Pharmacology and therapeutics

The impact of stratifying initial dose of corticosteroids by severity of pemphigus vulgaris on long-term disease severity
Anna Lyakhovitsky1, MD, Sharon Baum1, MD, Alon Scope1, MD, Boaz Amichai1, MD, A. Barzilai1, MD, Jacob Rimer2, and Henri Trau1, MD

1 Department of Dermatology, Sheba Medical Center, Tel-Hashomer, Israel, and 2Department of Immunology, Weizmann Institute of Science, Rehovot, Israel

Abstract
Background Pemphigus vulgaris (PV) is a life-threatening disease affecting skin and mucous membranes. The epitope spreading theory posits that uncontrolled PV can gradually worsen because of exposure of cellular antigens to the immune system. To this end, high-dose systemic corticosteroids have been advocated as rst-line treatment for patients with PV to achieve disease control. Objective To determine whether the initial dose of prednisone stratied by disease severity affects long-term disease severity. Methods A retrospective study was conducted on 58 patients with PV with at least ve years of follow-up from diagnosis. Patients were categorized into three groups according to the initial dose of prednisone treatment. Parameters analyzed included age, gender, disease severity at baseline and follow-up, hospitalizations, prednisone doses and adjuvant therapy at follow-up, and remission rate. Results Ten patients received initial low-dose prednisone or were treated initially without systemic CS, 19 patients received intermediate-dose prednisone, and 29 received highdose prednisone. Disease severity at presentation correlated directly with initial prednisone doses. The duration of the rst hospitalization and number of hospitalization days during the ve-year follow-up period were signicantly lower in the group treated with initial lowdose prednisone and similar for the groups treated with intermediate and high doses. Conclusions Disease severity of PV at presentation is a good predictor of the clinical course. Stratifying initial prednisone dose according to PV disease severity at presentation is appropriate.

Correspondence B. Amichai, MD Department of Dermatology Sheba Medical Center Tel-Hashomers Israel E-mail: boazam@clalit.org.il Conicts of interest: None.

Introduction Pemphigus vulgaris (PV) is a chronic and potentially lifethreatening autoimmune blistering disease affecting the skin and mucous membranes. Systemic corticosteroids (CS) are considered the rst-line treatment for PV, leading to a decline in mortality from almost 100% to an average of 6%. Mortality at present is mostly due to the adverse effects of high-dose and prolonged CS treatment and of various adjuvant immunosuppressive drugs used to control the disease.13 Eventually, PV subsides in most patients allowing discontinuation of systemic therapy. The remaining patients usually have only mild disease that is controlled with low-dose prednisone or with an adjuvant agent.1,2 Although many treatment strategies are available, few have been assessed in randomized trials.2,3 Since the introduction of CS treatment for PV, high doses of CS were advocated. Lever4,5 recommended an initial
International Journal of Dermatology 2011, 50, 10141019

1014

dose of 180 360 mg/d of prednisone to rapidly control the disease. Proponents of high-dose CS protocols argued that intermediate- and low-dose CS are not effective in inducing remission and would result paradoxically in higher cumulative CS doses with increased morbidity and mortality. A more recent argument in favor of rapid disease control using high-dose CS treatment has been the need to prevent the epitope spreading cascade. This theory posits that cellular antigens become exposed to the immune system during active PV, expanding the range of antibodies targeting autoimmune epitopes and leading to increased severity of disease. For example, mild PV that begins in oral mucosa, if inadequately controlled by high doses of CS, can spread to the skin and become severe PV.610 On the other hand, the severe side effects associated with high-dose CS treatment have led to protocols using moderate- and low-dose CS. Indeed, several trials showed that high doses of CS were not superior to
2011 The International Society of Dermatology

Lyakhovitsky et al.

Pemphigus treatment with oral prednisolone

Pharmacology and therapeutics

1015

moderate and low doses. Moreover, topical or intralesional CS treatment have been described for cases with mild PV.1,2,1012 The purpose of this study was to determine whether the initial CS dose stratied by disease severity affects the course of the disease and rate of remissions. More specically, we aimed to analyze whether initial low-dose CS therapy for mild disease had adverse effects on disease severity over time. Methods
Patients The computerized database at the Sheba Medical Center, a tertiary medical center in Israel, was searched using ICD-9 codes for all patients with PV who were admitted to the Department of Dermatology or visited the departments outpatient clinic between January 1985 and March 2005. The study was approved by the institutional review board. Inclusion criteria for the retrospective analysis included the following: 1 diagnosis of PV was conrmed by histopathology and by direct immunouorescence; 2 treatment of PV was initiated at our department; and 3 a follow-up of at least ve years from initial presentation was available. The approach in our department is to divide the patients using the Pemphigus Clinical Activity Score (PCAS) shown in Table 1, into the following subgroups: mild, moderate, and severe disease. The PCAS incorporates the extent of body surface area (BSA, %) involved by blisters or exudative lesions, number of cutaneous sites involved, the extent of mucosal involvement (number of mucosal sites involved), and severity of discomfort during drinking and eating. Mild disease included patients with a score of 14, moderate disease with a score of 58, and severe disease with a score of 912. Disease is also considered severe in cases that involved BSA more than 30% or in cases having severe discomfort in eating or drinking (Table 1).

Study protocol The patients were divided into three groups according to the initial dose of prednisone treatment: low (<0.5 mg/kg per day), intermediate (0.51 mg/kg per day), and high (more than 1 mg/kg per day). The following clinical and epidemiological parameters were analyzed and compared between the groups: disease severity according to PCAS; age at disease onset; gender; clinical form (oral, cutaneous, mucocutaneous); duration of the rst hospitalization; number of hospital admissions; total number of hospitalization days during a ve-year follow-up period from the diagnosis; adjuvant treatment used (number and type of agents); response to therapy after two and ve years of follow-up. Response to therapy was dened as follows: 1 complete remission off therapy absence of active lesions while the patient is off systemic therapy for at least two months; 2 complete remission on minimal therapy absence of active lesions while the patient is receiving minimal therapy (prednisone 10 mg/d and/or adjuvant agent) for at least two months; 3 partial remission off therapy transient new lesions healing within one week without treatment, while the patient is off systemic therapy for at least two months; and 4 partial remission on minimal therapy transient new lesions healing within one week without treatment, while patient is on minimal therapy (prednisone 10 mg/d and/or adjuvant agent) for at least two months.13

Table 1 Pemphigus clinical activity score (PCAS)

Extent of cutaneous involvement, sites 0 1 2 >2 (0) (1) (2) (3) Extent of mucosal involvement, sites 0 1 2 >2 (0) (1) (2) (3)

Statistical analysis The Statistical Package for the Social Sciences (SPSS) version 15.0 for Windows software was used for the data entry and analysis (SPSS Inc, an IBM company, Chicago, IL, USA). Normally distributed numerical data were summarized by their mean values and standard deviation. All analyses of paired data were performed using paired t-test. Condence intervals were extended to a level of 95%. P values below 0.05 were accepted as statistically signicant.

BSA, % <1% (0) 110% (1) 1130% (2) >30% (3*)

Discomfort in drinking and eating Absent (0) Slight (1) Signicant (2) Severe (3*)

Results Of 136 records patients with PV identied in the computerized database of Sheba Medical Center, 58 fullled the inclusion criteria for the study. Of the 58 patients included, 10 were treated initially without systemic CS (six) or with low-dose prednisone (four). All 10 patients had mild disease according to PCAS. In 19 patients, initial treatment was with intermediate-dose prednisone; 17 of these 19 had moderate PV and two had severe PV. In 29 patients, initial

Score range 012: mild disease, 14; moderate disease, 58; severe disease, 912 or less if it includes score marked by *.
2011 The International Society of Dermatology

International Journal of Dermatology 2011, 50, 10141019

1016

Pharmacology and therapeutics

Pemphigus treatment with oral prednisolone

Lyakhovitsky et al.

treatment was with high-dose prednisone; 27 of these 29 had severe PV and two had moderate PV.
Age of onset

The mean age at onset of the disease was 51 years (range 2572 years) (Table 2). There were no signicant age differences between the groups. The age of disease onset was similar between female and male patients. The age of onset was also similar among the subgroups treated with low-, intermediate-, and high-dose prednisone.
Female to male ratio

11.1 days (range 036 days), while those treated with intermediate- or high-dose prednisone were hospitalized for 27.4 days (range 089 days). Only the difference between low- and high-dose was statistically signicant (P < 0.02) (Table 3).
Parameters at disease follow-up

There were 38 (65.5%) female patients and 20 (34.5%) male patients. While female predominance was noted across the entire study population (F/M = 1.9 : 1), this gender inequality was particularly notable in the subgroup of patients treated with low-dose prednisone (F/ M = 9:1) (Table 2).
Clinical forms

Number of hospital admissions and hospitalization days during ve years from diagnosis The mean number of hospital admissions and total number of hospitalization days during the ve-year follow-up period correlated directly with the initial prednisone dose and with initial disease severity. The mean number of hospital admissions for all patients was 3.1 (range 013). Patients in the subgroup treated with low-dose prednisone were hospitalized on average 1.5 times, while patients in both the intermediate- and high-dose prednisone subgroups were hospitalized on average 3.7 and 3.2 times accordingly (Table 3).
Treatment (corticosteroid and adjuvant)

Fifty-three patients (91.4%) manifested oral lesions, and 43 of them (74.1%) had cutaneous involvement. Disease was limited to mucosa in 10 (17.24%) and to skin in ve patients (8.6%) (Table 2).
Duration of the rst hospitalization

The mean duration of the rst hospitalization was 24.4 days (range 089 days). A correlation was found between the initial dose of prednisone and duration of rst hospitalization. Patients treated initially with lowdose prednisone were hospitalized on average for
Table 2 Patient data: age, gender, and clinical form of

Of six patients who did not receive systemic CS at presentation, two did not require systemic CS during the veyear follow-up. Adjuvant treatments were used in all patients. The mean number of agents used was 2.2 and was not signicantly different among the low-, intermediate-, and highdose CS subgroups. However, dapsone was used more frequently in the subgroup of patients treated without or with low-dose prednisone (70%), while intravenous immunoglobulin was more frequently used in the subgroup treated with high-dose CS (28%) (Fig. 1). Response to treatment (remission rate) after 2 and after 5 years The disease activity decreased with time in most patients. The most signicant decrease in PCAS was found in the group of patients treated initially without systemic CS or with low-dose prednisone. After two years two (3.54%) patients were in complete remission and off therapy, nine (15.5%) were in complete
Table 3 Hospitalization data
First hospitalization, days (95% CI) 24.36 11.1 26.63 27.45 5.34 8.49 10.66 7.08 5 years hospitalization days (95% CI) 36.41 18.6 39.84 40.31 7.54 13.06 13.87 10.64 N hospitalizations (95% CI) 3.07 1.5 3.68 3.21 0.82 1.1 1.66 1.14

pemphigus vulgaris
Patients All (n = 58) Age, years (95% CI) 51.07 3.33 Skin/oral (%) Oral 10 (17.2) Skin 5 (8.6) Both 43 (74.1) Oral 2 (20.0) Skin 3 (30.0) Both 5 (50.0) Oral 4 (21.0) Skin 0 (0.0) Both 15 (79.0) Oral 4 (13.8) Skin 2 (6.9) Both 23 (79.3) F/M ratio 1.9

Low (n = 10)

53.9 8.55

Medium (n = 19)

50.47 5.08

1.38

High (n = 29)

50.48 5.06

1.64 Patients All (n = 58) Low (n = 10) Medium (n = 19) High (n = 29)

Three groups of patients according to the initial dose of prednisone treatment: all, all patients; low, low dose (<0.5 mg/kg per day); medium, intermediate dose (0.51 mg/kg per day); high, high dose (>1 mg/kg per day).
International Journal of Dermatology 2011, 50, 10141019

2011 The International Society of Dermatology

Lyakhovitsky et al.

Pemphigus treatment with oral prednisolone

Pharmacology and therapeutics

1017

Celcept

10% 16%

High IVIG
28% 16% 14% 10%

Medium

Low

Adjuvant agents

CyA

Cyt
10%

21% 26% 28%

MTX

37% 40% 76% 79% 80% 41% 42%

Figure 1 Adjuvant treatments. High,

Imuran

subgroup treated with high-dose prednisone; medium, subgroup treated with intermediate-dose prednisone; low, subgroup treated without or with low-dose prednisone. IVIG, intravenous immunoglobulin; CyA, cyclosporin A; Cyt, cyclophosphamide; MTX, methotrexate

Dapsone 0% 10% 20% 30% 40%

70%

50%

60%

70%

80%

90%

Patients
High subgroup treated with high-dose prednisone, Medium subgroup treated with intermediate-dose prednisone, Low subgroup treated without or with low-dose prednisone.

remission and on minimal therapy, none were in partial remission and off therapy, and 20 (34.5%) were in partial remission and on minimal therapy. After ve years, six (10.34%) of the patients were in complete remission and off therapy, eight (13.8%) were in complete remission and on minimal therapy, one (1.7%) was in partial remission and off therapy, and 21 (36.2%) were in partial remission and on minimal therapy (Table 4a,b). Discussion Our ndings support the rationale of stratifying the initial dose of systemic CS to severity of PV. In line with our department guidelines, we found a good correlation between disease severity at presentation and initial dose of systemic CS. During ve years of follow-up, we found that patients with mild disease severity at presentation that initially received low-dose CS were more likely to remain with mild disease and to require smaller doses of CS. The duration of rst hospitalization, number of admissions to the hospital, and total number of hospital days during the ve-year follow-up was signicantly lower in patients with mild disease and low initial CS dose than in patients with moderate or severe disease and with intermediate- or high-dose CS at presentation. It should be noted that of the six patients with PV that did not require systemic CS at initial presentation, two entered durable remissions without requiring systemic CS during ve years of follow-up. The difference in longterm disease severity was not attributed to difference in
2011 The International Society of Dermatology

the use of adjuvant agents between the subgroups. The number of adjuvant agents was similar in the subgroups. The more frequent use of dapsone in patients with milder PV, as opposed to the increased use of intravenous immunoglobulin in patients with severe PV, is a reection of our practice guidelines and is unlikely to alter long-term disease severity. Our results revealed a signicant difference between the subgroup of patients with mild disease and subgroups with moderate and severe disease, even though there were several similarities between the subgroups with moderate and severe disease (e.g. number of hospitalizations, number of hospital days during the ve-year follow-up). This suggests that there might be only two categories of cases: mild and moderatesevere. Although our impression is that there are moderate cases placed between mild and severe, the denitive answer to this question requires further studies. Our study population was comparable with previously published series with peak incidence in the sixth decade of life, female predominance, and distribution of lesions.1,2 In stratifying disease severity, we made an interesting observation of greater female preponderance among patients with mild disease. Several recent studies reported that remissions were much more common than previously reported.1416 For example, Bystryn and Steinman15 performed the retrospective study on 40 patients with PV and found that 25% of patients were in complete remission after two years, 50% after ve years, and 75% after 10 years. The rate of complete remission in our study was lower than in recent studies. We found that only 3.5% of patients were
International Journal of Dermatology 2011, 50, 10141019

1018

Pharmacology and therapeutics

Pemphigus treatment with oral prednisolone

Lyakhovitsky et al.

Table 4

(a) Prednisone dosage, mg/day initial, after 2 and 5 years and (b) Pemphigus: remission rates
Initial (95% CI) 2 years (95% CI) 5 years (95% CI)

Patients

(a) All (n = 58) 69.22 9.07 Low (n = 10) 11.5 9.25 Medium (n = 19) 57.36 2.94 High (n = 29) 96.89 6.82

17.46 9.25 15.21 21.77

4.32 3.73 5.7 7.39

11.13 6.25 11.84 12.34

2.5 3.81 4.51 3.74

Patients (b) All (n = 58)

Remission

2 years, n (%)

5 years, n (%)

CRNT CRMT PRNT PRMT CRNT CRMT PRNT PRMT CRNT CRMT PRNT PRMT CRNT CRMT PRNT PRMT

2 9 0 20 0 2 0 7 2 3 0 4 0 4 0 9

(3.4) (15.5) (0.0) (34.5) (0.0) (20.0) (0.0) (70.0) (10.5) (15.8) (0.0) (21.1) (0.0) (13.8) (0.0) (31.0)

6 8 1 21 0 2 1 6 3 2 0 6 3 4 0 9

(10.3) (13.8) (1.7) (36.2) (0.0) (20.0) (10.0) (60.0) (15.8) (10.5) (0.0) (31.6) (10.3) (13.8) (0.0) (31.0)

Low (n = 10)

Medium (n = 19)

High (n = 29)

CRNT, complete remission off (no) therapy; CRMT, complete remission on minimal therapy; PRNT, partial remission off (no) therapy; PRMT, partial remission on minimal therapy.

in complete remission without treatment and 15.5% in complete remission with minimal treatment after two years. After ve years, 10.3% of patients were in complete remission without treatment and 13.8% in complete remission with minimal treatment. We suggest that such differences may be related to the differences in the denition of remission. In our study, we adhered to the Consensus denitions,13 except for the denition of minimal treatment. While the Consensus regards the minimal treatment as prednisone 10 mg/d and/or an adjuvant agent reduced to half of the initial dose, we continued the full dose of an adjuvant agent for a long time, thus we dened the minimal treatment as prednisone 10 mg/d and/or an adjuvant agent without relation to its dose. Another possible explanation for the relatively low response rate in our study was that our subgroups of severe patients included 50% of cases, while in Bystryns group they included only 25%. This may be attributed to strict inclusion criteria, as less severe cases may not be followed up for ve years. In addition,
International Journal of Dermatology 2011, 50, 10141019

one can assume that our treatment protocol is less aggressive, giving a lower rate of remissions. The advantages of our study included using a score system, uniform treatment strategy, and relatively long-term follow-up. In our opinion, our score system was practical and useful for the retrospective study and included only clinical characteristics. From several scores reported, some were not appropriate for the retrospective study, and several used the dose of prednisone as a part of the initial severity score.1719 We believe that the score usually is created to decide about the dose of the treatment agent and should not be part of the initial severity score. This may be different for the score in follow-up. The present study has several limitations because of the retrospective design. The study group was relatively small, mostly because of strict inclusion criteria. There are also some difculties in comparing our results with those of previous studies, due to the differences in the scoring system and in the denition of remission. The limitations in assessing and comparing results from clinical trials of PV include: the rarity of the disease (leading to small number of participants), diversity in disease severity at onset, and the great variation in response to treatment. An additional difculty is that in studies performed there is no agreed common denition for grading disease severity, activity, therapeutic response of PV, and remission. Recently the Consensus was published in order to achieve the common denitions, but studies that were conducted until now still have not adhered to its denitions. In support of our ndings, several trials showed that protocols consisting of very high doses of CS are not superior to those consisting of moderate and low doses. Moreover, it was reported that mild cases of PV could be treated topically with CS solutions, ointments, or intralesional injections of CS. In conclusion, our retrospective review does not support the common practice of using an initial high dose of CS. The initial treatment is identied by the extent and rate of progression of the new lesions. Stratifying the CS dose by disease severity is of great importance because it gives an effective treatment and with minimum side effects from the CS, which are a main cause of mortality from PV today. There is an urgent need for large multicenter randomized controlled trials of treatment in patients with pemphigus in order to assess the optimal CS dose and the role of adjuvant immunosuppressive medications. References
1 Bystryn J, Rudolph JL. Pemphigus. Lancet 2005; 366: 6173.
2011 The International Society of Dermatology

Lyakhovitsky et al.

Pemphigus treatment with oral prednisolone

Pharmacology and therapeutics

1019

2 Jessop S, Khumalo NP. Pemphigus: a treatment update. Am J Clin Dermatol 2008; 9: 147154. 3 Martin LK, Werth V, Villanueva E, Segall J, Murell DF. Interventions for pemphigus vulgaris and pemphigus foliaceus. Cochrane Database Syst Rev 2009; 20: CD006263. doi:10.1002/14651858.CD006263.pub2. 4 Lever WF, Schaumburg-Lever G. Immunosuppressants and prednisone in pemphigus vulgaris: therapeutic results obtained in 63 patients between 1961 and 1975. Arch Dermatol 1977; 113: 12361241. 5 Lever WF, Schaumburg-Lever G. Treatment of pemphigus vulgaris. Results obtained in 84 patients between 1961 and 1982. Arch Dermatol 1984; 120: 4447. 6 Pigozzi B, Peserico A, Schiesari L, Alaibac M. Pemphigus foliaceus evolving into pemphigus vulgaris: a probable example of intermolecular epitope spreading conrmed by enzyme-linked immunosorbent assay study. J Eur Acad Dermatol Venereol 2008; 22: 242244. 7 Endo H, Rees TD, Hallman WW, Kuyama K, Nakadi M, Kato T, Kono Y, Yamamoto H. Disease progression from mucosal to mucocutaneous involvement in a patient with desquamative gingivitis associated with pemphigus vulgaris. J Periodontol 2008; 79: 369375. 8 Salato VK, Hacker-Foegen MK, Lazarova Z, Fairbey JA, Lin MS. Role of intramolecular epitope spreading in pemphigus vulgaris. Clin Immunol 2005; 116: 5464. 9 Tchernev G, Orfanos CE. Antigen mimicry, epitope spreading and the pathogenesis of pemphigus. Tissue Antigens 2006; 68: 280286. 10 Ratnam KV, Phay KL, Tan CK. Pemphigus therapy with oral prednisolone regimens. A 5-year study. Int J Dermatol 1990; 29: 363367.

11 Hietanen J, Salo OP. Pemphigus: an epidemiological study of patients treated in Finnish hospitals between 1969 and 1978. Acta Derm Venereol 1982; 62: 491496. 12 Smolle J. [Therapy of pemphigus. Critical remarks based on 44 clinical cases]. Hautarzt 1985; 36: 96102. 13 Murrell DF, Dick S, Ahmed AR, Amagai M, Barnadas MA, Borradori L, Bystrin JC, et al. Consensus statement on denitions of disease, end points, and therapeutic response for pemphigus. J Am Acad Dermatol 2008; 58: 10431046. 14 Herbst A, Bystryn JC. Patterns of remission in pemphigus vulgaris. J Am Acad Dermatol 2000; 42: 422427. 15 Bystryn J, Steinman NM. The adjuvant therapy of pemphigus. An update. Arch Dermatol 1996; 132: 203212. 16 Czernik A, Bystryn J. Kinetics of response to conventional treatment in patients with pemphigus vulgaris. Arch Dermatol 2008; 144: 682683. 17 Pftze M, Niedermeier A, Hertl M, Eming R. Introducing a novel Autoimmune Bullous Skin Disorder Intensity Score (ABSIS) in pemphigus. Eur J Dermatol 2007; 17: 411. 18 Rosenbach M, Murell DF, Bystrin JC, Dulay S, Dick S, Fakharzadeh S, et al. Reliability and convergent validity of two outcome instruments for pemphigus. J Invest Dermatol 2009; 129: 24042410. 19 Agarwal M, Walia R, Kochhar AM, Chander R. Pemphigus Area and Activity Score (PAAS)a novel clinical scoring method for monitoring of pemphigus vulgaris patients. Int J Dermatol 1998; 37: 158160.

2011 The International Society of Dermatology

International Journal of Dermatology 2011, 50, 10141019