Background: Vancomycin, a glycopeptide antibiotic routinely used to treat severe Gram-positive bacterial infections, is associated with the development of drug reaction with eosinophilia and systemic symptoms (DRESS) in individuals expressing HLA-A*32:01. Previous studies have identified the potential role of T-cells using HLA-A*32:01 positive healthy donor models. However, DRESS pathogenesis remains poorly defined and a deeper mechanistic understanding is now required to aid the diagnosis and prediction of vancomycin-induced DRESS. The present study aims to elucidate CD4+ and CD8+ T-cell involvement within the pathogenesis of vancomycin-induced DRESS following the isolation and functional study of cloned T-cells from hypersensitive patients. Methods: CD4+ and CD8+ vancomycin-responsive T-cell clones (TCCs) were generated by serial dilution from PBMC samples collected from suspected vancomycin-DRESS patients. Functionality of drug-responsive TCCs was assessed using T-cell proliferation ([ 3H]-thymidine). Cytokine analysis was performed using intracellular cytokine staining (ICS), ELISpot assay and LEGENDplex immunoassays. Results: Vancomycin-responsive TCCs expressing CD4+ and CD8+ phenotypes were successfully generated from suspected vancomycin-DRESS patients (n=3). CD45RO + memory T-cells were the primary activated population, with both CD4+ and CD8+ T-cells associated with the release of IFN-γ, IL-5, IL-13, granzyme B and perforin. Vancomycin-responsive CD4+ and CD8+ T-cells are activated by direct, pharmacological interactions, with antigen presentation possible through both HLA class I and HLA class II molecules. Conclusion: This study provides in vitro evidence for the dual role of antigen-specific CD4+ and CD8+ T-cells within the pathogenesis of vancomycin-induced DRESS. This has been demonstrated following the generation of cloned T-cells with strong vancomycin specificity from patients presenting with vancomycin-DRESS and positive for expression of HLA-A*32:01.

Utility and safety of skin tests in drug reaction with eosinophilia and systemic symptoms (DRESS): a systematic review Running title : Skin tests in drug reaction with eosinophilia and systemic symptoms (DRESS)Ying Xin Teo MB BChir1,2 , Peter Simon Friedmann MD FRCP1,2, Marta Ewa Polak PhD1, Michael Roger Ardern-Jones DPhil FRCP1,2Clinical Experimental Sciences, Faculty of Medicine, University of Southampton, Southampton SO16 6YD, United Kingdom2 Department of Dermatology, Southampton General Hospital, University Hospitals Southampton NHS Foundation TrustAcknowledgements : noneFunding sources: This work was supported by the British Skin Foundation [grant number BSF023F19].Conflict of interest: The authors have no conflicts of interests to declareText word count: 752

Background Determination of culprit drug in drug reaction with eosinophilia and systemic symptoms (DRESS) is crucial. Skin tests have been utilised, although it remains unclear how sensitive these are. We set out to determine the value of skin tests in the assessment of drug causality in DRESS. Methods A systematic literature search was conducted for publications from 1996 onwards of skin tests (skin prick test = SPT, patch test = PT, intradermal test = IDT) performed in clearly defined DRESS cases. Outcomes of testing, drug culpability assessments and challenge tests data were extracted. Results 17 articles met inclusion criteria. In 290 DRESS patients, patch testing was most frequent [PT = 97.2% (n=282), IDT = 12.4% (n=36), SPT = 3.1% (n=9)]. Positive results were noted in 58.4% (n=160/282) of PT, 66.5% IDT and 25% SPT. When confidence of drug causality was high (n=73 of 194), testing did not correlate well with clinical suspicion: PT 37.6%; IDT 36.5% (n=19 of 52). Direct comparison of skin testing with provocation testing (n=12) showed 83.3% correlation. Positive IDTs were reported in 8 negative PT cases. Conclusion Skin tests, particularly PT and IDT, have been reported as tools for diagnosis of causal drugs in DRESS. Heterogeneity in methodology, results analysis and reporting of cohorts, makes meta-analysis to determine sensitivity and specificity of published literature impossible and highlights weaknesses in the field. We propose that international collaboration is essential to harmonise the methodology and reporting measures from hypersensitivity testing studies in larger cohorts.