Research ArticleAutoimmunity

Long-Term Remissions of Severe Pemphigus After Rituximab Therapy Are Associated with Prolonged Failure of Desmoglein B Cell Response

See allHide authors and affiliations

Science Translational Medicine  06 Mar 2013:
Vol. 5, Issue 175, pp. 175ra30
DOI: 10.1126/scitranslmed.3005166

Abstract

Pemphigus is a severe blistering condition of the skin and mucosa caused by autoantibodies directed against desmogleins, which are a type of keratinocyte adhesion protein. B cell depletion by rituximab has short-term efficacy against pemphigus. We aimed to assess the long-term course of pemphigus patients after B cell depletion and to understand the immunological mechanisms that mediate long-lasting remissions. We evaluated the clinical course of 22 pemphigus patients treated with rituximab after a 79-month median follow-up and compared the anti-desmoglein B cell response and B and T lymphocyte subpopulations and repertoire between patients who achieved complete remission (CR) and those who had incomplete remission (IR). Thirteen patients (59%) experienced CR during the study, including 10 patients off treatment and 3 patients with prednisone doses <10 mg/day; 9 patients had IR. A marked increase was observed in the ratio of CD19+CD27 naïve B cells to CD19+CD27+ memory B cells. Indeed, patients in CR had a fourfold higher number of transitional B cells and interleukin-10–secreting regulatory B cells than those in IR. Furthermore, CR was associated with modification of the initial B cell repertoire and the disappearance of desmoglein-specific circulating immunoglobulin G–positive (IgG+) B lymphocytes, whereas a skewed B cell repertoire was observed in patients in IR. Thus, a blockage of B cell maturation, a prolonged repopulation with naïve B cells, and a delayed reappearance of memory B cells, which resulted in the disappearance of circulating desmoglein-specific IgG+ B lymphocytes, contribute to the long-lasting effectiveness of rituximab for treating pemphigus.

Introduction

Pemphigus is a rare organ-specific autoimmune disease (1) with an incidence estimated as 0.7 to 7 new cases per million per year (2, 3). It is characterized by the production of pathogenic autoantibodies directed against two desmosomal proteins involved in keratinocyte adhesion: desmoglein 1 and desmoglein 3 (4, 5). These antibodies are responsible for the disruption of desmosomes leading to the so-called acantholysis phenomenon, which results in the formation of skin and mucosal blisters (69). High doses of corticosteroids (CSs) remain the mainstay of first-line treatment, usually in conjunction with immunosuppressants (10). However, on standard therapy, relapses are frequent and long-term treatment is required, leading to high cumulative doses of CSs and severe CS side effects such as diabetes, hypertension, osteoporosis, or severe infections (1113).

Recently, new strategies based on B cell depletion have been proposed for the treatment of severe autoimmune disorders (1417). We previously reported the efficacy of a single cycle of rituximab in the treatment of severe types of pemphigus: 86% of patients achieved complete remission (CR) after a median follow-up of 34 months (18, 19). The short-term efficacy of rituximab in pemphigus patients has been confirmed in other series and case reports (2023). However, the long-term clinical course of pemphigus patients treated with rituximab has not been clearly assessed yet. In particular, the long-term evolution of the desmoglein 1– and desmoglein 3–specific B cell responses and their regulation after B cell reconstitution remain unknown. The evolution of the antigen-specific autoimmune B cell response in pemphigus may contribute to understand the mechanisms that maintain the long-term remission induced by rituximab in other autoimmune diseases, thus improving the management of these conditions and leading to an optimal use of rituximab.

The present study reports the long-term clinical course of the 22 patients with severe type of pemphigus initially treated with rituximab (18). To understand the mechanisms of the prolonged action of rituximab in pemphigus patients, we extensively assessed the evolution of the B cell subpopulations, B cell repertoire, and regulatory mechanisms affecting the desmoglein-specific B cell response over a 6-year follow-up period and compared these findings between patients who had a long-lasting CR after rituximab treatment and those who did not.

Results

Clinical course of patients

Baseline clinical characteristics of the 22 pemphigus patients are shown in Table 1 and table S1. Five patients with contraindications to CSs because of severe underlying medical conditions were treated with rituximab as first-line treatment, and 17 patients with recalcitrant or relapsing types of pemphigus were treated with rituximab as second- or third-line treatment. Twenty-one (95%) patients achieved disease control, with complete epithelialization of skin and/or mucosal lesions after a mean time of 3.1 months. Seventeen patients relapsed, including 8 patients who relapsed during the decrease of CS doses and 9 patients who relapsed after treatment withdrawal.

Table 1

Baseline characteristics and clinical course of the 22 pemphigus patients. CR, complete remission; IR, incomplete remission.

View this table:

Nine of the 17 relapsing patients were treated with a second cycle of rituximab, with 7 of them (78%: 95% confidence interval, 40 to 97%) achieving CR again. Nineteen of the 22 patients could be evaluated at the end of the study, after a 79-month median follow-up time. The three other patients died after a median survival of 29 months. One patient (aged 81 years) died from septicemia 18 months after rituximab treatment, whereas pemphigus was considered to be in incomplete remission (IR). The two other patients (aged 87 and 61 years) died from cardiovascular disease 29 and 51 months after rituximab treatment, respectively, and were in CR off therapy (n = 1) or on minimal therapy (n = 1). Finally, at the end of the study, 11 of the 19 living patients (58%: 95% confidence interval, 33 to 78%) achieved CR, including 9 patients in CR off treatment and 2 patients in CR on minimal treatment. The eight remaining patients still had prednisone doses higher than 10 mg/day due to active disease and hereafter are identified as patients in IR. Despite a limited number of patients in this group, 5 of 5 patients (100%) treated with rituximab as first-line treatment achieved CR off therapy at their last assessment time point compared with 5 of 17 patients (29%) who were treated with rituximab as second- or third-line treatment (P = 0.009).

Autoantibodies against desmoglein 1 and desmoglein 3

The changes of anti–desmoglein 1 and anti–desmoglein 3 antibody enzyme-linked immunosorbent assay (ELISA) values after rituximab treatment are shown in Fig. 1. We observed a major decrease in anti–desmoglein 1 and anti–desmoglein 3 ELISA values within the first 6 months after rituximab treatment that paralleled the improvement of skin and mucosal lesions. Mean ELISA values of anti–desmoglein 1 and anti–desmoglein 3 antibodies decreased from 112 ± 14 IU and 152 ± 17 IU at baseline to 23 ± 10 IU and 71 ± 19 IU at month 6, respectively (P = 0.001 and P = 0.009), and 23 ± 18 IU and 115 ± 29 IU 24 months after rituximab treatment (P = 0.003 and P = 0.02 versus baseline). Then, mean anti–desmoglein 1 and anti–desmoglein 3 ELISA values did not vary significantly from month 24 to month 79 after rituximab treatment, when they were measured at 42 ± 14 IU and 92 ± 21 IU (P = 1.00 and P = 0.09, respectively, month 24 versus month 79). The modification of anti–desmoglein 1 and anti–desmoglein 3 antibody ELISA values depending on clinical course of patients is shown in fig. S1. All patients who achieved CR off treatment and did not relapse during follow-up had a rapid, marked, and long-lasting decrease in anti–desmoglein 1 and anti–desmoglein 3 antibodies that remained undetectable up to 6 years after initial rituximab infusions (fig. S1, A and B). All the patients who relapsed were in IR at the end of the study and had either persistently high ELISA values of anti–desmoglein 1 and/or anti–desmoglein 3 antibodies or a reincrease of the corresponding antibodies before the relapse (fig. S1, C to F). However, some pemphigus vulgaris patients in CR of their mucosal lesions still had elevated anti–desmoglein 3 antibody ELISA values (fig. S1C).

Fig. 1

Evolution of anti–desmoglein (Dsg) 1 (open circles) and anti–desmoglein 3 (black circles) antibody ELISA values in sera of pemphigus patients after rituximab therapy. The dashed line represents the cutoff values proposed by the manufacturer for anti–desmoglein 1 and anti–desmoglein 3 antibodies. *P < 0.05; **P < 0.01; ***P < 0.001 versus baseline for desmoglein 3 (above) and desmoglein 1 (below).

Total immunoglobulin levels and antimicrobial antibody titers

Total immunoglobulin M (IgM) serum levels progressively decreased after rituximab treatment from 1.6 ± 0.2 g/liter at baseline to 0.9 ± 0.1 g/liter at month 24 (P = 0.007) and 1.2 ± 0.1 g/liter at month 79 (P = 0.01 versus baseline). On the contrary, total IgG serum levels did not vary significantly during the follow-up period (P = 0.2) (fig. S2A). Anti-TT (tetanus toxoid) IgG titers did not vary significantly from baseline to month 79 (P = 0.9). Anti-PCP (pneumococcal capsular polysaccharide) antibody titers did not vary from baseline to month 12 after rituximab infusions and were even measured at a higher value than at baseline at the end of the study (fig. S2B) (P = 0.02).

Effect of rituximab on B cell count and phenotype

A marked depletion of B lymphocytes was observed from 330 × 106/liter at baseline to 26 × 106/liter 6 months after rituximab infusion (P < 0.001). B cell reconstitution began between month 6 and month 9 after rituximab treatment, with all the patients having B lymphocyte counts between 1 × 106/liter and 140 × 106/liter (Fig. 2A). Indeed, the number of CD19+ B lymphocytes measured at month 24 was still much lower than at baseline (144 × 106/liter versus 330 × 106/liter, P = 0.01). Circulating B cells then progressively increased up to 215 × 106/liter at month 79, although this number remained significantly lower than at baseline (P = 0.02) (Fig. 2A). During B cell reconstitution, 9 months after rituximab infusion, an average of 24% of B cells expressed a CD24highCD38high phenotype, which is characteristic of transitional B cells migrating from the bone marrow to the periphery (Fig. 2B). The frequency of CD24highCD38high transitional B cells decreased progressively during the years after rituximab treatment (Fig. 2B). However, patients in CR had a 4.4-fold higher frequency of CD24highCD38high transitional B cells than those in IR at the end of the study (8.1 ± 1.3% versus 1.9 ± 0.6%, P = 0.002) (Fig. 3). The proportion of CD19+CD27 naïve B lymphocytes and CD19+CD27+ memory B lymphocytes after rituximab treatment is shown in Fig. 2 (C and D). The ratio between CD19+CD27 naïve B lymphocytes and CD19+CD27+ memory B lymphocytes increased more than fivefold from 3.6 at baseline to 18.6 at month 24 after rituximab treatment (P < 0.001). In addition, the naïve/memory B lymphocyte ratio did not vary significantly from month 24 to month 79 (P = 0.08). We did not observe significant differences in the proportion of memory and naïve B cell subpopulations between patients in CR and those in IR at the end of the study (P = 0.9 and P = 0.2, respectively) (Fig. 3).

Fig. 2

Evolution of peripheral blood B cell populations in pemphigus patients after rituximab therapy. (A) CD19+ B lymphocytes. (B) CD24highCD38high transitional B lymphocytes. (C) CD19+CD27 naïve B lymphocytes. (D) CD19+CD27+ memory B lymphocytes. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 3

Frequency of peripheral blood CD19+CD27 naïve B cells, CD19+CD27+ memory B cells, and CD24highCD38high transitional B cells evaluated in pemphigus patients 79 months after rituximab therapy depending on the patient’s clinical status. NP, newly diagnosed untreated pemphigus patients (white columns); IR, pemphigus patients in IR after rituximab therapy (black columns); CR, pemphigus patients in CR after rituximab therapy (striped columns). *P < 0.05; **P < 0.01; ***P < 0.001. ns, not significant.

Desmoglein 1– and desmoglein 3–specific B cells

The number of desmoglein 1– and desmoglein 3–specific circulating B cells was evaluated in 18 of the 19 living patients. The remaining patient had a very low number of circulating B cells at the end of the study, which did not allow the performance of this analysis. Flow cytometry was performed after staining desmoglein-specific B cells with fluorescently tagged recombinant desmoglein 1 and desmoglein 3.

Desmoglein 1– and desmoglein 3–specific circulating IgG+ B lymphocytes were detected in 90% of blood samples collected at baseline from pemphigus patients, with a mean count of 48 ± 20/107 B cells (Fig. 4). Conversely, negative control experiments using blood samples from healthy individuals only detected 2 ± 1 desmoglein-specific IgG+ B lymphocytes/107 B cells in 5 of the 10 samples tested, which corresponded to the cutoff value for these experiments. Finally, desmoglein 1– and desmoglein 3–specific IgG+ B lymphocytes were detected in blood samples from 5 of the 10 pemphigus patients in CR and 5 of the 8 patients in IR at a mean count of 11 ± 6/107 B cells and 15 ± 9/107 B cells, respectively, which was between five- and sevenfold higher than the mean count in samples from healthy individuals, and between three- and fourfold lower than the count in samples collected from pemphigus patients at baseline (P = 0.04 and P = 0.01, respectively).

Fig. 4

Number of peripheral blood desmoglein 1–specific (white columns) and desmoglein 3–specific (black columns) IgG+ and IgM+ B lymphocytes evaluated by flow cytometry in pemphigus patients at baseline and 79 months after rituximab therapy. HD, healthy individuals (negative control experiments). *P < 0.05; **P < 0.01; ***P < 0.001.

Desmoglein 1– and desmoglein 3–specific circulating IgM+ B cells were almost consistently detected in the blood of pemphigus patients from 90% of samples collected from patients in CR to 100% in samples collected from patients with active disease at baseline or in IR, whereas these cells were detected in 40% of the 10 negative controls from healthy individuals. Mean counts of desmoglein 1– and desmoglein 3–specific circulating IgM+ B lymphocytes in samples collected in pemphigus patients at baseline and in patients in IR were not statistically different: 158 ± 51/107 B cells and 147 ± 62/107 B cells, respectively (P = 0.8) (Fig. 4). Corresponding values were twofold lower in samples in pemphigus patients in CR, 71 ± 18/107 B cells (P = 0.06 versus baseline), and were very low in negative control samples from healthy individuals, 3 ± 3/107 B cells (P ≤ 0.001 versus pemphigus patients at baseline).

Regulatory T and B cells

Because B cells with a regulatory phenotype have been shown to be impaired in some autoimmune diseases (24, 25), we analyzed the ability of B cells from pemphigus patients to produce interleukin-10 (IL-10) under in vitro stimulation. Pemphigus patients who experienced CR after rituximab treatment had a higher frequency of IL-10–secreting cells and secreted higher amounts of IL-10 than either patients with active lesions at baseline or patients who experienced IR after rituximab treatment (8% versus 5 and 4%, P = 0.02, and 2278 pg/ml versus 920 and 1230 pg/ml, P = 0.02, respectively) (Fig. 5, A and B).

Fig. 5

IL-10–producing regulatory B cells among purified CD19+ blood B cells from pemphigus patients at baseline and 79 months after rituximab therapy. (A) Frequency of IL-10+ B cells among purified CD19+ B cells after in vitro stimulation with CpG + anti-μ antibodies. (B) IL-10 levels measured by ELISA in culture supernatants. (C) Frequency of CD24highCD38high transitional B cells among IL-10–producing regulatory B cells. *P < 0.05; **P < 0.01; ***P < 0.001.

Patients in CR had a more than fourfold higher frequency of CD24highCD38high transitional B cells than those in IR, which suggested that these transitional B cells might have a regulatory activity (25). We therefore assessed the frequency of IL-10+ regulatory B cells among CD24highCD38high transitional B cells. Patients in CR after rituximab treatment had a higher percentage of CD24highCD38high IL-10+ B cells than those who experienced IR (64 versus 39%, P = 0.01) (Fig. 5C). We did not observe any difference in the frequency of CD4+CD25highFoxp3+ T cells between patients in CR and those with active disease after rituximab treatment (fig. S3).

B cell repertoire analysis

Using the immunoscope method, we performed an extensive B cell repertoire analysis before and 6 years after rituximab treatment on three representative patients displaying persistent CR off therapy, relapse before achieving CR, and IR, respectively.

At baseline, before rituximab, a skewed repertoire and B cell expansions were found in all three patients. After rituximab therapy, B cell reconstitution was associated with new repertoire illustrated by new immunoscope profiles. The B cell immunoscope profile from the patient in CR off therapy became Gaussian again, whereas new expansions were observed in the repertoire from the relapsing patient, suggesting the reappearance of a new skewed repertoire. Conversely, the B cell repertoire reconstitution was associated with recurrent expansions in the patient who only had IR after rituximab treatment (fig. S4).

Discussion

Here, we demonstrate that rituximab can induce a long-term CR in a substantial proportion of patients with severe types of pemphigus. Indeed, 11 of the 19 living patients (58%) were in CR 6 years after the initial infusion of rituximab, including 9 patients (47%) in CR off treatment. Notably, all five patients treated with rituximab as the first-line treatment achieved long-lasting CR off treatment. Moreover, even though these patients probably had a less severe type of pemphigus than the 17 other ones, it must be underlined that these patients achieved CR of their disease, without any systemic CS treatment.

In accordance with the long-lasting clinical effect of rituximab, the B cell depletion induced by the drug lasted for many years: The mean number of CD19+ B lymphocytes measured 6 years after rituximab treatment remained much lower than at baseline, which is somewhat surprising considering the short half-life of rituximab (26, 27). It is likely that this long-lasting effect of rituximab was related to the B cell ontogeny and the major changes in B lymphocyte subpopulations observed after rituximab treatment, which included (i) an increase of CD19+CD27 naïve B lymphocytes; (ii) a decrease of CD19+CD27+ memory B lymphocytes, resulting in a marked and long-lasting change of the naïve/memory B cell ratio after rituximab treatment; (iii) the expansion of transitional B cells; and (iv) the expansion of IL-10–secreting regulatory B cells in patients who had CR.

B cell reconstitution that occurs after B cell depletion therapy is initially composed of reemerging naïve B cells, which secondarily give rise to the memory B cell compartment (28, 29). The reconstitution of the memory B cell compartment is markedly delayed after rituximab treatment (29, 30). Moreover, relapses in patients with rheumatoid arthritis have been shown to occur preferentially in patients who had higher number of circulating memory B cells during B cell repopulation (30, 31). Here, it is likely that the marked and long-lasting modification of the naïve/memory B cell ratio may have participated in the long-lasting therapeutic effect of rituximab in pemphigus patients. Indeed, we observed that patients in CR had a higher mean number of CD24highCD38high transitional B cells than those in IR with persistent active disease. These data are consistent with previous observations in patients with systemic lupus erythematosus (SLE) showing that the expansion of the transitional B cell subset was associated with a better clinical outcome and prolonged clinical remissions. Thus, the transitional B cell compartment that develops after B cell depletion therapy might have censoring effect on the autoimmune reaction (30).

Because transitional B cells have been shown to have regulatory properties in humans through the production of IL-10 (25), we evaluated the ability of circulating B lymphocytes from pemphigus patients to produce this cytokine. Pemphigus patients who were in CR after rituximab treatment had a higher number of IL-10–secreting B cells than those in IR. Most of the IL-10–secreting B cells were found within the transitional B cell subset, suggesting that transitional B cells may contribute to the regulatory B cell compartment. Because the functional impairment of these transitional B cells with regulatory properties has been associated with the occurrence of autoimmune diseases such as SLE (25), it is likely that these cells could be involved in the long-lasting decrease of the anti-desmoglein antibody response in pemphigus patients. Indeed, whereas a variable number of desmoglein-specific circulating IgG+ B lymphocytes could be detected at baseline in pemphigus patients and in those who experienced IR after rituximab treatment, these desmoglein-specific IgG+ B lymphocytes were barely detectable in patients who achieved long-lasting CR after rituximab. These findings are in accordance with the marked decrease of anti–desmoglein 1 and anti–desmoglein 3 IgG antibodies that we observed in most remitted patients. In contrast, desmoglein-specific circulating IgM+ B cells could be detected at different stages of pemphigus, with a close correlation between the number of these cells and disease activity. The almost complete disappearance of desmoglein-specific circulating IgG+ B lymphocytes and that of serum anti-desmoglein antibodies in remitted patients, whereas desmoglein-specific IgM+ B lymphocytes were still detectable, is likely explained by the blockage of the IgM-to-IgG switch, which might be favored by regulatory B cells (Fig. 6). Notably, the function of IL-10–secreting B cells was initially described in B cell depletion experiments in mouse models of experimental autoimmune encephalitis (32). In humans, regulatory B cells have been shown to decrease CD4+ T cell effector responses (33) and to down-regulate CD4+ T cell help (34).

Fig. 6

Proposed model of rituximab action in pemphigus patients. Step 1: Interaction between autoreactive T and B cell generates a positive feedback loop that sustains the differentiation in autoantibody-producing plasma cells, leading to acantholysis in pemphigus patients. Step 2: Rituximab treatment induces CD20+ B cell depletion, decreasing the T and B cell cooperation and blocking the generation of memory B lymphocytes and consequently new autoantibody-secreting plasma cells, leading to a decrease of circulating autoantibody titers. Step 3: B cell repopulation occurs between 6 and 9 months after rituximab treatment. Newly emerging B cells, with a naïve and transitional phenotype, give rise to functionally competent IL-10–secreting B cells that could inhibit T cell help, leading to a decrease of anti-desmoglein circulating B cells and thus a decrease of anti-desmoglein circulating antibodies (Ab).

Because desmoglein-specific CD4+ T cells have been shown to decrease after rituximab therapy (35), one could hypothesize that the regulatory B cells that we observed after rituximab treatment could be at least partly responsible for the down-regulation of desmoglein-specific CD4+ T cells through the secretion of IL-10 (Fig. 6).

We observed a discrepancy in patients in IR after rituximab treatment between the persistence of high levels of serum anti-desmoglein IgG antibodies and the barely detectable circulating desmoglein-specific IgG+ B lymphocytes. The persistent secretion of anti-desmoglein IgG antibodies in some patients might be attributed to long-lived plasma cells or to noncirculating long-lived memory IgG+ B cells, which transform into plasma cells and are not targeted by anti-CD20 therapy (36). Such a mechanism might also be involved in the persistence of anti-infectious antibodies after rituximab treatment. Indeed, the nonresponse to anti-CD20 therapy in rheumatoid arthritis patients has been recently associated with the presence of elevated levels of IgJ, a marker for antibody-secreting plasmablasts (37). Finally, the reconstitution of a new Gaussian B cell repertoire in one patient who achieved persistent CR after rituximab, whereas recurrent and/or new expansions in the B cell repertoire were observed in patients in IR, is in accordance with the disappearance of circulating anti-desmoglein antibodies in most of the remitted patients. High titers were still detected in patients with active disease, which suggests that the appearance of new expended peaks in these latter patients could correspond to the reemergence of autoreactive B cell clones.

The main limitations of this study are, first, the limited number of patients included with both patients treated with rituximab as first-line treatment and patients with CS recalcitrant and relapsing types of pemphigus. In particular, it would have been of particular interest to assess the course of patients treated with rituximab as first-line treatment in a higher number of cases. Indeed, the marked improvement observed in these patients might argue for an early use of rituximab in the course of pemphigus, which would allow using lower doses of CSs. However, the potential benefits of this first-line use of rituximab should be balanced with the possible severe side effects of rituximab. Here, one patient died from septicemia, and another had pyelonephritis. However, there are also severe side effects observed with the standard CS regimen, which are responsible for a significant morbidity and even mortality in pemphigus patients (11). Indeed, apart from the two patients who died from cardiovascular disease 29 and 51 months after the initial infusion of rituximab, which is not thought to be related to rituximab treatment, we did not observe any other delayed severe side effects during this extension phase of our initial study. In particular, no severe infections were observed despite the long-lasting B cell depletion. This observation is consistent with the persistence of an anti-infectious antibody response, as suggested by the sustained levels of the anti-TT and anti-PCP antibodies throughout the study. Another limitation of the study is that some of the relapsing patients were retreated with a second cycle of rituximab, whereas others only had a reincrease of CS doses. The rather high relapse rate (59%) during the years after rituximab treatment and the marked efficacy of a second course of rituximab in seven of the nine retreated patients argue for using a maintenance therapy with rituximab to avoid the occurrence of relapses as previously suggested in severe cases of systemic vasculitis (3840).

In conclusion, this study provides new insights in the understanding of the mechanism of the long-lasting effect of rituximab in pemphigus patients. It suggests that rituximab induces a prolonged and continuous repopulation with naïve B cells with a new repertoire that occurs after the B cell depletion, with a marked delay in the reappearance of memory B cells. This inhibition of B cell maturation is associated with a blockage of the IgM-to-IgG class switching process. The emergence of transitional B cells with regulatory properties is also likely involved in the prolonged immunosuppressive effect of rituximab and may lead to the marked decrease of desmoglein-specific IgG+ B cells and anti-desmoglein IgG antibodies in patients with CR. This study demonstrates that rituximab therapy can not only induce prolonged clinical remissions with a marked decrease of CS doses but also, in some cases, provide a functional cure for pemphigus, a chronic antibody-mediated autoimmune disease.

Materials and Methods

Patients

The 22 pemphigus patients included in our initial clinical trial (ClinicalTrials.gov number, NCT00213512) and treated with a single cycle of rituximab (18) were reassessed for clinical and immunological parameters after a minimum of 6 years after their inclusion (ClinicalTrials.gov number, NCT01299857). The present report on 22 patients includes one PNP patient who was enrolled in our first study (18), but whose data were not presented in the initial publication on the request of the editor. Peripheral blood mononuclear cells (PBMCs) were collected from the 19 patients still alive at the time of the present study between 72 and 93 months after the initial infusion of rituximab. Additionally, PBMCs were collected from 10 patients with a newly diagnosed pemphigus before any treatment. These patients are included in a clinical trial that is currently being performed to assess the usefulness of rituximab as first-line treatment (ClinicalTrials.gov number, NCT00784589). These PBMCs were used as positive controls in the experiments that assessed the number of desmoglein 1– and desmoglein 3–specific B cells at baseline, because these experiments needed a high number of lymphocytes than were available in sufficient quantity in samples collected at the baseline evaluation of our initial clinical trial (18). Additionally, PBMCs from 10 healthy blood donors were provided in buffy coats from the French Blood Agency for use as negative controls. This study was approved by the Ethics Committee of the North West area II in France and conducted according to the Declaration of Helsinki principles. Written informed consent was obtained from each patient.

New clinical examination was performed, and treatment history, including time of relapse and retreatment with rituximab, was collected in all the patients at a new late time point, at least 6 years after their inclusion in the initial study. Endpoints were defined according to the Consensus Statement for pemphigus (41). “CR off therapy” was defined as the absence of new or established lesions, as well as the healing of all skin and mucosal lesions with complete epithelialization when patients were off all systemic therapy. “CR on minimal therapy” was defined as the absence of new or established lesions, as well as the healing of all skin and mucosal lesions with complete epithelialization, when patients were still receiving minimal therapy—prednisone less than or equal to 10 mg/day and/or minimal adjuvant therapy for at least 2 months. Patients still receiving prednisone doses higher than 10 mg/day at the end of the study were identified as being IR, even if they had no skin or mucosal lesions at the time of examination. “Relapse” was defined as the occurrence of new skin blisters or oral erosions after having achieved disease control.

Immunological evaluation

A blood sample was collected to assess anti–desmoglein 1 and anti–desmoglein 3 antibody titers, total IgM and IgG levels, B and T lymphocyte counts and cell subset analysis, IL-10–secreting regulatory B cells, desmoglein 1– and desmoglein 3–specific B cells, and B cell repertoire.

Titers of IgG antibodies against desmoglein 1 and desmoglein 3 were measured with a desmoglein ELISA test (MESACUP Desmoglein Test, MBL Medical and Biological Laboratories) with 1:100 diluted serum. Serum levels of anti-TT and anti-PCP IgG antibodies were determined by ELISA as previously described (19).

The phenotype of PBMCs was determined by six-color flow cytometry with the murine monoclonal antibodies (mAbs) against CD3, CD4, CD8, CD5, CD19, CD20, CD21, CD22, CD23, CD24, CD27, CD38, CD56, and CD86 (Beckman Coulter and BD Biosciences). Regulatory T cells were identified with a combination of CD4, CD25, and Foxp3 antibodies according to the manufacturer’s instructions (eBioscience).

IL-10–secreting B cell analysis

Purified B cells (5 × 105) were cultured for 48 hours with CpG-B 2006 (3 μg/ml) and anti-human IgG+IgA+IgM (anti-Ig) antibody (Jackson ImmunoResearch) (20 μg/ml) in 1 ml of complete medium in 24-well flat-bottom tissue culture plates. Cells were first stained with anti-CD19, anti-CD24, anti-CD27, anti-CD38, and anti-CD5 mAb, and then fixed and permeabilized, followed by intracellular staining with anti-human IL-10 mAb (B-T10) or mouse IgG1 isotype control (Miltenyi Biotec). IL-10 was quantified in culture supernatant by ELISA (R&D Systems). All assays were carried out with duplicate samples.

Desmoglein-specific B cell detection

Purified mononuclear cells (20 × 106) were incubated for 1 hour at room temperature with histidine-tagged desmoglein 1 or desmoglein 3 (35). After washing, anti-histidine coupled with phycoerythrin (R&D Systems) was used to identify desmoglein-stained cells. B cells were characterized with anti-human CD19 and anti-human IgG and IgM antibodies (BD Biosciences). The number of desmoglein 1– and desmoglein 3–specific cells per million of cells was then determined.

B cell repertoire analysis with immunoscope method

IgG and IgM repertoire was characterized at the molecular level in B cell PBMCs. Complementary DNA was prepared from 20 × 106 PBMCs. VH gene usage and CDR3 analysis were performed with the “immunoscope” method coupled with real-time polymerase chain reaction (PCR) to provide quantitative information on VH usage. Briefly, PCRs were performed by combining a primer and a specific fluorophore-labeled probe for the constant region CH with one of eight primers covering the different VH1 to VH7 genes (42). Reactions were carried out with TaqMan 7300 (Applied Biosystems) and standard reagents from Applied Biosystems. Run-off reactions with a nested fluorescent primer specific to the constant region gene were then performed on PCR products. Fluorescent products were separated and analyzed on an ABI Prism 3730 DNA analyzer to determine CDR3 lengths.

Statistical analysis

Aggregate data are presented as means ± SEM. Prism software was used for statistical analysis. Analyses were performed with Fisher’s exact test, Wilcoxon’s matched pairs signed rank test, Kruskal-Wallis test with Dunn’s posttest, and the Mann-Whitney test for comparison of CD24highCD38high frequencies. A P value of less than 0.05 was considered significant for all analyses.

Supplementary Materials

www.sciencetranslationalmedicine.org/cgi/content/full/5/175/175ra30/DC1

Fig. S1. Evolution of anti–desmoglein 1 and anti–desmoglein 3 antibody ELISA values after rituximab therapy according to the clinical type of pemphigus [pemphigus vulgaris (PV), pemphigus foliaceus (PF), or paraneoplastic pemphigus (PNP)] and the clinical course of the disease.

Fig. S2. Evolution of total and anti-infectious immunoglobulin levels in pemphigus patients after rituximab therapy.

Fig. S3. Frequency of peripheral regulatory T cells evaluated in pemphigus patients 79 months after rituximab therapy.

Fig. S4. Analysis of the IgM and IgG B cell repertoire of the VH B cell profiles using the immunoscope technique at baseline and after rituximab therapy in three representative pemphigus patients.

Table S1. Baseline characteristics and follow-up of the 22 pemphigus patients treated with rituximab.

References and Notes

  1. Acknowledgments: We thank S. Duvert-Lehembre and C. Leblond from the “National Reference Center for autoimmune bullous diseases” and Vincent Ferranti for technical assistance and support, and D. Gilbert and D. F. Murrell for critical review of the manuscript. We also thank E. Houivet for validation of statistical analysis. We are grateful to N. Sabourin-Gibbs (Rouen University Hospital) for her help in editing the manuscript. Funding: Financial support for the study was provided by the “French Society of Dermatology.” Author contributions: N.C., D.P., P.J., and P.M. designed the study; N.C. performed the experiments; D.P. obtained funding; N.C., D.P., P.J., and P.M. analyzed clinical data; N.C., D.P., P.J., and P.M. wrote the manuscript; A.L. and B.L. performed immunoscope analysis; F. Caillot, S.C., S.L.C., M.M.-V., and S.J. performed and analyzed flow cytometry experiments; R.E. and M.H. produced desmoglein 1 and desmoglein 3 recombinant proteins; B.L.M. performed anti-TT and anti-PCP ELISA; C.B., P.B., F. Caux, C. Prost, E.D., M.-S.D., B.D., N.F., S.I.-H.-O., O.C., C. Pauwels, C. Picard, J.-C.R., M.S., E.T.-B., I.T., and M.D. collected clinical data. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article