suggesting that rituximab may be effective for all patients with pemphigus. The inadequate efficacy of rituximab treatment in some patients may be related to the presence of ADA. 26,27 In our study, one patient demonstrated reduced blood rituximab level after ADA was detected and serum anti-Dsg autoantibody titers were increased, resulting in failure to achieve remission. In the PEMPHIX trial, ADA was detected in 31.7% of patients (20 of 63), with no differences in the efficacy and safety of rituximab between patients with and without ADA. 24 In our study, ADA was detected in 20% of patients (four of 20) and no differences were observed in the rate of remission between patients with and without ADA.
Although there are still improvements to be made, such as countermeasures against recurrence and relapse, we will need to mature in the careful use of rituximab as the mainstay of future pemphigus treatment strategies.

ACK N OWLED G M ENTS
We thank all the patients and their families, the investigators, and the site staff who participated in the present study. This work Co., Ltd. for the support having provided the investigational product, drug safety information, PK/HACA analyses, and financial support.
We thank Ms Hanako Mizuno and Mariko Okajima for the management of grants used for this study.