Advertisement

Large International Validation of ABSIS and PDAI Pemphigus Severity Scores

Open ArchivePublished:October 06, 2018DOI:https://doi.org/10.1016/j.jid.2018.04.042
      The Pemphigus Disease Area Index (PDAI) and Autoimmune Bullous Skin Disorder Intensity-Score (ABSIS) scores have been proposed to provide an objective measure of pemphigus activity. These scores have been evaluated only on already treated patients mainly with mild to moderate activity. The objective was to assess the interrater reliability of ABSIS and PDAI scores and their correlation with other severity markers in a large international study.
      Consecutive patients with newly diagnosed pemphigus were enrolled in 31 centers. Severity scores were recorded during a 24-month period by the same two blinded investigators. Serum was collected at each visit for ELISA measurement of anti-desmoglein antibodies. The intraclass correlation coefficient (ICC) and Spearman rank correlation coefficient were calculated.
      A total of 116 patients with pemphigus vulgaris (n = 84) or pemphigus foliaceus (n = 32) were included. At baseline, the ABSIS and PDAI ICCs were 0.90 (95% confidence interval [CI] = 0.85–0.93), and 0.91(95% CI = 0.87–0.94), respectively. The ICCs for PDAI were higher in moderate and extensive pemphigus (ICC = 0.82, 95% CI = 0.63–0.92 and ICC = 0.80, 95% CI = 0.62–0.90, respectively) than in patients with intermediate (significant) extent (ICC = 0.50, 95% CI = 0.27–0.68). Conversely, the ICCs for ABSIS were lower in patients with moderate extent (ICC = 0.44, 95% CI = 0.004–0.74) than in those with intermediate or extensive forms, (ICC = 0.69, 95% CI = 0.51–0.81 and ICC = 0.75, 95% CI = 0.51–0.88, respectively). During patients’ follow-up, the ICCs of both ABSIS and PDAI scores remained higher than 0.70.
      ABSIS and PDAI skin (r = 0.71 and r = 0.75) but not mucosal (r = 0.32 and r = 0.37) subscores were correlated with the evolution of anti-DSG1 and anti-DSG3 ELISA values, respectively. ABSIS and PDAI scores are robust tools to accurately assess pemphigus activity.

      Abbreviations:

      ABSIS (Autoimmune Bullous Skin Disorder Intensity Score), CI (confidence interval), DLQI (Dermatology Quality of Life Index), ICC (intraclass correlation coefficient), M (month), PDAI (Pemphigus Disease Area Index), PF (pemphigus foliaceus), PGA (Physician Global Assessment), PV (pemphigus vulgaris)

      Introduction

      The precise assessment of skin and especially of mucosal lesions is difficult in pemphigus, a rare and severe autoimmune blistering disease. Pemphigus is caused by the production of auto-antibodies directed against desmogleins, which are desmosomal proteins responsible for the adhesion of keratinocytes (
      • Amagai M.
      • Klaus-Kovtun V.
      • Stanley J.R.
      Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion.
      ,
      • Anhalt G.J.
      • Labib R.S.
      • Voorhees J.J.
      • Beals T.F.
      • Diaz L.A.
      Induction of pemphigus in neonatal mice by passive transfer of IgG from patients with the disease.
      ,
      • Chee S.-N.
      • Murrell D.F.
      Pemphigus and quality of life.
      ,
      • Langan S.M.
      • Smeeth L.
      • Hubbard R.
      • Fleming K.M.
      • Smith C.J.P.
      • West J.
      Bullous pemphigoid and pemphigus vulgaris—incidence and mortality in the UK: population based cohort study.
      ,
      • Sebaratnam D.F.
      • Frew J.W.
      • Davatchi F.
      • Murrell D.F.
      Quality-of-life measurement in blistering diseases.
      ,
      • Stanley J.R.
      • Amagai M.
      Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome.
      ). Pemphigus may severely impair quality of life (
      • Chee S.-N.
      • Murrell D.F.
      Pemphigus and quality of life.
      ,
      • Sebaratnam D.F.
      • Frew J.W.
      • Davatchi F.
      • Murrell D.F.
      Quality-of-life measurement in blistering diseases.
      ,
      • Tabolli S.
      • Pagliarello C.
      • Paradisi A.
      • Cianchini G.
      • Giannantoni P.
      • Abeni D.
      Burden of disease during quiescent periods in patients with pemphigus.
      ). Many severity scores have been proposed to assess pemphigus severity but none of them is widely used, because these scores did not result from an international consensus of experts (
      • Martin L.
      • Murrell D.F.
      Measuring the immeasurable: a systematic review of outcome measures in pemphigus.
      ).
      However, the precise assessment of the extent of skin and mucosal lesions in pemphigus patients is a major goal both for clinicians to adapt treatment and for investigators to robustly evaluate new treatment options in clinical trials. Two severity scores, the Autoimmune Bullous Skin Disorder Intensity Score (ABSIS) and the Pemphigus Disease Area Index (PDAI), have been proposed to evaluate the extent of pemphigus lesions (
      • Pfütze M.
      • Niedermeier A.
      • Hertl M.
      • Eming R.
      Introducing a novel Autoimmune Bullous Skin Disorder Intensity Score (ABSIS) in pemphigus.
      ,
      • Rosenbach M.
      • Murrell D.F.
      • Bystryn J.-C.
      • Dulay S.
      • Dick S.
      • Fakharzadeh S.
      • et al.
      Reliability and convergent validity of two outcome instruments for pemphigus.
      ). The interrater reliability of ABSIS and PDAI scores has been evaluated in two preliminary studies that mainly included patients with mild to moderate pemphigus, many of whom were already treated and were assessed only once, with no follow-up evaluations (
      • Murrell D.F.
      • Dick S.
      • Ahmed A.R.
      • Amagai M.
      • Barnadas M.A.
      • Borradori L.
      • et al.
      Consensus statement on definitions of disease, end points, and therapeutic response for pemphigus.
      ,
      • Rahbar Z.
      • Daneshpazhooh M.
      • Mirshams-Shahshahani M.
      • Esmaili N.
      • Heidari K.
      • Aghazadeh N.
      • et al.
      Pemphigus disease activity measurements: Pemphigus Disease Area Index, Autoimmune Bullous Skin Disorder Intensity Score, and Pemphigus Vulgaris Activity Score.
      ). As stated in an editorial in the Journal of Investigative Dermatology, which presented one of these studies, “The assessment of large, well-defined populations with a formal evaluation, including the usual successive methodological steps, is needed before the PDAI and ABSIS scores should be used as a research instrument” (
      • Bastuji-Garin S.
      • Sbidian E.
      How to validate outcome instruments for pemphigus.
      ).
      Therefore, we conducted an international multicenter study to prospectively assess the interrater reliability of ABSIS and PDAI scores in a large population of 116 newly diagnosed pemphigus patients with a wide spectrum of disease activity (mild/moderate, significant, and extensive) who were followed for up to 2 years. In particular, we aimed to fulfill all the methodological criteria that were detailed in the editorial to thoroughly investigate the validity of ABSIS and PDAI scoring systems in pemphigus. We also assessed the correlation between these severity scores and other clinical and biological markers of severity, including Physician Global Assessment (PGA) score, Dermatology Quality of Life Index (DLQI), and ELISA values of serum anti-DSG1 and anti-DSG3 antibodies at baseline and after the start of treatment, to measure the improvement of lesions; we also measured these in relapsing patients to assess the worsening of skin and/or mucosal lesions.

      Objectives

      The primary objective was to assess the interrater reliability of ABSIS and PDAI scores at baseline, first on the whole population of patients and then on subpopulations depending on (i) type of involvement (skin, using the ABSIS skin and PDAI skin subscores, or mucosal, using the ABSIS mucosa and PDAI mucosa subscores) and (ii) pemphigus extent (moderate, significant, or extensive).
      Secondary objectives were (i) to assess the interrater reliability of the ABSIS and PDAI scores during patients’ follow-up, (ii) to assess the correlation of these scores with the PGA and DLQI scores and ELISA values of serum anti-DSG1 and anti-DSG3 antibodies, and (iii) to assess the time to complete the scores.

      Results

      Baseline characteristics of patients

      Overall, 116 patients (68 women and 48 men) were enrolled in the study. Thirty-two patients had pemphigus foliaceus (PF), and 84 had pemphigus vulgaris (PV) (23 with exclusive mucosal involvement, 32 with exclusive skin involvement, and 61 with mucosal and skin lesions). Mean age was 53 years (standard deviation = 14.9 years, range = 19–84 years).
      At baseline, median (range) ABSIS and PDAI activity scores of the whole population were 37 out of 206 points (0.5–124.5 points) and 27.5 out of 250 points (3–114 points), respectively. Median (range) PGA score was 6 out of 10 points (1–10 points), and median DLQI score was 8 out of 30 points (0–29 points). The number of patients with PV or PF; moderate, significant or extensive pemphigus; and median corresponding PDAI, ABSIS, PGA, and DLQI scores and anti-DSG1 and anti-DSG3 antibody ELISA values are shown in Table 1.
      Table 1Baseline clinical characteristics and desmoglein ELISA values of pemphigus patients
      CharacteristicsABSIS
      According to the thresholds proposed by Boulard et al. (2016).
      , Median (Range)
      PDAI
      According to the thresholds proposed by Boulard et al. (2016).
      ,

      Median (Range)
      PGA
      According to the thresholds proposed by Boulard et al. (2016).
      ,

      Median (Range)
      Anti-DSG1
      According to the thresholds proposed by Boulard et al. (2016).
      in IU/ml,

      Median (Range)
      Anti-DSG3
      According to the thresholds proposed by Boulard et al. (2016).
      in IU/ml,

      Median (Range)
      Whole population (n = 116)36.8 (0.5–124.5)27 (3–114)6 (1–10)129 (1– 2,065)240 (1– 5,217)
      Pemphigus type
       Pemphigus vulgaris (n = 84)39 (0.5–106)29 (4–114)6 (1–10)52 (1–2,065)750 (1–5,217)
      Only mucosal involvement (n = 23)39.3 (6–56)23 (5–76)6 (3–10)5 (1–75)162.5 (1– 3,800)
      Skin and mucosal involvement (n = 61)36 (0.5–106)32 (4–114)6 (1–10)118 (1– 2,068)910 (1– 5,217)
       Pemphigus foliaceus (n = 32)25.3 (1–124.5)24.5 (3–95)6 (1–9)640 (4– 2,800)1 (1–179)
      Pemphigus extent
      According to the thresholds proposed by Boulard et al. (2016).
       Mild/moderate (n = 26)6 (0.5–15)10 (3–14)4 (1–3)6 (1–1,960)185 (1– 4,450)
       Significant (n = 60)36 (17–52.8)27 (15–44)6 (4–7)89.5 (1– 2,800)183 (1– 5,217)
       Extensive (n = 30)57.5 (53–124.5)64.5 (45–114)8 (8–10)207.5 (1– 2,320)622.5 (1–3,975)
      Abbreviations: ABSIS, Autoimmune Bullous Skin Disorder Intensity Score; IU, international units; PDAI, Pemphigus Disease Area Index; PGA, Physician Global Assessment.
      1 According to the thresholds proposed by
      • Boulard C.
      • Duvert Lehembre S.
      • Picard-Dahan C.
      • Kern J.S.
      • Zambruno G.
      • Feliciani C.
      • et al.
      Calculation of cut-off values based on the ABSIS and PDAI pemphigus scoring systems for defining moderate, significant and extensive types of pemphigus.
      .

      Interrater reliability

      At baseline, a high interrater reliability was observed for both the ABSIS (intraclass correlation coefficient [ICC] = 0.90; 95% confidence interval [CI] = 0.85–0.93) and PDAI scores (ICC = 0.91, 95% CI = 0.87–0.94) and, to a lesser degree, for the PGA score (ICC = 0.80, 95% CI = 0.72–0.86).
      According to pemphigus extent, the baseline ICC values for PDAI were significantly higher in patients with mild/moderate (ICC = 0.82; 95% CI = 0.63–0.92) and extensive (ICC = 0.80, 95% CI = 0.62–0.90) extent than in patients with significant extent (ICC = 0.50, 95% CI = 0.27–0.68), with P = 0.017 and P = 0.022, respectively. ABSIS ICC values were borderline significantly higher in patients with significant (ICC = 0.69, 95% CI = 0.51–0.81) and extensive pemphigus (ICC = 0.75, 95% CI = 0.51–0.88) than in those with mild/moderate pemphigus (ICC = 0.44, 95% CI = 0.004–0.74), with P = 0.178 and P = 0.116, respectively. Finally, ICC (95% CI) values for PGA in mild/moderate, significant, and extensive pemphigus were 0.51 (0.07–0.83), 0.56 (0.37–0.71), and 0.65 (0.35–0.86), respectively.
      According to type of involvement (skin or mucosa), the ICC values (95% CI) for PDAI skin and ABSIS skin were 0.97 (0.96–0.98) and 0.96 (0.94–0.97), respectively, and those for PDAI mucosa and ABSIS mucosa were 0.91 (0.87–0.94) and 0.96 (0.94–0.97), respectively. In cases of discrepancy between investigators, the mean (± standard deviation) interrater difference was higher for assessment of mucosal lesions than for skin lesions: PDAI mucosa, 7.5 ± 10 points versus PDAI skin, 4.5 ± 4.3 points, and ABSIS mucosa, 6.7 ± 6.7 points versus ABSIS skin, 4.9 ± 6.6 points.
      Figure 1 shows the evolution of the PDAI, ABSIS, and PGA scores during the 2-year follow-up. Figure 2 shows that the ABSIS and PDAI ICCs remained higher than 0.70 during follow-up. In particular, the ICC values (95% CI) for PDAI and ABSIS at month (M) 1 were 0.84 (0.77–0.89) and 0.90 (0.88–0.95), respectively, and those at M3 were 0.70 (0.58–0.79) and 0.91 (0.86–0.94), respectively. The slightly lower ICC values observed at M3 for PDAI score corresponded to an ICC (95% CI) of 0.72 (0.61–0.80) for PDAI skin and of 0.87 (0.81–0.91) for PDAI mucosa.
      Figure thumbnail gr1
      Figure 1Evolution of ABSIS, PDAI, and PGA severity scores and DLQI quality of life score during patients’ follow-up. The vertical bars correspond to the 95% confidence intervals. ABSIS, Autoimmune Bullous Skin Disorder Intensity Score; DLQI, Dermatology Quality of Life Index; PDAI, Pemphigus Disease Area Index; PGA, Physician Global Assessment.
      Figure thumbnail gr2
      Figure 2Interrater reliability of ABSIS and PDAI scores (intraclass coefficient correlation) during patients’ follow-up. The vertical bars correspond to the 95% confidence intervals. ABSIS, Autoimmune Bullous Skin Disorder Intensity Score; PDAI, Pemphigus Disease Area Index.

      Correlation between severity scores and other markers of disease severity

      Baseline correlations

      At baseline, Spearman coefficient correlation was r = 0.57 (P < 0.0001) between ABSIS and PDAI scores, r = 0.68 (P < 0.0001) between PDAI and PGA scores, and r = 0.60 (P < 0.0001) between ABSIS and PGA scores. ABSIS skin and PDAI skin subscores were highly correlated (r = 0.87, P < 0.0001), as were mucosal subscores (r = 0.85, P < 0.0001) (Figure 3).
      Figure thumbnail gr3
      Figure 3Baseline correlations of total scores and skin or mucosa ABSIS and PDAI subscores with PGA and anti-DSG ELISA values. Scatter diagram depicting the baseline correlation between (a) ABSIS, PDAI, and PGA severity scores and (b) PDAI skin and ABSIS skin and PDAI mucosa and ABSIS mucosa subscores with anti-DSG1 and anti-DSG3 ELISA values, respectively. ABSIS, Autoimmune Bullous Skin Disorder Intensity Score; PDAI, Pemphigus Disease Area Index; PGA, Physician Global Assessment.
      At baseline, PDAI skin and ABSIS skin subscores were highly correlated with anti-DSG1 ELISA values: r = 0.84 (P < 0.0000) and r = 0.77 (P < 0.0001), respectively. PDAI mucosa and ABSIS mucosa subscores were also correlated with anti-DSG3 ELISA values: r = 0.62 (P < 0.0001) and 0.57 (P < 0.0001), respectively (Figure 3).
      Baseline ABSIS and PDAI scores were both weakly correlated with the DLQI score in the whole sample: r = 0.24 (P = 0.02) and r = 0.33 (P = 0.001), respectively. According to the type of pemphigus, the PDAI score was weakly correlated with the DLQI score in PV (r = 0.30, P = 0.01) and PF patients (r = 0.39, P = 0.05). No correlation was found between ABSIS and DLQI scores in the PV and PF subpopulations.

      Correlations during disease course

      Overall, 100 of 116 patients achieved complete remission (ABSIS and PDAI scores = 0) at any time during the study, including 18 patients (15.5%) at the M1 evaluation and 39 (33.6%) and 68 (58.6%) patients at the M3 and M6 evaluations, respectively.
      Among these latter patients, anti-DSG1 and anti-DSG3 antibodies were observed in 8 (11.7%) and 23 patients (33.8%), with mean ELISA values of 113.6 ± 62.9 IU/ml (anti-DSG1 antibodies) and 152.5 ± 235.8 IU/ml (anti-DSG3 antibodies), respectively. More precisely, 6 patients had anti-DSG1antibodies alone, 21 had anti-DSG3 antibodies alone, and 2 had both anti-DSG1 and anti-DSG3 antibodies.
      During the follow-up, ABSIS and PDAI scores remained correlated (r = 0.55, P < 0.0001). Both ABSIS and PDAI scores remained correlated with PGA: r = 0.45 (P < 0.0001) and r = 0.54 (P < 0.0001), respectively; they were no longer correlated with the DLQI score. We then evaluated the correlation between the ABSIS and PDAI scores during the first 3-month period after the start of treatment to know whether these scores are useful for evaluating the initial improvement of patients’ condition and the effect of treatment. From baseline to M1, ABSIS and PDAI scores remained correlated (r = 0.54, P < 0.0001), as they did for the M1 to the M3 evaluations (r = 0.68, P < 0.0001). Table 2 shows the correlation between ABSIS and PDAI scores during this initial period, depending on the mild, moderate, or severe type of pemphigus. Apart from a poor correlation (r = 0.2) from baseline to M1 in patients with mild types of pemphigus, the correlation between ABSIS and PDAI scores was between 0.44 and 0.78 for moderate and severe pemphigus subtypes during this initial 3-month period of treatment.
      Table 2Correlations between ABSIS and PDAI scores from baseline to M1 and M3 according to the mild, moderate, or severe types of pemphigus
      Type of PemphigusBaseline to M1M1 to M3
      Mild types
       Spearman rank correlation coefficient0.19860.6305
       95% CI–0.3258 to 0.62960.1805 to 0.8623
      P0.44470.0088
      Moderate types
       Spearman rank correlation coefficient0.43710.7795
       95% CI0.1218 to 0.67220.5846 to 0.8894
      P0.0068<0.0001
      Severe types
       Spearman rank correlation coefficient0.61340.4843
       95% CI0.1533 to 0.8549–0.05398 to 0.8045
      P0.01150.0673
      Abbreviation: CI, confidence interval; M, month.
      The absolute improvement of the PDAI skin and ABSIS skin activity subscores was highly correlated with the absolute decrease of anti-DSG1 antibodies: r = 0.75 (P < 0.0001) and r = 0.71 (P < 0.0001), respectively. Conversely, only a weak correlation was observed between the absolute change in the PDAI mucosa and ABSIS mucosa subscores and the absolute change in anti-DSG3 antibodies: r = 0.37 (P < 0.001) and r = 0.32 (P = 0.003), respectively. To further assess the correlation between severity scores and anti-DSG antibodies during disease course, we studied the correlation between the ABSIS, PDAI, and their respective skin and mucosal subscores with the evolution of anti-DSG1 and anti-DSG 3 ELISA values in the 17 patients who relapsed during the 2 years of follow-up. Because patients relapsed at different time points during follow-up, we compared the evolution of the ABSIS and PDAI scores and that of anti-DSG1 and anti-DSG3 antibodies between the last evaluation and serum collection performed during the period of complete remission (before relapse) and the next evaluation performed during skin and/or mucosal relapse. Whereas the ABSIS and PDAI scores remained correlated (r = 0.64, P = 0.0061), we did not observe a correlation between the PDAI skin and ABSIS skin and anti-DSG1 antibodies (r = 0.28, P = 0.2841 and r = 0.41, P = 0.1343, respectively), nor between the PDAI mucosa and ABSIS mucosa and anti-DSG3 antibodies (r = 0.24, P = 0.3636 and r = 0.27, P = 0.3014, respectively).

      Time for instrument completion

      The mean time for carrying out the PDAI was 5.2 minutes. It was progressively quicker, from 8.3 minutes at baseline to 3.6 minutes at month 24. Nine percent of investigators found the PDAI easy to use, 21% quite easy, 35% practical, 24% difficult, and 11% imprecise or too long. The mean time for ABSIS completion was 5.4 minutes, and this also decreased over time from 8.1 minutes at baseline to 3.9 minutes at month 24. Six percent of investigators found the ABSIS easy to use, 13% quite easy, 31% practical, 39% difficult, and 11% imprecise or too long.

      Discussion

      Our results definitely validate the ABSIS and PDAI scores, showing an excellent interrater reliability for both scores (ICC ≥ 0.90) and a good correlation between these two scores (= 0.55). The inter-rater reliability of the ABSIS was high in scoring patients with significant or extensive pemphigus (ICC = 0.69 and ICC = 0.75, respectively), whereas it was lower in scoring patients with mild/moderate extent (ICC = 0.44). This is likely because the ABSIS uses the rule of nine to estimate body surface involvement, which makes the scoring of patients with limited pemphigus extent difficult. The PDAI score had a high interrater reliability for scoring patients with mild/moderate and extensive pemphigus (ICC = 0.82 and ICC = 0.80, respectively) and a lower reproducibility for assessing patients with intermediate extent (ICC = 0.50). This is a quite common weakness in many scoring systems, such as the Psoriasis Area Severity Index in psoriasis, which is due to the heterogeneity of patients with intermediate disease extent (
      • Paul C.
      • Gourraud P.-A.
      • Bronsard V.
      • Prey S.
      • Puzenat E.
      • Aractingi S.
      • et al.
      Evidence-based recommendations to assess psoriasis severity: systematic literature review and expert opinion of a panel of dermatologists.
      ). It may seem surprising that ICC estimates were lower in subgroups than for the overall sample: for example, 0.91 for the PDAI ICC for the overall sample compared with 0.82, 0.50, and 0.80 for mild/moderate, significant, and extensive forms, respectively. However, this was expected, because ICC estimates are known to be lower when observations are more homogeneous, which is the case in subgroups compared with the overall sample. In case of discrepancy, the difference between investigators was rather low for the assessment of skin lesions (4.5 to 5 points) and slightly higher for the assessment of mucosal lesions (6.7 to 7.5 points).
      Our results show a higher reproducibility for both PDAI and ABSIS scores than was reported by
      • Rosenbach M.
      • Murrell D.F.
      • Bystryn J.-C.
      • Dulay S.
      • Dick S.
      • Fakharzadeh S.
      • et al.
      Reliability and convergent validity of two outcome instruments for pemphigus.
      in a study of 15 patients (0.76 and 0.77, respectively) and a lower reproducibility than was reported by
      • Rahbar Z.
      • Daneshpazhooh M.
      • Mirshams-Shahshahani M.
      • Esmaili N.
      • Heidari K.
      • Aghazadeh N.
      • et al.
      Pemphigus disease activity measurements: Pemphigus Disease Area Index, Autoimmune Bullous Skin Disorder Intensity Score, and Pemphigus Vulgaris Activity Score.
      in a cross-sectional study (0.98 and 0.97, respectively). These two studies mainly included already-treated patients, most of whom had low to moderate disease activity as a consequence and who had been evaluated once only with no follow-up.
      The ABSIS and PDAI scores were not only reliable instruments to measure pemphigus extent at baseline, but they also showed a high reproducibility during follow-up, both to measure the improvement of lesions after the start of treatment and also to assess the worsening of skin and/or mucosal lesions in relapsing patients. This feature is particularly important for assessing the evolution of skin and/or mucosal lesions under treatment, because as previously reported, only anti-DSG1, but not anti-DSG3, antibodies seem useful for following the course of pemphigus patients (
      • Abasq C.
      • Mouquet H.
      • Gilbert D.
      • Tron F.
      • Grassi V.
      • Musette P.
      • et al.
      ELISA testing of anti-desmoglein 1 and 3 antibodies in the management of pemphigus.
      ,
      • Patsatsi A.
      • Kyriakou A.
      • Giannakou A.
      • Pavlitou-Tsiontsi A.
      • Lambropoulos A.
      • Sotiriadis D.
      Clinical significance of anti-desmoglein-1 and -3 circulating autoantibodies in pemphigus patients measured by Area Index and Intensity score.
      ). Indeed, this study showed that anti-DSG3 antibodies were weakly correlated with the evolution of mucosal lesions over time, suggesting that the PDAI and ABSIS mucosa subscores are more accurate than anti-DSG3 ELISA values for assessing the evolution of mucosal lesions and adapting treatment. In particular, whereas the correlation between ABSIS and PDAI scores in relapsing patients (r = 0.64) remained in the same order of magnitude as at baseline (r = 0.55), these scores were only weakly correlated with the evolution of anti-DSG antibodies under treatment, reinforcing the usefulness of the ABSIS and PDAI scores in the follow-up of patients.
      The evolution of the PGA score was also correlated with the ABSIS and PDAI scores. However, the PGA score is based on a physician’s subjective impression, with no clear definition of what the different marks from 0 to 10 correspond to. Additionally, it does not allow separate assessment of the evolution of skin and mucosal lesions. Similarly, the evolution of the DLQI score was not correlated with the evolution of pemphigus lesions and cannot be used as a tool to adapt treatment in pemphigus patients. The recently published Autoimmune Bullous Disease Quality of Life and Treatment of Autoimmune Bullous Disease Quality of Life scores might be better correlated with disease activity (
      • Tjokrowidjaja A.
      • Daniel B.S.
      • Frew J.W.
      • Sebaratnam D.F.
      • Hanna A.M.
      • Chee S.
      • et al.
      The development and validation of the treatment of Autoimmune Bullous Disease Quality Of Life Questionnaire, a tool to measure the quality of life impacts of treatments used in patients with autoimmune blistering disease.
      ).
      Finally, these scores are feasible in clinical practice, taking an average of approximately 5 minutes, which is a little bit longer than the time previously reported (
      • Rahbar Z.
      • Daneshpazhooh M.
      • Mirshams-Shahshahani M.
      • Esmaili N.
      • Heidari K.
      • Aghazadeh N.
      • et al.
      Pemphigus disease activity measurements: Pemphigus Disease Area Index, Autoimmune Bullous Skin Disorder Intensity Score, and Pemphigus Vulgaris Activity Score.
      ). Investigators generally considered the PDAI easier to use than the ABSIS.
      This study has several major strengths. First, it was performed prospectively on a large, multicenter, international cohort of pemphigus patients who were followed up for 2 years. Because it only included newly diagnosed pemphigus cases not yet treated, this study allowed assessment of these scores not only in patients with mild to moderate pemphigus activity but also in patients with significant and extensive pemphigus. It fulfills all the methodological criteria that were recently detailed as necessary to thoroughly investigate the validity of the ABSIS and PDAI scoring systems in pemphigus (
      • Bastuji-Garin S.
      • Sbidian E.
      How to validate outcome instruments for pemphigus.
      ).
      Selection bias is unlikely in this study, because it included consecutive patients recruited both in secondary and tertiary care centers with mild/moderate, significant, or extensive pemphigus. The M1 and M3 dates for early evaluation of these scores were chosen by a panel of international experts, because these dates correspond to common time points used to evaluate disease activity and the effect of treatment in most clinical trials on pemphigus.
      Few patients (15.5% and 33.6%, respectively) were in complete remission at these dates, thus allowing a valid assessment of these scores in patients who still had active disease. Despite the extremely high interrater reliability, evidenced both at baseline and at the M1 and M3 evaluations, weekly evaluations would have reinforced the validity of our results.
      Overall, this study provides strong evidence that the ABSIS and PDAI scores are robust tools to accurately assess pemphigus activity, both at the time of diagnosis and during disease course, which is of major interest for clinicians to adapt treatment. The high interrater reliability of these scoring systems will allow valuable intergroup comparisons of disease activity in randomized clinical trials. Finally, one might question which scoring system investigators should choose. We found that the PDAI score had a higher reproducibility than the ABSIS in scoring patients with mild pemphigus. Conversely, it had a lower reproducibility than the ABSIS in assessing patients with intermediate extent. Overall, 30% and 19% of investigators found the PDAI and ABSIS scores, respectively, easy or quite easy to use. Finally, the PDAI has been developed by a consensus of international experts, whereas the ABSIS was proposed by German investigators. Further research is needed to assess the prognostic value of these scores.

      Patients and Methods

      Study Population

      We conducted a prospective, international, multicenter study in 31 French, German, Italian, Swiss, and Croatian departments of dermatology (secondary and tertiary care centers) between July 2009 and May 2015. Consecutive patients aged 18 years or older with newly diagnosed pemphigus were included. Diagnosis of either PV or PF was based on (i) characteristic clinical features; (ii) histological analysis of a skin or mucosal biopsy showing acantholysis, intraepithelial blistering, or eosinophilic spongiosis; (iii) direct immunofluorescence examination showing IgG and/or C3 deposits on keratinocyte cell membrane; and (iv) detection of circulating anti-DSG1 and/or anti-DSG3 autoantibodies by ELISA assays (
      • Amagai M.
      • Komai A.
      • Hashimoto T.
      • Shirakata Y.
      • Hashimoto K.
      • Yamada T.
      • et al.
      Usefulness of enzyme-linked immunosorbent assay using recombinant desmogleins 1 and 3 for serodiagnosis of pemphigus.
      ,
      • Kasperkiewicz M.
      • Ellebrecht C.T.
      • Takahashi H.
      • Yamagami J.
      • Zillikens D.
      • Payne A.S.
      • et al.
      ). Treatments were not controlled in this study. They varied among countries and mainly consisted of oral corticosteroids alone or associated with conventional immunosuppressants or rituximab in some patients.
      All patients gave signed informed consent before inclusion. The study was approved by the corresponding local ethics committee.

      Assessment of disease extent

      Disease extent was evaluated at baseline and during the follow-up visits at M1, M3, M6, M12, and M24 by the same two investigators blinded to the results of each other and using ABSIS, PDAI, and PGA scores. All investigators were dermatologists with extensive experience in the diagnosis and treatment of pemphigus patients.
      The ABSIS score of skin involvement is based on the extent of the body surface area assessed using Wallace’s “rule of nines” and type of skin lesions (
      • Livingston E.H.
      • Lee S.
      Percentage of burned body surface area determination in obese and nonobese patients.
      ). The body surface area value is then multiplied by an index reflecting the predominant lesions: 1.5 (erosive exudative lesions, bullae, or Nikolsky sign positivity), 1.0 (erosive dry lesions), or 0.5 (re-epithelialized lesions). ABSIS oral involvement is evaluated by scoring 11 mucosal sites by 1 (presence of lesions) or 0 (absence of lesions) and by completing a subjective severity scale based on discomfort during eating and drinking. The ABSIS ranges from 0 to 206 points, with 150 points for skin involvement, 11 points for oral involvement, and 45 points for subjective discomfort; higher scores denote worse disease. A score between 0 and 16 corresponds to moderate pemphigus, between 17 and 52 to significant pemphigus, and higher than 52 to extensive pemphigus (
      • Boulard C.
      • Duvert Lehembre S.
      • Picard-Dahan C.
      • Kern J.S.
      • Zambruno G.
      • Feliciani C.
      • et al.
      Calculation of cut-off values based on the ABSIS and PDAI pemphigus scoring systems for defining moderate, significant and extensive types of pemphigus.
      ).
      The PDAI has a score ranging from 0 to 263 points, with 250 points representing disease activity (120, 10, and 120 points for skin, scalp, and mucosal activity, respectively) and 13 points representing disease damage. However, the damage component was not included in our analysis. For skin activity assessment, 12 anatomic sites are assigned a score according to disease extent: 0 (no lesions), 1 (1–3 lesions, up to 1 lesion > 2 cm in any diameter, all ≤ 6 cm), 2 (2–3 lesions, at least 2 lesions > 2 cm, all ≤ 6 cm), 3 (> 3 lesions, all ≤ 6 cm), 5 (> 3 lesions and/or 1 lesion > 6 cm), or 10 (> 3 lesions and/or at least 1 lesion > 16 cm or entire area affected). Scalp activity is assigned a score based on the presence of blisters, erosions, or erythema of 0 (no activity), 1 (one quadrant affected), 2 (two quadrants affected), 3 (3 quadrants affected), 4 (whole skull affected), or 10 (at least 1 lesion > 6 cm). For mucosal activity assessment, 12 mucosal sites are assigned a score based on the presence of erosions or blisters: 0 (absent), 1 (1 lesion), 2 (2–3 lesions), 5 (> 3 lesions or 2 lesions > 2 cm), or 10 (entire area). A score between 0 and 14 corresponds to moderate pemphigus, between 15 and 44 to significant pemphigus, and higher than 44 to extensive pemphigus (
      • Boulard C.
      • Duvert Lehembre S.
      • Picard-Dahan C.
      • Kern J.S.
      • Zambruno G.
      • Feliciani C.
      • et al.
      Calculation of cut-off values based on the ABSIS and PDAI pemphigus scoring systems for defining moderate, significant and extensive types of pemphigus.
      ).
      The PGA is a visual analog 10-point scale, based on a physician’s subjective impression from 0 (no lesions) to 10 (worst skin and mucosae condition imaginable). It has been used in clinical trials because it is fast and easy to use (
      • Tabolli S.
      • Pagliarello C.
      • Paradisi A.
      • Cianchini G.
      • Giannantoni P.
      • Abeni D.
      Burden of disease during quiescent periods in patients with pemphigus.
      ).
      Patients’ quality of life was evaluated by the DLQI translated into different languages (
      • Kasperkiewicz M.
      • Ellebrecht C.T.
      • Takahashi H.
      • Yamagami J.
      • Zillikens D.
      • Payne A.S.
      • et al.
      ). It includes 10 questions, with a total score between 0 and 30. The DLQI was used because the Autoimmune Bullous Disease Quality of Life and Treatment of Autoimmune Bullous Disease Quality of Life were not available at the time the study was designed.
      Serum samples were collected at each visit, stored on site, and centrally analyzed at the end of the study in the Immunology Laboratory of Rouen University Hospital for measurement of anti-DSG1 and anti-DSG3 antibody ELISA values, using commercially available ELISA-DSG1 and ELISA-DSG3 assays (Euroimmun, Lübeck, Germany). ELISA values higher or equal to 20 IU/ml were considered positive. Beyond 200 IU/ml, which corresponds to the upper limit of the assay, additional dilutions were performed.

      Statistical analysis

      Scores were prospectively recorded on standardized forms. Patients with more than one missing score out of four were excluded from the analysis. The target sample size (n = 100) was calculated for the primary objective of this study, which was to assess the interrater reliability through estimation of the ICC with good precision, as measured by the width of the 95% ICC CI. With n = 100, the expected width was ±0.15 for an ICC of 0.5 and ±0.07 for an ICC of 0.8. These figures translated into respective widths of ±0.21 and ±0.10 for half the overall sample size (n = 50), that is, the expected size of the subgroup with significant extent, and into respective widths of ±0.31 and ±0.14 for one quarter of the overall sample size (n = 25), that is, the expected size of the subgroups with mild/moderate or extensive extent.
      Interrater reliability was assessed by estimating the ICC overall and according to severity (mild/moderate, significant, extensive). This was done at each study visit for the overall ABSIS and PDAI scores, their respective skin and mucosal subscores (ABSIS skin and PDAI skin; ABSIS mucosa and PDAI mucosa), and per severity subgroups (mild/moderate, significant, extensive).
      Correlations between ABSIS and PDAI scores and with other severity markers (PGA, DLQI, and anti-DSG1 and anti-DSG3 antibody ELISA) were assessed using Spearman rank correlation coefficient. Correlations were assessed for baseline values of these scores or other severity markers and for their absolute changes between time 0 and M6. Correlations were not calculated after 6 months of follow-up, because severity scores were returned to 0 or almost 0 in most patients. SAS, version 9.3 (SAS institute, Cary, NC) and GraphPad Prism, version 5.0 (GraphPad Software, San Diego, CA) software were used.

      Acknowledgments

      The study was supported by grants from the Seventh Framework Programme of the European Union (FP7-HEALTH-2007-2.4.4-1), natural course and pathophysiology of rare diseases, and the French Society of Dermatology .

      References

        • Abasq C.
        • Mouquet H.
        • Gilbert D.
        • Tron F.
        • Grassi V.
        • Musette P.
        • et al.
        ELISA testing of anti-desmoglein 1 and 3 antibodies in the management of pemphigus.
        Arch Dermatol. 2009; 145: 529-535
        • Amagai M.
        • Klaus-Kovtun V.
        • Stanley J.R.
        Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion.
        Cell. 1991; 67: 869-877
        • Amagai M.
        • Komai A.
        • Hashimoto T.
        • Shirakata Y.
        • Hashimoto K.
        • Yamada T.
        • et al.
        Usefulness of enzyme-linked immunosorbent assay using recombinant desmogleins 1 and 3 for serodiagnosis of pemphigus.
        Br J Dermatol. 1999; 140: 351-357
        • Anhalt G.J.
        • Labib R.S.
        • Voorhees J.J.
        • Beals T.F.
        • Diaz L.A.
        Induction of pemphigus in neonatal mice by passive transfer of IgG from patients with the disease.
        N Engl J Med. 1982; 306: 1189-1196
        • Bastuji-Garin S.
        • Sbidian E.
        How to validate outcome instruments for pemphigus.
        J Invest Dermatol. 2009; 129: 2328-2330
        • Boulard C.
        • Duvert Lehembre S.
        • Picard-Dahan C.
        • Kern J.S.
        • Zambruno G.
        • Feliciani C.
        • et al.
        Calculation of cut-off values based on the ABSIS and PDAI pemphigus scoring systems for defining moderate, significant and extensive types of pemphigus.
        Br J Dermatol. 2016; 175: 142-149
        • Chee S.-N.
        • Murrell D.F.
        Pemphigus and quality of life.
        Dermatol Clin. 2011; 29: 521-525
        • Kasperkiewicz M.
        • Ellebrecht C.T.
        • Takahashi H.
        • Yamagami J.
        • Zillikens D.
        • Payne A.S.
        • et al.
        Pemphigus. Nat Rev Dis Primers. 2017; 3: nrdp201726
        • Langan S.M.
        • Smeeth L.
        • Hubbard R.
        • Fleming K.M.
        • Smith C.J.P.
        • West J.
        Bullous pemphigoid and pemphigus vulgaris—incidence and mortality in the UK: population based cohort study.
        BMJ. 2008; 337: a180
        • Livingston E.H.
        • Lee S.
        Percentage of burned body surface area determination in obese and nonobese patients.
        J Surg Res. 2000; 91: 106-110
        • Martin L.
        • Murrell D.F.
        Measuring the immeasurable: a systematic review of outcome measures in pemphigus.
        Australas J Dermatol. 2006; 47: A32-A33
        • Murrell D.F.
        • Dick S.
        • Ahmed A.R.
        • Amagai M.
        • Barnadas M.A.
        • Borradori L.
        • et al.
        Consensus statement on definitions of disease, end points, and therapeutic response for pemphigus.
        J Am Acad Dermatol. 2008; 58: 1043-1046
        • Patsatsi A.
        • Kyriakou A.
        • Giannakou A.
        • Pavlitou-Tsiontsi A.
        • Lambropoulos A.
        • Sotiriadis D.
        Clinical significance of anti-desmoglein-1 and -3 circulating autoantibodies in pemphigus patients measured by Area Index and Intensity score.
        Acta Derm Venereol. 2014; 94: 203-206
        • Paul C.
        • Gourraud P.-A.
        • Bronsard V.
        • Prey S.
        • Puzenat E.
        • Aractingi S.
        • et al.
        Evidence-based recommendations to assess psoriasis severity: systematic literature review and expert opinion of a panel of dermatologists.
        J Eur Acad Dermatol Venereol. 2010; 24: 2-9
        • Pfütze M.
        • Niedermeier A.
        • Hertl M.
        • Eming R.
        Introducing a novel Autoimmune Bullous Skin Disorder Intensity Score (ABSIS) in pemphigus.
        Eur J Dermatol. 2007; 17: 4-11
        • Rahbar Z.
        • Daneshpazhooh M.
        • Mirshams-Shahshahani M.
        • Esmaili N.
        • Heidari K.
        • Aghazadeh N.
        • et al.
        Pemphigus disease activity measurements: Pemphigus Disease Area Index, Autoimmune Bullous Skin Disorder Intensity Score, and Pemphigus Vulgaris Activity Score.
        JAMA Dermatol. 2014; 150: 266-272
        • Rosenbach M.
        • Murrell D.F.
        • Bystryn J.-C.
        • Dulay S.
        • Dick S.
        • Fakharzadeh S.
        • et al.
        Reliability and convergent validity of two outcome instruments for pemphigus.
        J Invest Dermatol. 2009; 129: 2404-2410
        • Sebaratnam D.F.
        • Frew J.W.
        • Davatchi F.
        • Murrell D.F.
        Quality-of-life measurement in blistering diseases.
        Dermatol Clin. 2012; 30: 301-307
        • Stanley J.R.
        • Amagai M.
        Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome.
        N Engl J Med. 2006; 355: 1800-1810
        • Tabolli S.
        • Pagliarello C.
        • Paradisi A.
        • Cianchini G.
        • Giannantoni P.
        • Abeni D.
        Burden of disease during quiescent periods in patients with pemphigus.
        Br J Dermatol. 2014; 170: 1087-1091
        • Tjokrowidjaja A.
        • Daniel B.S.
        • Frew J.W.
        • Sebaratnam D.F.
        • Hanna A.M.
        • Chee S.
        • et al.
        The development and validation of the treatment of Autoimmune Bullous Disease Quality Of Life Questionnaire, a tool to measure the quality of life impacts of treatments used in patients with autoimmune blistering disease.
        Br J Dermatol. 2013; 169: 1000-1006