In our Slavic DH patients, the ratio between men and women was 1.6
1, which is consistent with the reports that DH affects men slightly more often [18
In this study, we attempt to assess diagnostic usefulness and accuracy of tTG/npG ELISA tests in the management of Slavic DH patients. Curiously, the research literature data about DH and studies on its pathomechanism are amazingly scanty in relation to the other autoimmune blistering dermatoses, while it seems that DH pathogenesis is far more complex. As mentioned above, GSE, both CD and DH, is a growing medical problem, particularly among young people, with nutritional impact on their health status. Thus, specific, reliable, and objective criteria for diagnosing and monitoring of DH should be established. DIF of nonlesional skin remains a definitive laboratory test for diagnosing this disease, however its invasiveness is a serious limitation for screening. Indirect immunofluorescence is a time-consuming, expensive, and subjective technique; then, ELISA seems to be free from those disadvantages. At present, ELISA is the method of choice for serological screening of DH. However, there is a problem what kind of ELISA-based test is the best. The medical diagnostics market offers a wide range of ELISA kits with biotechnologically obtained eTG, tTG, and npG as the most frequent antigen sources for diagnosing DH. Literature data indicated that there are discrepancies regarding the specificity and repeatability of test results. Some findings showed that 52% of DH patients have anti-eTG IgA elevated [20
], which is consistent with our previous study [21
]. On the other hand, several communications revealed that more than 90% of DH patients have anti-eTG IgA elevated [19
]. Such divergent results might be caused by evaluating series of DH patients differing in severity of cutaneous rash. In the light of this, there is an urgent need to find/develop a new or modified antigen/epitope that would make the DH diagnostics more accurate. Thus, the usefulness of tTG/npG ELISA-based tests should be considered. Literature associated with this problem demonstrated that there is also incompatibility about the level of anti-tTG IgA. Some investigators [20
] obtained only 25% of anti-tTG IgA-positive results in DH patients, whereas other researchers achieved about 79% of positive results in their patients [19
]. Our own experience in this area is satisfactory: 90% of DH patients examined in this study had anti-tTG IgA above normal range. Important issue remains the composition and structural design of the antigen/epitope. Byrne et al. [24
] indicated that the use of novel mutagenic variant of tTG lacking the catalytic triad decreases the binding of IgA to the mutant tTG with the mean reduction of 58% in DH and even bigger mean reduction of 79% in CD samples. Fernández et al. [25
] analyzed six different human anti-tTG ELISA kits. This group of researchers showed that there are differences in the sensitivity and specificity of the human tTG ELISA assays. Furthermore, they suggested that diagnostic accuracy of tests was significantly improved by adjusting the cutoff thresholds according to ROC curve analyses, which was done in our study. Manufacturer's cut-off is not standardized to each laboratory conditions, therefore the standardization based on ROC curve analyses should be recommended to all ELISA tests performed for diagnostic purposes. Interestingly, Fernández et al. [25
] demonstrated that the correction of the cut-off with the use of the ROC curve analysis modifies the decision limit in more than 50% in five of the six examined anti-human tTG ELISA kits. Hence, there is the evidence that the way/source of production and further modification of antigen have the impact on test accuracy. Researchers and diagnosticians should take it into account before the choice of the appropriate test.
], investigators indicated the usefulness of synthetic deamidated gliadin-derived peptides (GDR) as antigen, which is useful for the detection of sensitivity to gluten in anti-tTG IgA seronegative DH patients. There is a hypothesis that they are the most reliable tools for identification of gluten sensitivity in DH patients [12
]. Our study demonstrated the presence of anti-npG IgA in 90% of examined DH patients, which may imply that usage of the ELISA test measuring IgA antibodies to this antigen broadens the information necessary to make the correct diagnosis. Presented results correspond to those obtained by Kasperkiewicz et al. [23
] indicating that 84% of DH patients have IgA to npG detected with GAF-3X ELISA. In this study, we determined the cutoff value for examined DH patients in anti-tTG ELISA as the level 17.199
RU/mL (for all examined groups), while the manufacturer's cut-off is 20
RU/mL. In anti-npG ELISA, the cut-off for examined population is 24.633
RU/mL (for 2 combinations of groups) and 24.08
RU/mL (for DH and healthy groups), while the manufacturer's cut-off is 25
RU/mL. Thus, there are negligible differences between them (especially in case of antigliadin GAF-3X ELISA). It can be explained by the fact that kits used are well standardized in a genetically similar population. It is known that the cutoff point can vary depending on examined population, for example, different cutoff value of anti-tTG ELISA for Italian and Spanish populations [28
]. Interestingly, the cutoff points in our studied groups were almost identical. It may be due to the fact that diseased group, IgA/neutrophil-mediated non-DH dermatoses, was chosen as pathogenetically most closely related to DH, as far as cutaneous pathology is concerned. Moreover, AUC obtained in this study was equal 1, which may be due to a perfect separation of the values of the examined groups. It should be stressed here that we had precise inclusion criteria and each diagnosis had to be confirmed with the combination of microscopic and biochemical/molecular techniques. Thus, using the new cutoff values derived from ROC curves, the diagnostic accuracy of the tests is improved (sensitivity and specificity of 100%). It could be due to fact that the tests were adapted to our native population.
There were no individuals with positive anti-tTG/npG IgA antibodies in the healthy controls, although we did not make the biopsy for DIF in order to rule out DH in these subjects presenting no cutaneous lesions whatsoever. In addition, the percentage of cases with positive anti-tTG/npG IgA antibodies in IgA/neutrophil-mediated non-DH dermatoses group has been quite low (4% of anti-tTG ELISA and 8% of anti-npG ELISA). Still, it cannot be excluded that some such cases had low-grade GST with no overt clinical symptoms in addition to their IgA/neutrophil-mediated non-DH dermatosis. Interestingly, findings obtained by Ludvigsson et al. [30
] in a nationwide cohort study indicated that individuals with CD were at increased risk of psoriasis both before and after CD diagnosis. Thus, one should be aware that CD can coexist with other dermatoses, not only with DH. In the light of the above, it is suggested that recognition of anti-npG IgA alone does not mean that one is dealing with DH as a cutaneous manifestation of GSE, and DIF of the uninvolved skin still remains crucial for diagnosing DH.
In our study, we demonstrated that the determination of anti-tTG/npG IgA by means of ELISA is a precise method to broaden the body of knowledge about DH patients in Slavic population. However, what was noticed by Fernández et al. [25
] is that it is necessary to select the ELISA kit with the highest sensitivity and specificity and recalculate the cutoff threshold using samples from any given native population. This sequence of actions is essential for the laboratory diagnostician to provide reliable information to the clinician to facilitate making right decision on the implementation of troublesome therapy with the risk of potentially dangerous sideeffects.
Statistically significant correlation between levels of IgA antibodies to tTG and npG in DH group (r
= 0.4149) found in this study and no such correlation in healthy controls (r
= 0.2231) might suggest that the anti-tTG and anti-npG IgA antibodies in the IgA/neutrophil-mediated DH, but not in healthy individuals, are produced in the coordinated way in contrast to healthy individuals. These findings are in agreement with our previous data [31
]. This might correspond with a suggestion that, as the catalytic site of tTG seems to be targeted by IgA autoantibodies, intermolecular epitope spreading from the gliadin epitopes to the catalytic site of tTG does take place in CD [24
In recent years, the cutaneous immunopathology of DH has become the area of extensive studies in humans and using animal models of the disease [21
]. We feel that the key issue in understanding the blister formation in DH is not simply linking it to the GSE, but evaluating instead local cutaneous factors, conceivably neutrophils' Fc receptors involvement, at the human skin level.
Nonetheless, it should be kept in mind, as far as the issue of importance of npG in pathogenesis and diagnostics of GSE is concerned, that within the gliadin peptide the N-terminal proline residue and the C-terminal glutamine residues were reported to be essential for antibody recognition in addition to the deamidated glutamine residue [34
]. Finally, our data might suggest that Anti-tTG IgA ELISA is marginally superior to Antigliadin (GAF-3X) IgA ELISA for differential diagnosis of cutaneous itchy rashes suspected to be DH at the clinical level. Plausibly, in connection with the data that IgA1 deposits predominate in the skin of DH sufferers, modification of the above ELISA tests to enable the determination of IgA1 subclass antibodies to tTG and npG would be even more valuable for differentiating IgA/neutrophil-mediated dermatoses.