Prognostic factors in pemphigus vulgaris and pemphigus foliaceus M. Saha, 1 B. Bhogal, 1 M.M. Black, 1 D. Cooper, 2 R.W. Vaughan 3 and R.W.

Transcription

1 CLINICAL AND LABORATORY INVESTIGATIONS BJD British Journal of Dermatology Prognostic factors in pemphigus vulgaris and pemphigus foliaceus M. Saha, 1 B. Bhogal, 1 M.M. Black, 1 D. Cooper, 2 R.W. Vaughan 3 and R.W. Groves 1 1 Department of Immunodermatology, St John s Institute of Dermatology, London SE1 9RT, U.K. 2 Department of Statistics, GSTS and MRC Centre for Transplantation, Guy s Hospital, London SE1 9RT, U.K. 3 Department of Clinical Transplantation Laboratory, GSTS and MRC Centre for Transplantation, Guy s Hospital, London SE1 9RT, U.K. Summary Correspondence Monika Saha. monikasaha@doctors.org.uk Accepted for publication 31 August 2013 Funding sources This work was generously supported by the British Skin Foundation, The Immunodermatology Fund of the Guy s and St Thomas Charity and the National Institute for Health Research Comprehensive Biomedical Research Centre at Guy s and St Thomas NHS Foundation Trust/King s College London. Conflicts of interest None. DOI /bjd Background Pemphigus typically has a chronic course, although there is great variability in disease duration (DD) and time taken to disease remission (DR) between individuals with the disease. The reasons for this are unclear. Objectives To explore the prognostic influence of epidemiological, clinical, immunological and genetic factors on disease course and remission in pemphigus vulgaris (PV) and pemphigus foliaceus (PF). Methods This was a retrospective study of patients with PV and PF, recruited from a single UK centre. Direct and indirect immunofluorescence and enzyme-linked immunosorbent assay studies for antidesmoglein (Dsg) antibodies were used to assess immunological factors. Polymerase chain reaction with sequence specific primers (PCR-SSP) was used to assess the Class II human leukocyte antigen status of patients. Prognostic endpoints investigated were time to initial first DR and total DD. Results Ninety-five patients were recruited (79 PV and 16 PF). Patients of Indo- Asian origin were significantly associated with longer DD than White-British patients (P = 0029). In addition, younger age at onset was associated with a worse prognosis in terms of DD: the mean age at presentation of patients with DD of < 5 years was 49 years (SEM = 34) compared with 40 years (SEM = 19) in those with DD > 5 years (P = 0039). A higher initial intercellular antibody titre on normal human skin substrate was associated with a greater time to initial DR (P = 0007) and high anti-dsg 3 levels at baseline were associated with a longer total DD (P = 003). Conclusions Ethnic group, age at presentation, initial intercellular antibody titre and initial Dsg 3 antibody levels all had a significant impact on prognosis of pemphigus. What s already known about this topic? Pemphigus typically has a chronic course, although there is great variability in disease duration and time taken to disease remission between individuals with the disease. The reasons for this are unclear. Certain genetic, epidemiological, clinical and immunological factors may be implicated in prognosis of pemphigus. What does this study add? This is the first detailed exploration of prognostic factors in both PV and PF in a U.K. population. We have identified that ethnic group, age at presentation, initial intercellular antibody titre and initial desmoglein 3 antibody levels all had a significant impact on prognosis of pemphigus. 116 British Journal of Dermatology (2014) 170, pp

2 Prognostic factors in Pemphigus, M. Saha et al. 117 Pemphigus is a rare autoimmune blistering disease that is characterized by pathogenic auto-antibodies to key desmosomal proteins central to epidermal integrity. Pemphigus typically has a chronic course, with an average disease duration (DD) of 10 years. 1 However, there is great variability in disease length, some patients having active disease for a few months to a few years whereas others may have a protracted course extending over several decades. Previous work (Table 1) has identified a number of factors that may be implicated in prognosis of pemphigus Ethnic group, 3 gender and age at onset 2 are the main epidemiological variables that have been suggested to play a role in determining disease outcome. Some studies have suggested that a history of smoking may have a protective role in pemphigus. 4 6 Although pemphigus foliaceus (PF) has been regarded as a milder disease with a more favourable prognosis than pemphigus vulgaris (PV), one study has found little difference in prognosis. 7 A number of authors have suggested that the timing of initial treatment in pemphigus may correlate with prognosis, although opinions have been divided. Savin et al. 2 suggested that early treatment of patients was associated with a greater risk of mortality. Subsequently, others have found the contrary, that is, early response to therapy was important in predicting early induction of remission. 1 The site and severity of the initial disease have both been suggested to be important in determining prognosis: a good prognosis has been associated with those who had skin disease only at presentation, 8 10 and patients with mild to moderate disease severity at diagnosis were twice as likely to reach long-lasting disease remission (DR) compared with those with severe disease. 1 Pemphigus is an autoimmune disease, and there is little doubt that immunological factors are integral to the pathogenesis of the disease. Pemphigus antibodies tend to correlate with disease activity, 11 but their role in predicting the course of disease has not been established. Presence of both antidesmoglein (Dsg) 1 and 3 antibodies in PV tends to be associated with more aggressive disease, 12 but the impact of the Dsg profile has not been investigated in terms of DD and time to initial DR. Certain major histocompatability complex (MHC) alleles are associated with pemphigus, with human leukocyte antigen (HLA) DRB1*04 and *14 alleles being strongly associated with the disease in various ethnic groups. 13 However, there is currently little data on the effect of MHC alleles on the clinical course of pemphigus. The aim of this study was to determine whether epidemiological and clinical factors, immunological parameters or genotype influence the clinical course of pemphigus in a well-characterized group of patients recruited from a single U.K. centre. Materials and methods Patients Patients were recruited retrospectively from paper and electronic databases ( ) at the Immunodermatology Laboratory at St John s Institute of Dermatology, London, U.K. Detailed clinical, epidemiological and immunological data were collected from medical records. Epidemiological factors investigated were: age of onset, gender, ethnic background and smoking history. Clinical factors explored were: subtype of disease, initial site of disease (skin, mucosal or both), initial prednisolone dose, time to DR, total DD and follow-up period. Immunological factors investigated were: direct immunofluorescence (DIF), including the presence of complement 3 binding, baseline and 1 year indirect immunofluorescence (IIF) titres of pemphigus antibody, and anti-dsg 1 and anti-dsg 3 levels. Patients with either PV or PF were eligible for entry into the study according to the following: (i) characteristic clinical features of pemphigus, (ii) a skin or mucosal biopsy showing acantholysis and/or (iii) baseline immunopathology DIF with positive intercellular binding with IgG +/ complement 3 and positive IIF with circulating intercellular IgG antibodies, (iv) stored sera available for retesting for immunofluorescence studies, (v) clinical Table 1 Previous epidemiological studies on pemphigus and prognosis Date Population size (n) Location Primary endpoints References U.S.A. Mortality Rosenberg et al Scotland Mortality Savin Israel Mortality Seidenbaum et al. 8 Duration of clinical remission No. of exacerbations U.S.A Clinical remission Herbst and Bystryn Tunisia Relapse Zaraa et al. 7 Clinical remission Mortality Iran Clinical remission Kavusi et al Israel Clinical outcome Mimouni et al Tunisia Remission period Khaled et al. 20 Mortality Korea Clinical remission Kim et al. 21 British Journal of Dermatology (2014) 170, pp

3 118 Prognostic factors in Pemphigus, M. Saha et al. history available and (vi) a minimum of 4 years follow-up data available. In a subset of cases (n = 57), pretreatment serum was not available due to diagnostic testing occurring in other centres and countries. In these cases, the first serum stored in the immunodermatology laboratory was used as a baseline, which normally coincided with the patient s first visit to St John s Institute of Dermatology. The only exclusion criteria were: (i) subtypes of pemphigus other than vulgaris or foliaceus and (ii) if sera or clinical information were not available. Ethical approval for this study was obtained from the Royal Marsden Research Ethics Committee in September Endpoints Two independent endpoints were chosen: time to first DR and the DD. DR was the period (months) from diagnosis to first complete clearing of clinical lesions (which then remained clear for at least 1 year) on minimal immunosuppression (< 10 mg prednisolone +/ a steroid sparing immunosuppressant). DD was the period from diagnosis (years) to complete remission of clinical disease and withdrawal of all immunosuppression. DD was assigned only after observation for 1 year in order to detect relapse. For Kaplan Meier analysis, DD was assigned for all patients, including those with persisting active disease or ongoing immunosuppression, as the period from diagnosis to most recent follow-up. In addition, a subgroup of patients who had clinical disease for less than 5 years (defined as total DD of < 5years) was analysed using logistic regression. For this analysis, patients who had not been followed up for a minimum of 5 years were excluded. Serum collection and antibody testing Serum samples for each patient were identified from an archival store and IIF was performed on baseline and 1-year postbaseline samples. Two IIF substrates were used; normal human skin and marmoset oesophagus (Harlan Laboratories, Derby, U.K.). Commercially available enzyme-linked immunsorbent assay (ELISA) kits (MBL laboratories, Nagoya, Japan) were used to determine specific anti-dsg 1 and anti-dsg 3 levels at baseline and at 1 year. Threshold values for positive Dsg 1 and Dsg 3 antibody levels were taken as 30 U. 14 Change in intercellular antibody and Dsg 1 and 3 levels were assessed by recording the difference between the initial sample and that at 1 year, categorizing them into either negative (reduction in antibody level), positive (increase in antibody level) or no change in antibody. DIF data were collected on all patients for whom this was available at presentation. Polymerase chain reaction with sequence specific primers Medium to high resolution HLA Class II typing was performed using polymerase chain reaction with sequence specific primers (PCR-SSP). HLA DR status was classed in three groups: British Journal of Dermatology (2014) 170, pp *04, *14 or none (no *04 or *14 alleles), accounting for the majority of recognized pemphigus susceptibility alleles. Statistical analysis Survival analysis with the Cox regression model was used to assess the relationship of potential prognostic risk factors to our two independent endpoints, i.e. DD and time to first DR. Kaplan Meier analysis was used to generate standard survival curves for each variable. 15 SPSS version 16 (IBM, Armonk, NY, U.S.A.) was used for this analysis. The comparison of survival curves was based on the logrank test and a P value of < 005 was regarded as statistically significant. Results Ninety-five patients were identified, 79 with PV and 16 with PF. Only relevant survival plots are shown (negative plots have not been included). Disease duration and disease remission There was a wide variation in DD and time taken to achieve DR in the patients included in the study (Fig. 1a,b). There was a highly significant positive correlation between these two parameters (r = 055) (Fig. 1c). Epidemiological factors The predominant ethnic group in our cohort was Indo-Asian (n = 37) (Table 2). The second most common group was White British (n = 29). For purposes of analysis, ethnic origins were grouped into three clusters: (i) Indo-Asians, (ii) White Europeans, which included White British, other Europeans and Jewish patients and (iii) Other, which included Afro-Caribbean and remaining groups. We identified a significant difference in DD between patients from White European and Indo-Asian ethnic groups (P = 0029), with Indo-Asian background being significantly associated with longer DD (Fig. 2a) in comparison with White European, though there was no impact on DR (Fig. 2b). Analysis of ethnic status in a subgroup of patients with very short DD (< 5 years) further confirmed a significant difference between the ethnic groups (P = 0016). Only 3/35 (9%) Indo-Asians compared with 15/42 (36%) of White-Europeans, had DD < 5 years. Age at presentation varied between 16 and 78 years, with a mean age of 42 years. We did not find any overall association between age at presentation and DD or DR. However, in the DD < 5 years group, age at presentation was associated with slightly older patients. The mean age at presentation of patients with DD < 5 years was 49 years (SEM 34) compared with 40 years (SEM 19) in those who had a DD > than 5 years (P = 0039). In the whole group (n = 95), 55 were female and 40 male (ratio 14:1). Sex did not have any impact on DD or DR.

4 Prognostic factors in Pemphigus, M. Saha et al. 119 Disease duration (years) (a) (b) 0 0 (c) Smoking history was available for a subgroup of patients (74/95), and we found 61/74 were lifelong non-smokers, six were current smokers and seven were ex-smokers. Thus 18% of patients had either a past or current history of smoking compared with 82% who were non-smokers. There was no significant difference between the DD and DR endpoints in the smokers (current and ex) and non-smokers. Clinical factors r = 0 55 P = < Time taken to disease remission (months) Fig 1. (a) Variation in disease duration (DD) of all patients with pemphigus in the study. (b) Variation in time taken to disease remission (DR) [excluding patients who did not reach DR (n = 46)]. (c) Scattergraph of patients who reached DR and their DD. There were 79 patients with PV and 16 with PF. Clinical subtype did not have an effect on either DD (P = 0249) or DR (P = 046). For patients with PF, mean duration of disease was 111 years compared with 102 years in the PV group. The mean initial prednisolone dose in the studied population was 5025 mg, with a range of mg. Patients were divided into 59 mg and 60 mg groups. The first prednisolone treatment dose had no impact on DD (P = 095) or DR (P = 0881). Table 2 Racial distribution (regions and countries of origin) of patients with pemphigus in the study group Ethnic group No. of patients Indo-Asian (n = 37) Gujarat, India 5 Punjab, India 1 Goa, India 1 Unknown, India (born in Africa) 2 Gujarat, India (born in Africa) 9 Bangladesh 2 Pakistan 4 Sri Lanka 2 Tanzania 1 India (unknown) 10 White European (n = 46) U.K. 29 Jewish ancestry 8 Greece/Cyprus 3 Italy 1 Poland 1 Turkey 2 Unknown 2 Afro-Caribbean 8 Other China 1 Israel 1 Iran 1 Mixed race 1 The most common site of presentation in PV was the mucous membranes (40/76, 53%), followed by combined skin and oral presentation (21/76, 28%) and the least common presentation was skin only (15/76, 20%). We were unable to identify an effect of initial site of presentation on either DD and DR endpoints. Serological factors Pemphigus vulgaris Baseline intercellular antibody titres to normal human skin and marmoset oesophagus were highly variable. The titre of initial intercellular antibody on normal human skin was not related to DD (P = 050), although the baseline IIF was significantly related to time taken to DR (P = 0007) (Fig. 2c). Thus, a higher initial normal human skin IIF titre was associated with a longer time to DR. In contrast, baseline IIF titre on marmoset oesophagus was not significantly associated with DD (P = 050) or time taken to DR (P = 085). In addition, change in intercellular antibody titre (0 1 year) on both normal human skin (P = 013, P = 061) and marmoset oesophagus (P = 029, P = 062) had no effect on DD or DR, respectively. Dsg 1 and 3 data were available for 84 patients. Dsg antibody levels were divided into three groups for analysis: (i) 0 29 (negative), (low medium Dsg antibody levels) and > 100 (high Dsg antibody level). A significant association was found between baseline Dsg 3 levels and DD (P = 003) British Journal of Dermatology (2014) 170, pp

5 120 Prognostic factors in Pemphigus, M. Saha et al. (a) P = (c) P = Baseline IIF on normal human skin Ethnic group White European Indo-Asian Other (b) (d) Ethnic group White European Indo-Asian Other P = 0 03 Baseline Dsg >100 Fig 2. (a) Survival plots of major ethnic group vs. disease duration (DD). There was a significant difference in DD between White-Europeans and Indo-Asian patients (P = 0029). (b) Survival plots of major ethnic group vs. time taken to first disease remission (DR), showing no significant difference between ethnic groups. (c) Survival plots of baseline intercellular antibody titre on normal human skin substrate vs. time taken to DR. There is a significant association (P = 0007), indicating that a higher baseline indirect immunofluorescence (IIF) was associated with a longer time to DR. (d) Survival plots of baseline desmoglein 3 (Dsg 3) level and DD, demonstrating that a higher baseline Dsg 3 was associated with longer DD. Table 3 Human leukocyte antigen (HLA) class II DRB1 alleles in patients with pemphigus vulgaris and pemphigus foliaceus Number of patients Pemphigus vulgaris (n = 67) Pemphigus foliaceus (n = 12) HLA DRB1 allele group *04 *14 Two risk factor alleles (*04 or *14) None DR, and change in Dsg 3 antibody levels did not relate to either DD or DR. In addition, we could not identify a significant effect of the Dsg 1 and 3 profile with either DD or DR. Pemphigus foliaceus There were only 16 patients in this group and neither baseline IIF titre on normal human skin nor change in IIF titre (0 1 years) showed any association with DD or DR. In addition, there was no association between baseline Dsg 1 levels or change in Dsg 1 levels (0 1 years) with either DD or time taken to DR. (Fig. 2d). However, we could not identify an association between baseline Dsg 1 and DD (P = 058) or DR (P = 061) or between baseline Dsg 3 and DR (P = 021). Change in Dsg 1 antibody (0 1 year) level as a predictor of DD approached significance (P = 006) but did not relate to Direct immunofluorescence DIF data at presentation was available for 42 PV and all PF patients. All had positive intercellular staining with IgG and of those, 35 patients had positive staining with complement 3. Positive DIF with complement 3 had no effect on DD or DR. British Journal of Dermatology (2014) 170, pp

6 Prognostic factors in Pemphigus, M. Saha et al. 121 Genetic factors HLA typing data was available in 79/95 patients. Many previous studies have shown that Class II HLA DRB1 allele groups *04 and *14 are associated with pemphigus and 94% (63/ 67) of our PV patients had either *04 or *14 or both alleles (Table 3). Of patients with PF, 92% had at least one of these alleles. We found no significant difference in DD between patients who had either DRB1*04 or *14 compared with those who had neither of these alleles (P = 064). Seven of the 79 patients were homozygous for either DRB1*04 or *14 alleles. Homozygosity for either did not have a significant effect on DD, although this was approaching significance (P = 0071). Discussion DD and time taken to DR were highly variable in patients with pemphigus. Interestingly, there was a positive correlation between both DD and DR, suggesting that if it is difficult to achieve initial control (DR) patients are more likely to have a longer overall duration of disease. Certain epidemiological factors were shown to be important in predicting the above endpoints. The most striking significant result was the impact of ethnic group on DD in pemphigus. Indo-Asian patients were significantly associated with longer DD and were less likely to fall into the short (< 5 years) duration subgroup than White European patients. Ethnic group has been reported as a prognostic factor in pemphigus previously; for example, Jewish patients have been found to have a higher mortality rate, although the increased mortality in the Jewish group was attributable to underlying treatment of the disease. 3 However, this is the first time that an association between Indo-Asian ethnicity and poor prognosis, in terms of DD, has been demonstrated. The reasons underlying this are likely to be multi-factorial. Pemphigus in Indian patients has been associated with the presence of both Dsg 1 and 3 antibodies leading to a more aggressive clinical phenotype, which may also be associated with a longer DD. 12 A potential environmental factor maybe the high dietary content of garlic and tannins in Indo-Asians, both of which have been linked to disease development. 22,23 Overall, DD and DR were not significantly associated with age at onset. However, subgroup analysis of patients with DD of < 5 years revealed that age at onset did relate to DD, with older patients being more likely to be in the short duration group. Although Savin reported that an older age of onset was associated with a worse prognosis, this was in terms of mortality rather than DD and his study group was much smaller. 2 Previous epidemiological studies showed a decreased risk of pemphigus in ex or current smokers, compared with controls. 16 National statistical data from the UK 2007 household survey 17 indicate that the prevalence of smoking in the UK smokers was 21% and, although the incidence of smoking in our study (176%) was lower than this, we were unable to identify a relationship between smoking history and DD or DR in our patient group. The clinical subtype of pemphigus did not appear to impact on DD or DR, highlighting the fact that PF can be just as long and difficult to control as PV. Consistent with this, a previous smaller study in Israel also found no difference between PV and PF in terms of clinical course and prognosis. 7 Our data failed to demonstrate that an initial higher prednisolone dose was associated with decreased DD or DR. Other groups have reported no improvement in disease response with very high doses of prednisolone 8 suggesting that, due to the morbidity associated with high dose prednisolone, treatment with an initial moderate dose may be advantageous. The most common site of disease presentation in PV in our cohort was the mucous membranes, in keeping with previous reports. 18 In addition, the initial site of presentation of disease in PV patients did not affect prognosis. This is in contrast to previous data suggesting that combined mucocutaneous presentation is associated with a poorer prognosis A large Israeli study (PV = 88, PF = 3) found that a good prognosis (in terms of percentage reaching clinical remission and percentage of exacerbations) was associated with those who had skin disease only. 8 This may be related to selection bias, as the cohort of patients examined in the present study were all derived from a tertiary referral centre. It is also unusual in our experience for PV to present with skin involvement alone. Our study found that initial intercellular antibody titre to normal human skin was significantly related to time taken to DR in PV patients. Surprisingly, baseline Dsg 1 levels, which often correlate with IIF on normal human skin substrate, were not significantly associated with time taken to DR. This may relate to the restricted dynamic range of ELISA testing for Dsg 1, and further work is needed to explore this further. Additionally, higher baseline Dsg 3 antibody levels were a predictor of longer DD. Taken together, these findings indicate that immunological parameters early in the disease may provide useful long-term predictive data. Although presence of a Dsg 3 + /1 + profile in PV patients has been associated with aggressive disease, 12 in our study it was not associated with longer DD or time taken to DR. No immunological variables analysed in PF related to DD or time to DR, most likely reflecting the relatively small group analysed, making statistical analysis difficult. In conclusion, patients of Indo-Asian origin and with a younger age of onset had a worse prognosis. In addition, immunological parameters early in the disease course predicted prognosis of disease. We have demonstrated for the first time that initial intercellular antibody titre to normal human skin substrate may predict time taken to DR and baseline Dsg 3 level may predict total DD in PV patients. Our findings are of direct clinical relevance, as these patients can be identified, treated accordingly and monitored long-term. This is the first detailed exploration of prognostic factors in both PV and PF in a U.K. population. From a clinical British Journal of Dermatology (2014) 170, pp

7 122 Prognostic factors in Pemphigus, M. Saha et al. perspective, it is hoped that these data will aid dermatologists in managing and predicting outcome in patients with this serious and debilitating disease. References 1 Herbst A, Bystryn JC. Patterns of remission in pemphigus vulgaris. J Am Acad Dermatol 2000; 42: Savin JA. Some factors affecting prognosis in pemphigus vulgaris and pemphigoid. Br J Dermatol 1981; 104: Carson PJ, Hameed A, Ahmed AR. Influence of treatment on the clinical course of pemphigus vulgaris. J Am Acad Dermatol 1996; 34: Mehta JN, Martin AG. A case of pemphigus vulgaris improved by cigarette smoking. Arch Dermatol 2000; 136: Valikhani M, Kavusi S, Chams-Davatchi C et al. Pemphigus and associated environmental factors: a case-control study. Clin Exp Dermatol 2007; 32: Valikhani M, Kavusi S, Chams-Davatchi C et al. Impact of smoking on pemphigus. Int J Dermatol 2008; 47: Zaraa I, Mokni M, Hsairi M et al. Pemphigus vulgaris and pemphigus foliaceus: similar prognosis? Int J Dermatol 2007; 46: Seidenbaum M, David M, Sandbank M. The course and prognosis of pemphigus. A review of 115 patients. Int J Dermatol 1988; 27: Kavusi S, Daneshpazhooh M, Farahani F et al. Outcome of pemphigus vulgaris. J Eur Acad Dermatol Venereol 2008; 22: Chams-Davatchi C, Valikhani M, Daneshpazhooh M et al. Pemphigus: analysis of 1209 cases. Int J Dermatol 2005; 44: Creswell SN, Black MM, Bhogal B et al. Correlation of circulating intercellular antibody titres in pemphigus with disease activity. Clin Exp Dermatol 1981; 6: Harman KE, Gratian MJ, Bhogal BS et al. A study of desmoglein 1 autoantibodies in pemphigus vulgaris: racial differences in frequency and the association with a more severe phenotype. Br J Dermatol 2000; 143: Tron F, Gilbert D, Mouquet H et al. Genetic factors in pemphigus. J Autoimmun 2005; 24: Harman KE, Gratian MJ, Seed PT et al. Diagnosis of pemphigus by ELISA: a critical evaluation of two ELISAs for the detection of antibodies to the major pemphigus antigens, desmoglein 1 and 3. Clin Exp Dermatol 2000; 25: Bland JM, Altman DG. Survival probabilities (the Kaplan Meier method). BMJ 1998; 317: Brenner S, Tur E, Shapiro J et al. Pemphigus vulgaris: environmental factors. Occupational, behavioral, medical, and qualitative food frequency questionnaire. Int J Dermatol 2001; 40: Robinson S, Lader D. Smoking and drinking among adults General Household Survey 2007, UK National Statistics. 18 Regezi JA, Sciubba JA, Jordan RCK. Vesiculobullous diseases. In: Oral Pathology: Clinical Pathologic Correlations, 6th edn St Louis: Elsevier Saunders, 2012; Mimouni D, Bar H, Gdalevich M et al. Pemphigus, analysis of 155 patients. J Eur Acad Dermatol Venereol 2010; 24: Khaled A, Taazayet SB, Ben Alaya N et al. The course and prognosis of pemphigus in 47 Tunisian patients. J Eur Acad Dermatol Venereol 2011; 27: Kim MR, Kim HC, Kim SC. Long-term prognosis of pemphigus in Korea: retrospective analysis of 199 patients. Dermatology 2011; 223: Tur E, Brenner S. Contributing exogenous factors in pemphigus. Int J Dermatol 1997; 36: Ruocco V, Brenner S, Lombardi ML. A case of diet-related pemphigus. Dermatology 1996; 192: Rosenberg FR, Sanders S, Nelson CT. Pemphigus: a 20 year review of 107 patients treated with corticosteroids. Arch Dermatol 1976; 112: British Journal of Dermatology (2014) 170, pp