Author Affiliations: Departments of Dermatology (Drs Lunardon, Fett, Stanley, Werth, and Payne), Hematology/Oncology (Ms Tsai and Dr Tsai), and Biostatistics (Dr Propert), University of Pennsylvania, and Veterans Affairs Medical Center (Drs Fett and Werth), Philadelphia.
Background We conducted a retrospective study of patients with pemphigus vulgaris (n = 24) and foliaceus (n = 7) treated with adjuvant rituximab to determine efficacy and adverse events. The end point for efficacy was complete remission of disease taking no or minimal therapy.
Observations Eighteen patients (58%) achieved the study end point. Of these, 13 patients achieved complete remission off systemic therapy. Patients achieving the study end point had a median disease duration before rituximab therapy of 19 months vs 86 months in those not achieving the end point (P = .01). For the 18 patients achieving the end point, the median (SD) duration of remission was 19 (2) months. Eight of these 18 patients (44%) relapsed from 6 to 17 months after treatment. Serious adverse events attributed to rituximab treatment (osteomyelitis or phlegmon) occurred in 2 patients (6%). In paired serum samples from 10 patients before and after rituximab treatment, the percent change in serum desmoglein index value (median, −80%) was unrelated to the percent change in pneumococcal antibodies (median, +8%) (Spearman rank correlation coefficient r = −0.2).
Conclusions Patients treated with rituximab earlier in the course of disease may have better outcomes. A discussion of rituximab's mechanism of action supports the rationale for early therapy. Prospective clinical studies are necessary to substantiate this observation.
Pemphigus is a group of rare but potentially fatal dermatologic conditions in which autoantibodies against epidermal adhesion proteins known as desmogleins (Dsgs) cause blisters in the mucous membranes and/or skin.1 Pemphigus foliaceus is characterized by superficial skin blisters caused by autoantibodies to Dsg1. Pemphigus vulgaris demonstrates deeper blisters in the suprabasal layer of the epidermis. Pemphigus vulgaris is typically associated with autoantibodies to Dsg3 in mucosal-dominant disease and to Dsg3 plus Dsg1 in mucocutaneous disease. Widespread skin and mucosal blistering can lead to fatal complications, including malnutrition, dehydration, and sepsis. Current therapies aim to decrease antibody production by generally suppressing the immune system. However, chronic immune suppression also risks complications, including fatal infection and secondary cancers. The challenge in pemphigus treatment, therefore, is to balance risk of disease with risk of therapy.
Rituximab, an anti-CD20 monoclonal antibody, has shown impressive 86% and 100% efficacy in 2 prospective trials of 21 and 11 pemphigus patients, respectively, although 1 of 21 patients (5%) experienced fatal septicemia.2,3 Other prospective and retrospective analyses have reported variable efficacy and serious infections, raising debate as to the appropriate clinical use of rituximab in pemphigus.4- 9 To begin to form hypotheses as to which patient characteristics may lead to better clinical response, we conducted a retrospective study of all pemphigus patients examined at our medical center and treated with rituximab (N = 31). Herein, we report the efficacy and adverse effects of their treatment. Our observations indicate that rituximab therapy early in the course of disease is significantly more likely to lead to complete disease remission on no or minimal systemic therapy.
We conducted a retrospective single-center study of all pemphigus patients examined at the University of Pennsylvania and treated with rituximab from March 1, 2005, through November 30, 2010, with follow-up through January 31, 2012. All studies were approved by the institutional review board. Diagnoses were confirmed by clinical presentation, histology, and immunofluorescence or enzyme-linked immunosorbent assays.
The study population included 24 patients (77%) with pemphigus vulgaris and 7 (23%) with pemphigus foliaceus, with 18 women (58%) and 13 men (42%) (Table 1). The median (range) age was 50 (26-86) years. The median (range) disease duration before rituximab treatment was 41 (3-234) months. Before rituximab treatment, all patients were treated with systemic therapies, most commonly corticosteroids with or without mycophenolate mofetil or azathioprine. Because of a lack of response, contraindications, or adverse effects with these therapies, patients were treated with adjuvant rituximab. All patients were on systemic immunosuppressive therapies at the time of first rituximab infusion, indicated by a footnote in Table 1. Although disease severity scores were not prospectively determined, all patients had severe and/or refractory disease that prompted rituximab adjuvant therapy.
Clinical response to rituximab was assessed in accordance with the consensus statements for pemphigus.10 Outcomes were classified as complete remission (absence of skin and mucosal lesions for ≥2 months) and partial remission (transient new lesions that heal within 1 week). Outcomes could be achieved off therapy, on minimal therapy (with doses defined for each medication), or on doses greater than minimal therapy. The end point of the study was complete remission of disease on no or minimal therapy.
After rituximab therapy, all 31 patients experienced clinical improvement of disease activity. Overall, 18 patients (58%) achieved complete remission of disease on no or minimal systemic therapy, with similar efficacy in pemphigus vulgaris and pemphigus foliaceus (58% and 57%, respectively) (Table 1). Of these 18 patients, 13 achieved complete remission off all systemic therapy. Analysis of demographic data indicates that patients achieving complete remission on no or minimal therapy had a median (interquartile range) disease duration before rituximab therapy of 19 (14-41) months compared with 86 (55-144) months in patients not achieving the study end point (P = .01 by Wilcoxon rank sum test) (Figure 1). Study end points were not significantly associated with sex, age, or dosing regimen.
Figure 1. Complete remission of disease on no or minimal therapy is associated with use of rituximab earlier in the course of disease. Median (interquartile range) disease duration before rituximab therapy was 19 (14-41) months for patients achieving the study end point compared with 86 (55-144) months in those not achieving the study end point (P = .01 by Wilcoxon rank sum test). The horizontal line in the middle of each box indicates the median, while the top and bottom borders of the box mark the 75th and 25th percentiles, respectively. The whiskers above and below the box mark the 90th and 10th percentiles. The point beyond the whiskers is an outlier beyond the 90th percentile.
Of the 18 patients achieving the end point, 12 required a single cycle of rituximab, 3 required 2 cycles of rituximab, and 3 required 3 cycles of rituximab to attain complete remission. The individual rituximab regimens and response are further detailed in Table 2. Rituximab was administered using either an oncologic regimen consisting of 4 weekly infusions of 375 mg/m2 or a rheumatologic regimen consisting of two 1000-mg infusions separated by 2 weeks. Two patients did not receive full infusions because of scheduling conflict (patient PV15) and hospitalization for syncope due to anemia (patient PV24). Patients received additional cycles of rituximab for 1 of 2 reasons: to improve outcome or to treat relapse. Responses to rituximab retreatment for the first indication were mixed. Of the 15 patients retreated to improve outcome, 8 improved, achieving complete remission on no or minimal therapy, whereas the other 7 either were not able to taper to minimal therapy or did taper to no or minimal therapy but disease recurred within 2 months. One patient receiving 6 cycles of rituximab achieved partial remission of disease on topical corticosteroids but did not experience further improvement with repeated infusions.
The second indication for additional cycles of rituximab was to treat relapse. Of the 18 patients achieving the study end point, 10 (56%) did not relapse and have remained in complete remission, with follow-up from 8 to 28 months after rituximab infusion. The median (SD) relapse-free remission time was 19 (2) months (Figure 2). Eight of the 18 patients (44%) relapsed 6 to 17 months after rituximab infusion (indicated in parentheses in Table 2). Of the 8 patients who relapsed, 4 were retreated with rituximab; 2 attained complete remission off therapy, 1 attained partial remission off therapy, and the other attained complete remission on medication doses greater than minimal therapy. Of the other 4 patients, 2 were recently retreated with rituximab with follow-up pending; 1 is being treated with prednisone because of concurrent pregnancy, and the final patient is being observed while in partial remission off therapy.
Figure 2. Kaplan-Meier plot of time to relapse after rituximab infusion in patients achieving complete remission (n = 18). Circles indicate duration of follow-up for patients remaining in complete remission (n = 10). The median (SD) duration of remission was 19 (2) months. Dotted lines indicate the 95% CI.
Although it was not a primary study end point, 5 patients achieved partial remission of disease off therapy after a single cycle of rituximab with a median (range) duration of response of 21 (13-80) months. Partial remission is defined as the presence of transient lesions that heal within 1 week without use of any medication, including topical corticosteroids. All 5 patients considered this a favorable outcome and did not desire retreatment. Two patients relapsed after 16 and 21 months and were retreated with rituximab, again achieving partial remission off therapy.
Serious adverse events (SAEs) were identified according to US Food and Drug Administration definitions, including events that are fatal or life-threatening, require hospitalization, or cause disability or permanent damage. Before rituximab therapy, 5 patients (16%) experienced 5 SAEs attributed to disease or its treatment, including pulmonary embolism, avascular necrosis, suicidal ideation, psychosis with suicide attempt, and severe blistering with multiorganism superinfection. These SAEs were observed during the median 41-month period between disease onset and the start of rituximab therapy. During the median 28 months of follow-up after rituximab infusion, 5 patients experienced 6 SAEs, 2 of which were attributed to rituximab. One patient (patient PF3) developed osteomyelitis 4 months after rituximab infusion while on 10 mg prednisone daily. The treatment successfully tapered to 1 mg prednisone daily and the patient remained in complete remission. However, the patient subsequently died of a fatal gastrointestinal bleed, attributed to higher-dose corticosteroids prescribed for complications after a corneal transplant. One patient (patient PV20) was treated with 6 cycles of rituximab, achieved partial remission on topical corticosteroids, and developed perirectal phlegmon and intrapelvic abscesses requiring surgical debridement. Other SAEs occurring after rituximab infusion but not directly attributed to rituximab included hospitalization for syncope due to anemia, prostate cancer, and melanoma. Serum immunoglobulin levels were not available for the patients in our study.
Serum anti-Dsg antibody levels were determined in all pemphigus patients for whom serum samples were available before and after rituximab therapy (n = 10). Because Dsg enzyme-linked immunosorbent assay is not the standard of care for diagnosis or management of pemphigus, paired serum samples were not available for all pemphigus patients. Serial dilutions of patient serum (1:100 up to 1:12 800) were tested by Dsg enzyme-linked immunosorbent assay (MBL International) and an index value was calculated as previously described.11 Serum antibodies against pneumococcal capsular polysaccharide were measured in the same serum samples by enzyme-linked immunosorbent assay (Binding Site).
In 10 paired serum samples from pemphigus patients before and after rituximab administration, a median change of −80% in serum Dsg index value was observed after rituximab therapy. In the same group of patients, all of whom had received multiple courses of rituximab (range, 2-6 courses), serum pneumococcal antibody concentration was stable with a median change of +8% (Figure 3). There was no correlation between the levels of serum anti-Dsg and antipneumococcal antibodies (Spearman rank correlation coefficient r = −0.2). One patient sample demonstrated an 81% decrease in Dsg index value but a 607% increase in pneumococcal antibody concentration. This patient had developed fever, pharyngitis, and sinusitis responding to oral antibiotic therapy. Blood culture results were negative; throat and respiratory cultures were not performed.
Figure 3. Changes from before to after rituximab therapy in serum levels of desmoglein and pneumococcal capsular polysaccharide antibodies were unrelated (n = 10; Spearman rank correlation coefficient r = −0.2). A median change of −80% in desmoglein index value vs +8% in pneumococcal antibody concentration was observed.
A recent meta-analysis9 encompassing 153 pemphigus patients reported a complete remission rate of 65% after rituximab therapy. Most prior studies used complete remission of disease as the end point for rituximab efficacy, regardless of the dose of systemic therapy. However, complete remission on no or minimal therapy has been identified as the consensus end point for current pemphigus studies.10 Although it is impossible to retrospectively analyze the efficacy of most past studies under the consensus end point, complete remission off oral systemic therapy was observed in 23% of patients in the meta-analysis, 38% of 21 patients treated with a single cycle of rituximab, and 100% of 11 patients treated with rituximab and intravenous immunoglobulin.2,3,9
Our experience supports the early use of rituximab for pemphigus therapy on the basis of both efficacy and safety. Overall, 18 patients (58%) achieved the study end point, with 13 of these patients (42% of total) achieving complete remission off therapy. Seven patients (23%) achieved complete remission off therapy after a single cycle of rituximab. Achievement of the study end point was associated with early rituximab therapy of disease (Figure 1). There are limitations to our ability to substantiate this latter observation, including the retrospective study design, small number of patients, and potential for confounding by indication (ie, that patients treated later in the disease course have more refractory disease). However, published prospective studies12- 14 of pemphigus patients treated with first-line corticosteroid-sparing agents, such as mycophenolate mofetil or azathioprine, did not identify any cases of complete remission off therapy, raising the question for future prospective studies as to whether rituximab should be a first-line corticosteroid-sparing agent in pemphigus. An increased potential for better outcome with early use of rituximab is logical on the basis of the mechanism of CD20-targeted B cell depletion. Rituximab therapy, if 100% efficient, would deplete most of the B cell lineage, from pro-B cells in the bone marrow that have not yet formed the antibody repertoire to antigen-experienced memory B cells in the periphery. Repopulating B cells would produce new heavy and light chain combinations that ideally would not reproduce the pathogenic autoreactive clones. Antibody-secreting cells (plasma cells), which can survive for weeks to decades, are CD20- and hence are not directly targeted by rituximab. Data from our study and others2,15,16 showing a selective decrease in pemphigus autoantibodies after rituximab therapy suggest that pemphigus autoantibodies are preferentially produced by short-lived plasma cells, which are more dependent on the CD20+ memory B cell pool for replenishment. In contrast, protective antibodies against pneumococcus, tetanus, varicella zoster virus, or Epstein-Barr virus are unaffected in the serum, likely because of secretion from long-lived plasma cells. Factors determining the production of short-lived vs long-lived plasma cells are uncertain, although continued autoantigen stimulation in the setting of more long-standing disease could encourage the production and/or retention of long-lived autoreactive plasma cells that would be resistant to rituximab therapy. Consequently, early intervention with rituximab could be more likely to result in modification of the disease process, with possible long-term remission of disease.
Concern about rituximab safety stems from multiple studies. Two of 21 patients (10%) in the prospective case series of rituximab in pemphigus experienced severe infections, including fatal septicemia.2 A meta-analysis9 of 153 pemphigus patients treated with rituximab showed that 7% developed serious infections, with 2 fatalities (1.3%). In our study, 2 patients (6%) experienced SAEs attributed to rituximab, both nonfatal infections. However, serious infection, as well as other life-threatening SAEs, occurred in pemphigus patients before rituximab therapy. Thus, although risk of infection is an important consideration with rituximab, pemphigus patients, because of their underlying disease as well as its treatment, are susceptible to infection and other SAEs regardless of rituximab therapy.
In conclusion, our experience shows that rituximab is an effective therapy for pemphigus, potentially more so when given early in the course of disease. In addition, rituximab appears to be as safe as conventional immunosuppressive therapies. Current standard of care typically indicates rituximab for late therapy of pemphigus, after conventional treatments have failed.1,17 Our study suggests better clinical outcome with early rituximab treatment of pemphigus. Future prospective clinical trials would be valuable to determine the relative safety and efficacy of rituximab as a first-line corticosteroid-sparing agent in pemphigus compared with mycophenolate mofetil or azathioprine.
Correspondence: Aimee S. Payne, MD, PhD, Department of Dermatology, University of Pennsylvania, 217A Clinical Research Bldg, 415 Curie Blvd, Philadelphia, PA 19104 (firstname.lastname@example.org).
Accepted for Publication: April 14, 2012.
Published Online: June 18, 2012. doi:10.1001/archdermatol.2012.1522
Author Contributions: Drs Lunardon, Stanley, Tsai, and Payne had full access to all the data and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Lunardon, Stanley, and Payne. Acquisition of data: K. J. Tsai, Fett, Stanley, Werth, D. E. Tsai, and Payne. Analysis and interpretation of data: Lunardon, Propert, Stanley, Werth, and Payne. Drafting of the manuscript: Lunardon, Propert, and Payne. Critical revision of the manuscript for important intellectual content: K. J. Tsai, Propert, Fett, Stanley, Werth, D. E. Tsai, and Payne. Statistical analysis: Propert. Obtained funding: Payne. Administrative, technical, and material support: Lunardon, K. J. Tsai, D. E. Tsai, and Payne. Study supervision: Payne.
Financial Disclosure: Dr Tsai reports that he has partial stock ownership of Biogen IDEC. Dr Werth reports that she serves as a consultant for Lupus Foundation of America, Pfizer, Medimmune, Genentech, Novartis, Celgene, Stiefel, Rigel, Astion, Amgen, Infinity Pharmaceuticals, sanofi-aventis, and UBC; has research grants from Amgen and Celgene; and has partial stock ownership of UV Therapeutics; and that University of Pennsylvania holds the copyright for the Cutaneous Lupus Erythematosus Disease Area and Severity Index and the Cutaneous Disease and Activity Severity Index.
Funding/Support: This work was supported by the Department of Dermatology, University of Milan, Milan, Italy (Dr Lunardon), and grants AR052672 (Dr Stanley), AR002207 (Dr Werth), AR053505 and AR057001 (Dr Payne), CTSA UL1-RR-024134 (Dr Propert), and Skin Disease Research Center grant AR057217 (Drs Stanley and Payne) from the National Institutes of Health, Bethesda, Maryland.
Lunardon L, Tsai KJ, Propert KJ, Fett N, Stanley JR, Werth VP, Tsai DE, Payne AS. Adjuvant Rituximab Therapy of PemphigusA Single-Center Experience With 31 Patients. Arch Dermatol. 2012;148(9):1031-1036. doi:10.1001/archdermatol.2012.1522