Relationship of Anti-Desmoglein-3 Antibody Titres by ELISA with ClinicalABSIS Scoring in Pemphigus Vulgaris

Introduction

Pemphigus vulgaris (PV) is a severe, potentially life-threatening autoimmune blistering disorder characte-rized by autoantibodies against Desmoglein (Dsg) 3.¹ Currently, the diagnosis of PV is based on a concert of clinical features, histopathology, and immuno- fluorescence.^2,3 Recently, an enzyme-linked immuno-sorbent assay (ELISA) test demonstrating antibodies against Dsg1 and 3 (Dsg-3) has been developed, which is highly sensitive and specific.⁴ The sensitivity of ELISA is 96% in Pemphigus Foliaceus for Dsg 1 and about 98% in PV for Dsg-3, whereas specificity is around 99%.⁵ Zhou et al., demonstrated that the sensitivity of ELISA was 84.8% and that the specificity was 96.7%.⁶

Dsg-3 titres are an important tool for diagnosis and monitoring disease activity, as well as for deciding treatment, but are not routinely used in clinical practice in India, in contrast to the Western countries.⁴ Although recent studies have documented that ELISA is a more accurate test for confirming the diagnosis of PV than Indirect IF, ELISA tests are underutilized in clinical laboratories.⁶ There has been one study from North India about Dsg 3 utility.⁷ This study aimed to detect antibodies to Dsg-3 in PV using ELISA and correlate levels of Dsg-3 with the clinical severity score (ABSIS Score).

Materials and Methods

The study was conducted at a tertiary care hospital after obtaining IEC clearance. Twenty- seven patients diagnosed as PV over two years were enrolled for performing ELISA for Dsg-3 after taking informed consent.

These cases were diagnosed according to the S2k guidelines for diagnosis of Pemphigus, it is recommended to make a diagnosis of PV/PF if the following constellation of findings is present:

• Compatible clinical picture and positive DIF.

• Appropriate clinical picture and reactivity with Desmoglein 3 by ELISA.

• Compatible clinical picture corresponding histo-pathology, and positive indirect IF microscopy on monkey oesophagus sections.⁸

The demographic, clinical, and treatment details were collected from the medical records. The diagnosis was confirmed by histopathology and/or IF. Each patient was scored using the ABSIS score, which has a maximum score of 206. The ABSIS score was chosen not just because of its simplicity and ease of evaluation but because it provides both qualitative and quantitative information, thus improving its interobserver reliability. It assesses certain subjective qualities, such as relief of pain in eating, which may go unaccounted for by a clinician who only assesses the number of oral and skin lesions. It consists of a skin involvement score and an oral involvement score. The skin involvement score uses the rule of nine, which is used in burns measurement, to assess the percentage of involvement of blisters and erosions on the skin combined with a weighting factor for the stage of the blistering and erosions [skin involve-ment total score = % body surface area (BSA) × weighting factor = 0-150 points]. Erosive, exudative lesions have a weighting factor of 1.5; erosive, dry lesions have a weighting factor of 1.0; and re-epithelialized lesions have a weighting factor of 0.5. Oral involvement is based on two scores: the extent of the disease (presence of lesions) and the severity (discomfort during eating and drinking) of the disease. The extent is given a score of 0 or 1 (absence or presence, respectively) for 11 different mouthparts. The severity of oral lesions is assessed by evaluating the pain/bleeding associated with certain foods. The maximum scores for oral involvement are 11 for extent and 45 for severity.¹

Two ml of serum was collected from 27 patients on two occasions, at least 6 weeks apart. The first sample was collected at the time of diagnosis, and the second, a minimum of 6 weeks after instituting therapy. Most patients were treated using methylprednisolone pulse therapy, except four, using rituximab. Semi-quantitative ELISA was done for Dsg-3 on both serum samples. A correlation of ELISA titres with the clinical severity scoring was done.

A semi-quantitative in vitro assay of anti Dsg-3 was done on the patient’s serum samples using EUROIMMUN Anti-Desmoglein 3 ELISA (IgG) kit (EUROIMMUN, Lübeck, Germany). The kit literature recommends the upper limit of the normal range (cut-off) as 20 relative units (RU/ml). EUROIMMUN recommends interpreting the results as follows: < 20 RU/ml - negative, >= 20 R/ml - positive.

This recommendation is based on data from an ROC analysis using results from 71 patient samples with Pemphigus vulgaris and 470 control samples. A specificity of 99.4% at a cut-off value of 15.9 RU/ml was obtained.

Statistical analysis

The statistical software SPSS 11 was used for data analysis. Continuous variables were reported using the mean and standard deviation or median and interquartile range as appropriate. Categorical variables were reported using frequency and percentage. The median comparison was done using a non-parametric test. The cutaneous, oral mucosal, and total ABSIS scores were calculated before each sample collection and compared with the ELISA titres. Spearman rank correlation was done to compare the clinical severity score with the ELISA titers. The results were graphically represented. All the analysis was considered statistically significant at a 5 % P value. (P value< 0.05).

 Results

Twenty-seven patients with PV were enrolled in the study after obtaining informed consent.

The mean age of the patients was 42.8 years (21 to 65 years) with a male-to-female ratio of 1.8:1. Of these 27 patients, nine patients had only cutaneous involvement whereas the rest had both skin and mucosal involvement during the first clinical assessment.

The minimum interval between the first and second assessments was six weeks, and the maximum interval was 29 weeks. The mean interval was 12.2 weeks. (Tables 1, 2 and 3).

We correlated both the components of the ABSIS score (cutaneous and oral mucosal) separately and the total ABSIS score with the ELISA titres. This correlation was done for the first and second samples separately.

We received a positive association for all except the correlation of the oral mucosal score of the second set of samples with the ELISA titres, which showed a negative association. However, for all these correlations, the P value was > 0.05, indicating that statistically significant differences were not found (Table 4).

Discussion

With the use of corticosteroids and immunosuppressive therapy, the survival rate in cases of PV has improved dramatically.⁹ However, some recent studies have found an increase in the number of deaths. Older age of onset, i.e., people in the age group of 60-70 years, are likely to have a bad prognosis. An increased inflammatory status characterized by ESR>30 mm/hour is said to be associated with increased disease activity and increased risk of cardiovascular disease, all of which are bad prognostic indicators for this disease.¹⁰

Anti-Dsg antibodies act as an independent risk factor for PV. PV patients with anti-Dsg 1 titres of >100 U/ml are associated with a lower overall survival. However, the mean relapse-free time is significantly shorter in anti-Dsg3 positive patients compared to anti-Dsg3 negative patients. The presence of high titres of antibodies against Dsg-1 signifies a worse prognosis, especially in mucocutaneous PV. In the cutaneous form, it may suggest an underlying cutaneous infection.¹⁰

Currently, DIF, IIF, and ELISA are the tools that are commonly being used for prognosis. Saha et al., in addition, used PCR-SSP (Polymerase chain reaction with sequence-specific primers) to detect Class II HLA DRB1 allele groups *04 and *14 which are known to be associated with Pemphigus.¹¹

Although antibodies against Dsg-3 are the principal antibodies involved in PV, both Dsg-3 and Dsg-1 antibodies are required to induce skin blisters. However, only patients with Dsg-3 antibodies without detectable Dsg-1 antibodies develop oral blisters. This could explain the differential requirements of Dsg-1 and Dsg-3 antibodies in blister formation of mucosa vs skin. Also, there are very few studies on Dsg-3 titres at diagnosis and their prognostic significance, with varied results. In this study, we detected only Dsg-3 titres (and not Dsg-1) owing to financial constraints.¹²

Weiss et al. found a constant increase in mean ELISA values with worsening disease, suggesting a correlation with disease progression and can be used to monitor the disease. Thus, it is imperative to accurately measure the baseline ELISA values for Dsg 3 and 1.¹³ Relapses are common in PV, and the prediction of relapse generally relies upon clinical, histological, and immunofluorescence features. Recently, ELISA has been found helpful in predicting the occurrence of relapses.¹⁴

Anand et al. found that patients who did not relapse had mean anti-Dsg1 and anti-Dsg3 values comparatively lower than other patients. They found that anti-Dsg 1 and 3 levels were highest in the pre-treatment phase and subsequently rose again in the occurrence of a relapse ELISA allows us to analyze many samples in a relatively short period.¹⁴ It also provides quantitative and reproducible data for the measurement of Dsg 1 and 3.¹⁴

In our study, in 14 out of 27 cases, the variation of Dsg 3 levels was found to correlate with variations in ELISA values, i.e., an increase/decrease in the total ABSIS score was accompanied by an increase/decrease in the ELISA titres, respectively. (Table 1) In a study done by Abdel et al., anti-Dsg3 antibody titre was positive in 25/26 PV patients (96.2%), whereas Khandpur et al. found that, of the 54 cases, 53(98.15%) had elevated anti-Dsg3 titre.^15,16

In 10 cases in the present study, the ELISA score had increased/ remained high despite a decline in the total ABSIS score (Table 2). This could be due to an impen-ding relapse or the fact that levels take time to become negative. Pfutze et al., supported this hypothesis in their study where they assessed ABSIS in 13PV patients for six months after initiation of immunosuppressive therapy and found that the decrease of the ABSIS skin score was accompanied by a gradual decrease of anti- Dsg1 and anti-Dsg3.¹⁷ A study by Rahbar et al., to determine the predictors of anti-Dsg3 values, revealed that other unknown independent variables are involved in high titres of anti-Dsg 3, possibly non-pathogenic antibodies that do not contribute to disease activity.¹⁸

In 3 cases, an increase in the total ABSIS score was accompanied by a decreased/ negative ELISA value (Table 3). This can be seen in cases where the antibodies are directed against proteins other than Dsg-3, such as Dsg-1, Desmocollin, etc. Also, the antibodies could be directed against different subclasses of Dsg-3, which this study did not estimate. Harman et al., also cited the exact reason for the above-mentioned discrepancy.¹⁹

The answer to the discrepant results in the last two scenarios could also lie in other factors, such as genetic factors or factors related to the storage of the serum samples. Harman et al., hypothesized that this may be due to the simplicity of our scoring system, which may be unable to detect subtle influences on severity.¹⁹

We found a positive association between the total score and cutaneous ABSIS score and the ELISA titres in the first and second sets of samples. However, on correlating the oral mucosal score with the ELISA titres in the first sample, we found a positive association. However, a negative one was found correlating with the second set of samples.

De et al. compared salivary and serum Dsg-1 and Dsg-3 levels with ABSIS scores in patients with Pemphigus. While they found a correlation between serum and salivary Dsg-3 levels, there was no correlation between salivary or serum Dsg-3 levels with ABSIS score, similar to our findings. In contrast, Dsg-1 was found to correlate with the ABSIS score.⁷ Patsatsi et al. also reported similar results, where Dsg-3 scores did not correlate with disease activity, but Dsg1 did. Dsg-3 was detectable 6 months after remission, and ABSIS score only seemed to correlate with Dsg-3 after 12 months, especially in those patients with predominant mucosal lesions. This indicates that while Dsg-3 is helpful for initial diagnosis, it does not appear to mirror disease activity.²⁰

Thus, while the total score and the cutaneous ABSIS score correlated positively with the Dsg-3 ELISA titres, no definite correlation was found with the oral mucosal score. However, the P value for these correlations was, however > 0.05, indicating that statistically significant differences were not found.

The limitations of this study were the small sample size, ELISA done only for Dsg-3, and not Dsg-1 due to financial constraints. Also, only two samples were collected per patient and no further follow-up samples were taken. The time between the first and second assessments was variable as the patients did not always return for the second assessment in the stipulated time.

Conclusion

In conclusion, although ELISA for Dsg 3 is a very sensitive tool for the initial diagnosis of PV, its' utility in assessing disease activity is restricted and should be interpreted in the clinical context.

Sources of support

None

Conflict of interest

None