Background: Contact Dermatitis (CD) is a one of the most common inflammatory skin diseases. Distinguishing the different types of CD, especially allergic CD (ACD) versus non-allergic forms, is often challenging and requires additional investigations based on patch-testing. Early detection of relevant allergy biomarkers in active CD lesions could refine and simplify the management of CD patients.
Objectives: To characterize the molecular signatures of active CD lesions.
Materials and methods: We studied the expression of 12 allergy biomarkers (identified in a recent transcriptomic study) by qRT-PCR in eczema lesions of 17 CD patients. ACD was established on patch-test (PT) results and through exposure assessment. Molecular signatures of active lesions were compared to those of reference chemical allergens and irritants, and machine learning algorithm was used to develop a model to predict the allergic nature of the skin reaction.
Results: 6 of the 17 CD patients were diagnosed with ACD. The remaining 11 patients did not react to any of the allergens tested. Gene profiling of active CD lesions revealed two distinct molecular patterns: patients harbouring signatures similar to reference allergens (n=7), or irritants (n=10). Among the 7 patients with an allergy signature, we found 6 patients with confirmed ACD, and 1 other subject for whom it was not possible to identify the culprit allergen. In contrast, the 10 other patients had negative PT, suggesting that these patients developed non allergic reactions. Finally, using a machine learning approach, we confirmed that our biomarker combination correctly classified ACD from non-allergic CD.
Conclusions: Molecular signatures from active lesions could help to stratify patients and open new avenues for the development of a future point of care diagnosis of CD.

_Background_: Contact Dermatitis (CD) is a one of the most common inflammatory skin diseases. Distinguishing the different types of CD, especially allergic CD (ACD) versus non-allergic forms, is often challenging and requires additional investigations based on patch-testing. Early detection of relevant allergy biomarkers in active CD lesions could refine and simplify the management of CD patients.
_Objectives_: To characterize the molecular signatures of active CD lesions.
_Materials and methods_: We studied the expression of 12 allergy biomarkers (identified in a recent transcriptomic study) by qRT-PCR in eczema lesions of 17 CD patients. ACD was established on patch-test (PT) results and through exposure assessment. Molecular signatures of active lesions were compared to those of reference chemical allergens and irritants, and machine learning algorithm was used to develop a model to predict the allergic nature of the skin reaction.
_Results_: 6 of the 17 CD patients were diagnosed with ACD. The remaining 11 patients did not react to any of the allergens tested. Gene profiling of active CD lesions revealed two distinct molecular patterns: patients harbouring signatures similar to reference allergens (n=7), or irritants (n=10). Among the 7 patients with an allergy signature, we found 6 patients with confirmed ACD, and 1 other subject for whom it was not possible to identify the culprit allergen. In contrast, the 10 other patients had negative PT, suggesting that these patients developed non allergic reactions. Finally, using a machine learning approach, we confirmed that our biomarker combination correctly classified ACD from non-allergic CD.
_Conclusions_: Molecular signatures from active lesions could help to stratify patients and open new avenues for the development of a future point of care diagnosis of CD.

ESDR236 - Detection of skin allergy biomarkers for point of care diagnosis of contact dermatitis

50th Annual Meeting of the European Society for Dermatological Research. VIRTUAL MEETING