METHODS FOR DIAGNOSIS AND MONITORING OF TOXIC EPIDERMAL NECROLYSIS
20240003879 ยท 2024-01-04
Inventors
Cpc classification
G01N2333/70596
PHYSICS
C07K16/2896
CHEMISTRY; METALLURGY
C12N15/1138
CHEMISTRY; METALLURGY
C07K2317/24
CHEMISTRY; METALLURGY
International classification
C07K16/28
CHEMISTRY; METALLURGY
C12N15/113
CHEMISTRY; METALLURGY
Abstract
In the present invention, inventors investigate the representation of T cell subsets in Toxic epidermal necrolysis (TEN) a life-threatening cutaneous adverse drug reaction (cADR), characterized by massive epidermal necrosis. To better understand why skin symptoms are so severe in TEN disease, inventors conducted a prospective immunophenotyping study on skin samples and blood from 18 TEN patients, using mass cytometry and next generation TCR sequencing. Deep sequencing of the T cell receptor CDR3 repertoire revealed massive expansion of unique CDR3 clonotypes in blister cells. Over-represented clonotypes were mainly effector memory CD8+CD45RACCR7 T cells, and expressed high levels of cytotoxic (Granulysin and Granzymes A & B) and activation (CD38) markers. Thus present invention relates to non-invasive, specific and rapid methods for diagnostic and monitoring Toxic Epidermal Necrolysis. More specifically present invention relates to methods for diagnosis and/or monitoring of Toxic Epidermal Necrolysis through detection of a specific population of T lymphocytes in a subject. The present invention also relates to a method of preventing or treating a Toxic Epidermal Necrolysis in a subject in need thereof.
Claims
1. A method for assessing a subject's risk of having or developing Toxic Epidermal Necrolysis and treating the subject, comprising i) determining in a sample obtained from the subject the level of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ markers, and ii) administering a CD38 inhibitor to a subject identified as having a level of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ markers that is higher than a corresponding reference value.
2. The method according to claim 1, wherein the sample is a blood sample or immune primary cells or a blister sample or a skin sample.
3. The method according to claim 2, wherein the immune primary cells are selected from the group consisting of PBMC, WBC and T lymphocytes.
4. A method for monitoring and treating a Toxic Epidermal Necrolysis comprising i) determining the level of a population of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ markers in a sample obtained from the subject at a first specific time of the disease, ii) determining the level of a population of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ markers in a sample obtained from the subject at a second specific time of the disease, and iii) administering a CD38 inhibitor to a subject identified as having a level determined at step ii) that is higher than the level determined at step i).
5. An in vitro method for monitoring the treatment of Toxic Epidermal Necrolysis comprising the steps of i) determining the level of a population of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ in a sample obtained from the subject before the treatment, ii) determining the level of a population of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ markers in a sample obtained from the subject after the treatment, iii) comparing the level determined at step i) with the level determined at step ii) and iv) concluding that the treatment is efficient when the level determined at step ii) is lower than the level determined at step i).
6. The in vitro method for monitoring according to claim 4, wherein the sample is a blood sample or immune primary cells or blister sample or skin sample.
7. The in vitro method for monitoring according to claim 4, wherein the immune primary cells selected from the group consisting of PBMC, WBC and T lymphocytes.
8. The method according to claim 1, wherein the level of the population of T lymphocytes having cell surface expression of CD8+CD45RACCR7CD38+ is determined by clonal expansion of said population.
9. A method of preventing or treating a Toxic Epidermal Necrolysis in a subject in need thereof, comprising, administering to the subject a therapeutically effective amount of a CD38 inhibitor.
10. The method according to claim 8 wherein the CD38 inhibitors is selected from: a) an inhibitor of CD38 activity and/or b) an inhibitor of CD38 gene expression.
11. The method according to claim 10 wherein said inhibitor of CD38 activity is a small organic molecule, an antibody, a CAR T cell or an aptamer.
12. The method according to claim 11, wherein the antibody is selected from the group consisting Daratumumab, Isatuximab, MOR202, TAK-079, TAK-169, AMG424 or GBR 1342.
13. The method according to claim 10 wherein the inhibitor of CD38 gene expression is an antisense oligonucleotide, a nuclease, siRNA, shRNA, or ribozyme nucleic acid sequence.
14. The method according to claim 9, wherein the subject is identified having a high level of T lymphocytes CD8+CD45RACCR7CD38+ in a biological sample, wherein the level by the methods of claim 1.
15. The method according to claim 14, wherein the biological sample is a blood sample or immune primary cells or a skin sample.
16. (canceled)
Description
FIGURES
[0190]
[0191]
[0192]
[0193] The black bar illustrates the threshold value from which TCR V chains were considered as highly expanded (using Tukey's rule for the detection of outliers, i.e. Q3+1.5IQR).
[0194]
[0195]
[0196]
[0197] *Of note, as no anti-V3 mAb exists for CyTOF, dominant TCRV3+ cells in patient TEN-10 (which represent 90% of total CD8+ T cells in skin) were gated by negative selection. We gated cells negative for TCR-V21.3+, -V13.2+ and -V7.2+ expression.
[0198]
[0199] NGS mice were reconstituted with 1010.sup.6 PBMCs from a healthy donor, and treated by two-weekly injections of an anti-CD38+ mAb (Daratumumab, at 100 or 300 microg/mouse). Control group received PBS. Results depict the percentage +/SD of CD38+ fraction among human CD8+ T cells present in the spleen, as evaluated by flow cytometry 28 days after PBMCs injection.
[0200]
[0201] NGS mice were reconstituted with 1010.sup.6 PBMCs from a healthy donor, and treated by two-weekly injections of an anti-CD38+ mAb (Daratumumab, at 100 or 300 microg/mouse). Control group received PBS. The percentage of humanization was measured by flow cytometry at day 5 and at day 12 after PBMC injection. It was calculated by dividing the percentage of human blood CD45+ cells/the percentage of total (mouse+human) blood CD45+ cells.
[0202]
[0203] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, 1 year after disease recovery. Animals were then administrated with lamotrigine (the culprit drug; 0.1 mg/kg/day) or vehicle by oral gavage, every day, from day 4. Results depict the kinetic of human CD45+ cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol, or the spleen at day 29. Results are expressed as mean and individual % of humanization, calculated according to the following formula: % human CD45+ cells/% (mouse+human) CD45+ cells. Six mice per group were used in this experiment.
[0204]
[0205] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient (1 year after disease recovery) or from a healthy donor. Animals were then administrated with lamotrigine (the culprit drug; 0.1 mg/kg/day) or vehicle, by oral gavage every day, from day 4. Results depict the expansion of human CD45+ cells measured by flow cytometry in the spleen of NSG mice, 29 days after cell transfer. Results are expressed as mean and individual % of humanization, calculated according to the following formula: % human CD45+ cells/% (mouse+human) CD45+ cells. Six mice per group were used in this experiment.
[0206]
[0207] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient (1 year after disease recovery). Animals were then administrated with lamotrigine (the culprit drug; 0.1 mg/kg/day), by oral gavage every day, from day 4. Results depict the expression of CD38 (A) and Granzyme B and Granulysin (B) markers on CD4+ and CD8+ T cells as measured by flow cytometry in the spleen of NSG mice, 43 days after cell transfer.
[0208]
[0209] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb injections from day 4 (preventive mode). Results depict the kinetic of human CD8+Vbeta7.1+ T cell expansion measured by flow cytometry in the spleen of NSG mice at day 31. Results are expressed as mean and individual % of Vbeta7.1+ cell fraction among total CD8+ T cells. Ten mice per group were used in this experiment. Of note, 3.2% of Vbeta7.1+ cells were detected in CD8+ T cells at the time of cell transfer (day 0).
[0210]
[0211] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb injections from day 4 (preventive mode). Results depict the kinetic of human CD4+CD38+/ and CD8+CD38+/ T cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol. Results are expressed as mean and individual number of CD38+(A) and CD38 (B) cells/mL blood. Ten mice per group were used in this experiment.
[0212]
[0213] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; or vehicle by oral gavage, every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb injections from day 4 (preventive mode). Results depict the kinetic of human CD4+ and CD8+ T cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol. Results are expressed as mean and individual number of CD4+ and CD8+ T cells/mL blood. Ten mice per group were used in this experiment.
[0214]
[0215] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; 0.1 mg/kg/day), every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb injections from day 4 (preventive mode). Results depict the kinetic of human cytotoxic CD8+CD38+/GranzymeB+Granulysin+ T cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol. Results are expressed as mean and individual number of CD8+CD38+GranzymeB+Granulysin+ and CD8+CD38+GranzymeB+Granulysin+ cells/mL blood (A). Details for total CD8+GranzymeB+Granulysin+ T cells are also shown (B). Ten mice per group were used in this experiment.
[0216]
[0217] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb treatment from day 4 (preventive mode) or from day 29 (curative mode). Results depict the kinetic of human CD8+CD38+/ T cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol. Results are expressed as mean and individual % of CD38+ and CD38 fractions among total CD8+ T cells. Five mice per group were used in this experiment.
[0218]
[0219] NSG animals were adoptively transferred at day 0 with 1.10.sup.6 millions PBMCs collected from a TEN patient, and successively administrated with lamotrigine (the culprit drug; every day, from day 4. In parallel, NSG mice were injected i.p. with 200 mg of daratumumab or a control isotype twice weekly, starting mAb treatment from day 4 (preventive mode) or from day 29 (curative mode). Results depict the kinetic of human cytotoxic CD8+ T cell expansion measured by flow cytometry in the blood of NSG mice throughout the protocol. Results are expressed as mean and individual number of CD8+GranzymeB+Granulysin+CD38+/ T cells/mL blood. Five mice per group were used in this experiment.
EXAMPLE
[0220] Methods:
[0221] Study Design
[0222] Patients were prospectively recruited by the drug allergy reference center at the Hospices Civils de Lyon (France) between 2014 and 2018. TEN or MPE diagnoses were based on the definition published by the RegiSCAR study group (43) (44). Only patients with a probable or a definite diagnosis of TEN or MPE were enrolled in this study. Culprit drugs in TEN patients were determined according to the Algorithm for Drug Causality for Epidermal Necrolysis (ALDEN) (45). For MPE patients, the main putative drug was also determined. We collected demographic and clinical information, including sex and age, as well as underlying diseases (i.e. the disease the culprit drug was prescribed for), comorbidities, duration of drug exposure before TEN/MPE onset and HLA-A/B genotyping results. HLA-A/B genotypes were determined by reverse PCR-sequence-specific oligonucleotide hybridization (LABType SSO, One Lambda). Complementary information were also collected for TEN patients: SCORTEN (SCORe of Toxic Epidermal Necrosis) at diagnosis, which aim to predict the severity of the disease (46) and percentage of skin detachment assessed by E-Burn smartphone application (Android Play store). The latter was determined when the patient was first diagnosed with TEN (initial), and when maximum involvement was observed (final). We enrolled 20 healthy donors as controls.
[0223] Local ethical committee approved the study and written informed consent was obtained from each participant. Given the observational nature of the translational study, there was no randomization or formal blinding process for the investigators.
[0224] Sample Collection and Processing
[0225] Skin Samples
[0226] Skin samples for TEN mainly consisted of blister fluids and for 3 patients blister fluids and skin biopsies. Supernatant was collected and cells were repeatedly washed in complete RPMI before subsequent processing. In cases of MPE and patients TEN-15, -17 and -18, 6-mm.sup.2 biopsies were performed directly into lesional erythematous skin. Abdominal skin leftovers, from healthy donors undergoing elective plastic surgery, were used as control biopsies. Skin cells were extracted by mechanical dissociation and enzymatic digestion (2 hours at 37 C. in RPMI supplemented with collagenase type 1 (1.25 U/mL, Sigma-Aldrich, Saint Quentin Fallavier, France), DNAse (4 KU/mL, Sigma-Aldrich) and HEPES buffer (5%)), before to be passed through a 100 mm cell strainer (ThermoFischer Scientific, Dardilly, France) to obtain single cell suspensions. Cell viability was determined by trypan blue exclusion.
[0227] Blood Samples
[0228] PBMCs from healthy donors and patients were isolated from whole blood samples (in Lithium-Heparin coated tubes) using Ficoll-histopaque (Ficoll-Paque PLUS, GE Healthcare Life Sciences) density gradient centrifugation, and cell viability was assessed as described above.
[0229] Depending on experiments, samples were either frozen in liquid nitrogen according to standard procedures, or immediately stained for immunophenotyping analysis.
[0230] Flow Cytometry Analysis
[0231] Flow cytometry was carried out using fluorescently labelled mAbs, recognizing human CD3 (7D6; Thermo Fisher Scientific, Les Ulis, France), CD4 (VIT4; Miltenyi biotech, Bergish Glabach, Germany) and CD8 (SK1; Biolegend, San Diego, California, USA) proteins. V-beta (V) chain repertoire expression was assessed using a kit of 24 TCR-V mAbs (IOTest BetaMark, Beckman Coulter, Roissy, France; which includes approx. 70% of the expressed human TCR V domains: TCR V 1, 2, 3, 4, 5.1, 5.2, 5.3, 7.1, 7.2, 8, 9, 11, 12, 13.1, 13.2, 13.6, 14, 16, 17, 18, 20, 21.3, 22, 23) and viability discrimination was performed by incubating cells with Live/dead eFluor-506 (eBioscience, San Diego, California, USA).
[0232] Cells were analyzed on a LSR FORTESSA flow cytometer (BD Biosciences, Franklin Lakes, New Jersey, USA) and data were analyzed using FlowJo software (v10, Ashland, Oregon, USA).
[0233] For TCR sequencing experiments, some dominant CD8+ TCR V+ cells were sorted on a FACSARIA IIu device (BD Biosciences).
[0234] Mass Cytometry Analysis
[0235] Mass cytometry antibodies were obtained as pre-conjugated metal-tagged antibodies from Fluidigm (South San Francisco, California, USA) or generated in-house by conjugating unlabelled purified antibodies (from Myltenyi or Beckman Coulter) to isotope-loaded polymers using Maxpar X8 Multi-Metal Labelling Kit (Fluidigm). After titration on Nanodrop ND 1000 (ThermoFischer) antibodies were diluted in antibody stabilization buffer (Candor-Biosciences, Wangen im Allgu, Germany) with 0.5% sodium azide (Sigma). A detailed list of the antibodies used in this study is provided in supplementary materials (Table S2). Cell identification was performed using Irridium-Intercalator (Fluidigm) and viability discrimination was assessed by staining cells with Cisplatin (194Pt, Fluidigm). In some experiments, cells were fixed and permeabilized using Cytofix/Cytoperm solution (Cytofix/Cytoperm, BD Biosciences, Le Pont de Claix, France) and next intra-cellularly stained with human anti-Granulysin, anti-Granzyme A, anti-Granzyme B, and anti-Perforin mAbs.
[0236] Before acquisition on HELIOS mass cytometer (Fluidigm) cells were resuspended in half-diluted Four Element Calibration Beads (Fluidigm), and data set were normalized with CyTOF software using Finck algorithm (47). Flow Cytometry Standard (FCS) 3.0 files were imported into FlowJo software v10, and analyses included standard gating to remove beads, aggregates or dead cells, and further identify main leucocyte subsets (data not shown).
[0237] High-Dimensional Mass Cytometry Data Analysis
[0238] An inverse hyperbolic sine transformation was applied to analyze TCR + CD8+ T cells (n=300 per samples, all CyTOF samples were used (Table 1), except skin samples from MPE-9 and -12, which were excluded from the analysis due to very low CD8+ T cell number, and TEN-18 samples due to technical problem). Data were next clustered using FlowSOM algorithm (48) (with FlowSOM R pluging downloaded in FlowJo v10). A self-organizing map (SOM) was first trained to gather all cells into 100 distinct nodes based on their similarities in high dimensional space (i.e considering the relative MFI of 16 markers simultaneously: CCR7, CD45RA, CD27, CD38, CD56, CD57, CD107a, CD137, CD226, CD253, CD255, Granzyme A, Granzyme B, Granulysin, Perforin, Annexin A1, and excluding cell-lineage: CD45, CD14, CD19, TCR, TCR, CD8a, CD8b, CD4, CD38, CD56, NKP46, CD11b, CD11c, TCRV14-Ja18, TCRV7.2. SOM nodes were subsequently grouped in different clusters (each representing different CD8+ T cell subsets) using K-parameter and/or K-Finder R package (https://arxiv.org/pdf/1811.07356.pdf) (based on the Tracy Widom algorithm to approximate K in sparse data matrices, K being the number of relevant clusters in a population). FlowSOM clusters were visualized as integrated (i.e. including all samples) or disease phenotype minimal spanning trees, and heatmaps showing the integrated or individual MFI of each marker per cluster were generated with FlowJo or Excel. Additional hierarchical metaclusterings were performed, using the gplots R package based on the Euclidean distance and Ward-linkage (49), to determine the immunophenotype or the frequency of each cluster per samples.
[0239] DNA Isolation and High-Throughput Sequencing of TCRa/b CD3R Regions
[0240] DNA was isolated from frozen total blister, skin and PBMC samples using QIAamp DNA Micro Kit (Qiagen, Courtaboeuf, France), according to manufacturer's instructions. Then, TRB and TRA CDR3 regions were amplified and sequenced using ImmunoSEQ assay (Adaptive Biotechnologies). In brief, bias-controlled V and J gene primers were used to amplify rearranged V(D)J segments spanning each unique CDR3b/a, and amplicons were next sequenced (at approx. 20 coverage) using the Illumina HiSeq platform. The assay was performed at survey level (detection sensitivity: 1 cell in 40,000). After correcting sequencing errors via a clustering algorithm, TCR/ V, D and J genes were annotated according to the IMGT database (http://www.imgt.org).
[0241] Sequencing data were analyzed according to the ImmunoSEQ Analyzer V.3.0 (http://www.immunoseq.com). Diverse TCR repertoire metrics were explored: frequency and overlap of highly expanded clones, respective nucleotide or amino acid CDR3 sequences, usage and pairing of TRB/AV, TRBD and TRB/AJ families or diversity of the TCR repertoire (clonality index based on Shannon's entropy).
[0242] Transduction of the V- and V-Chains of the TCR into Skw3 Cell Lines
[0243] Skw3 cell lines (Leibniz Institute DSMZ, Brunswick, Germany, (50)) were transduced as described previously (51). In brief, rearranged human variable TCR - and genes identified by TCR sequencing were synthesized by custom gene synthesis (GeneUniversal, Newark, Delaware, USA) and cloned into retroviral pMSCV Vector (Takara Bio USA Inc, Mountain View, California, USA) containing puromycin and neomycin resistance genes respectively. The resulting retroviruses were used to transduce the TCR-defective Skw3 cell line, which also expresses the human CD8 coreceptor. The TCR-transduced cells outgrowing in selective medium were picked, and the expression of the correct TCR and was further assessed by flow cytometry, using a FACS-Canto-I device (BD-Biosciences, San Jose, CA, USA). The transduced cells with stable TCR expression were selected for assessment of reactivity and specificity, which was measured by TCR-induced CD69 expression.
[0244] TCR-Transductant Stimulation Assay
[0245] Skw3 cell lines expressing the cognate TCR and chains were cocultured with autologous EBV-transformed B-lymphoblastoid cell lines (52) at 1:2 ratio at 37 C. Tested drugs were added to the cocultures with the indicated concentrations. After 24 h, cells were stained with anti-human CD3 (Biolegend) and anti-human CD69 (Biolegend) and analysed by flow cytometry. Levels of CD69 expression were monitored in 10,000 CD3.sup.+ events. Experiments were repeated at least 2 times.
[0246] Statistical Analysis
[0247] P-values were calculated with two-tailed independent Student's t tests or one-way analysis of variance (ANOVA) using GraphPad Prism software (v8, San Diego, California, USA). P-values <0.05 were considered significant.
[0248] The Tukey's rule for the detection of outliers (75th percentile (Q3)+1.5 inter-quartile range (IQR)) was used to identify highly expanded TCR V chains. Of note, all TEN, MPE and healthy donor data for each V chain were compiled to calculate IQR.
[0249] Results
[0250] Skin and blood samples were collected from 18 TEN and 14 MPE hospitalized patients at the acute phase of their disease. Samples were recovered within 0-2 days after their hospital admission and diagnosis, and within 0-5 days after the first symptoms (mainly fever and/or skin rash). Hence, majority of samples were collected before the peak of the disease, characterized for TEN patients by the maximal percentage of skin detachment (Table 1 & data not shown). Noteworthy, the majority of patients displayed very diverse HLA genotypes. A*02 and B*44 were the most represented loci (Table 1). A careful investigation of causative drug(s) associated to skin symptoms revealed a large variability in terms of drug nature or mode of action. The same molecule was reported as culprit drug only for pairs of TEN patients (Allopurinol for patients TEN-1 & -3; Sulfamethozaxole/Trimethoprim for TEN-2 & -5; and Ceftriaxone for TEN-10 & -11, (Table 1)).
[0251] Immunophenotype of Leukocytes Infiltrating the Skin of TEN Patients
[0252] We first examined the immunophenotype of cells infiltrating the skin of TEN patients by mass cytometry (cyTOF) and subsequent computational data analysis. Blister cell samples obtained from 7 TEN patients were analyzed by CyTOF using a panel of 29 antibodies (Table S2), enabling mapping of all major peripheral blood mononuclear cell subsets (data not shown). We detected a large predominance of conventional T lymphocytes (TCR+; meanstandard deviation (SD)=71.318.8%) among hematopoietic CD45+ cells, along with a minor infiltration of monocytes (CD14+ subset, 13.478.6%), NK (TCR-CD56+, 5.87.2%) cells, and very few gamma delta T (TCR+, 1.92.8%), B (CD19+, 0.60.6%) or dendritic cells (CD11c+, 3.45.9%) (
[0253] Similarly to TEN, the inflamed skin of MPE patients was infiltrated by conventional T lymphocytes (63.819.5% of hematopoietic CD45+ cells), and to a lesser extent, by CD14+ monocytes (12.38.1%) and NK cells (4.85.8%) (
[0254] Finally, we detected no major difference in the immunophenotype of cells circulating in the blood of TEN, MPE patients and healthy donors, with CD8+ T cells representing approximately a quarter of total TCR+ cells in all the tested samples (data not shown).
[0255] Collectively, these results thus confirm that the blistering and inflamed skin of TEN patients is extensively infiltrated by CD8+ T cells (14) (26) (20). By contrast, no major skewing was recorded for unconventional lymphocytes, NK cells or monocytes.
[0256] Clustering of Skin CD8+ T Cells into 7 Phenotypic FlowSOM Subsets
[0257] As CD8+ T cell-mediated cytotoxicity is key in the initiation and formation of drug-induced lesions, we investigated in detail the molecular cytotoxic expression patterns of CD8+ T cells in TEN blisters. We performed high dimensional profiling and investigated the (co-)expression of several cell death-associated molecules (Granulysin, Granzyme B, Granzyme A, Perforin, but also TRAIL (CD253), TWEAK (CD255), Annexin A1, CD107a), as well as different activation markers (CD27, CD38, CD56, CD57, CD137, CD226). Using concatenated CyTOF data from different samples (skin and peripheral blood mononuclear cells (PBMCs) from TEN, MPE but also healthy donors), we ran FlowSOM, a self-organizing map (SOM) clustering algorithm, to assess the heterogeneity of the CD8+ T cell population present in the different patients. FlowSOM first stratified the CD8+ T cell population into 100 nodes. Projected as minimal spanning tree (data not shown), each SOM node groups cells with similar phenotypes, with the node size representing the number of cells within that node (illustrations of minimal spanning tree obtained for each tissue sample are also shown (data not shown). SOM nodes were next gathered in 4 main clusters, as automatically calculated using K-finder Tree-level approach algorithm. Because K-finder approach did not capture the full diversity of the concatenated population (data not shown), we decided to increase the FlowSOM clustering to 7 distinct clusters (clusters A to G). To define the phenotype identity of each cluster, we generated a heatmap showing the integrated median fluorescence intensity (iMFI) values of each marker in each FlowSOM cluster (
[0258] A Polycytotoxic Signature Typifies Lesional CD8+ TEN T Cells
[0259] The in-depth FlowSOM analysis allowed a comparison of the frequency of the CD8+ T cell clusters in lesion (blisters and skin) and blood samples from TEN and MPE patients, and from healthy individuals (
[0260] These results thus establish that the major subset of TEN blister CD8+ T cells displays a hallmark CD38+ polycytotoxic effector memory cell phenotype (cluster C).
[0261] Restricted TCR V Repertoire Among TEN Blister and Blood CD8+ T Cells
[0262] Parallel to these studies, we also addressed TCR usage of T cells present in TEN blisters. FACS analysis conducted on 24 of the most common Vbeta (V) chains found a highly restricted TCRV repertoire usage in the 13 TEN patients tested, with single V expansions ranging from as much as 20% to 80% of total TCR-V chains expression, when compared to healthy donors (
[0263] Although less marked than in TEN blisters, TCR VD expansions were observed in CD8+ T cells (but not CD4+ T cells) from TEN PBMCs, with notable biases in patients TEN-3, -4, -5-6, -10, -11, -13 and -15 (data not shown). In contrast, a limited number of TCR V expansions were detected in CD8+ and CD4+ T cells isolated from MPE skin (
[0264] Massive Oligoclonal Expansion of Distinct CDR3 Clones in the Skin and Blood of TEN Patients
[0265] As FACS cannot catch the full spectrum of the TCR repertoire, we next used high-throughput sequencing (HTS) of the TCR CDR3 regions (the antigen recognition domains) to evaluate sample clonality. HTS was performed on total blister, skin and PBMC samples from TEN and MPE patients.
[0266] Investigations of TCR repertoire diversity, measured using Shanon entropy-based clonality index metric, first revealed the presence of a highly clonal repertoire in the blisters of approximately half of TEN patients (
[0267] In-depth analysis of TRBV repertoire next confirmed the existence of preferential TCR biases in the skin of 12 out of 15 TEN patients, which were the result of very limited numbers of CDR3 clonotype expansions (ranging from >10% to 90% of total TCR sequences for combined top 5 clones; except TEN-2, -8 and -14) (data not shown). Of note, clone-tracking analyses revealed (i) that expanded clones expressed the same VD chains cells as those observed by FACS (data not shown) and (ii) no sharing of identical TCR CDR3 nucleotide (data not shown) or amino acid (data not shown) sequences among the 15 TEN patients. An interesting exception was noted for one clone from patients TEN-6 and TEN-10, which shared amino acid but not nucleotide sequence (data not shown). As these patients were exposed respectively to Norfloxacin and Ciprofloxacin quinolones and both expressed HLA-B*73:03 (Table 1), potential epitope cross-reactivity expansion of clones sharing identical TCR beta chains.
[0268] Another interesting observation was noted for the 2 dominant clones from patient TEN-9 (
[0269] Conversely to TEN, similar TRBV repertoire analysis revealed that clonal expansions were rare for MPE patients, and were usually lower than 5% (data not shown).
[0270] Clonotypes that were massively expanded in the TEN blisters were also found elevated in the blood of respective patients, at least for the top 5 clones (data not shown). This result then indicates that the massive infiltration of unique clonotypes in TEN blisters was likely to be consecutive to a previous clonal expansion in the lymphoid organs. Only for patient TEN-15, and to a lesser extent for patients TEN-6 and TEN-11, were some of the highly expanded skin clones not represented in the blood (data not shown).
[0271] T Cell Repertoire Diversity and Clonal Expansion of Blister Clones Circulating in Blood Correlates with TEN Severity
[0272] TEN severity, assessed here as the percentage of final skin detachment, varied significantly after hospital admission (
[0273] Combined with the lack of major TCR CDR3 biases found in MPE samples (skin or blood) (
[0274] V-Expanded CD8+ T Cells Display the Polycytotoxic Phenotype Over-Represented in TEN Samples
[0275] Thereafter, by taking advantage of mass cytometry, we were able to track back highly V-expanded CD8+ T cells in the blisters and blood of TEN patients to analyse their phenotype. We first demonstrated that CD8+ T cells expressing dominant V chains (FACS analysis) displayed very high levels of Granulysin and CD38 markers, when compared to their non-dominant CD8+V+ T cell counterparts (
[0276] This analysis confirms that major V-expanded CD8+ T cells display the polycytotoxic phenotype that is over-represented in TEN samples.
[0277] Expanded Clones in Blisters and Blood are Drug-Specific
[0278] Ultimately, we sought to determine whether highly expanded and activated clones were drug specific. To this end, we FACS sorted dominant CD8+ TCRV+ T cells present in the blister fluids or the blood of 4 TEN patients (TEN-3,-7,-10,-15), and sequenced their TRAV repertoire. For most dominant clones, a productive rearrangement (data not shown) encoding a functional TCR alpha-chain, as well as a second non-productive TCR alpha-locus rearrangement (data not shown) were identified. Then, the productive rearranged TCR and TCR chains were transduced into Skw3 cells, a TCR defective lymphoma line (28) (data not shown). After verification of sustained and stable TCR expression (data not shown), transduced Skw3 cells were stimulated with the culprit (or control) drug in presence of autologous Epstein Barr Virus (EBV)-transformed B cells generated from patient's PBMCs. The following day, we measured CD69 expression at the surface of Skw3 cells, as marker for TCR stimulation.
[0279] Results showed a positive dose response for patient TEN-3 with oxypurinol (the metabolite of allopurinol, the culprit drug for TEN-3), but not with the parent drug or an irrelevant drug (sulfamethoxazole) (Table 2 & data not shown). A positive response was also found for patient TEN-7 with the culprit pantoprazole (Table 2). In contrast, we failed to detect robust CD69+ expression in transductants generated from patients TEN-10 and TEN-15, stimulated respectively with ceftriaxone and ciprofloxacin or levofloxacin and metronidazole (Table 2).
DISCUSSION
[0280] The main objective of our study was to gain further insights into TEN pathophysiology by tracking immune cells that are present in the skin and the blood of patients at disease onset. Our results confirm that CTLs are the main leucocyte subset found in TEN blisters, followed by a minor infiltration of CD14+ monocytes and NK cells; but we failed to repeatedly detect unconventional cytotoxic lymphocytes such as NKT, MAIT or gamma-delta T cells. Strikingly, deep sequencing of the T cell receptor CDR3 repertoire revealed that there was a massive expansion of unique CD8+ T cell clones in TEN patients (both in skin and blood), which express an effector memory phenotype and an elevated level of cytotoxic or inflammatory/activation markers such as Granulysin, Granzymes A & B or CD38. By transducing and chains of the expanded clones into immortalized T cells, we demonstrate that some of these clones were drug-specific. Importantly, T cell repertoire diversity analysis revealed that clonal expansion of blister clones circulating in the blood of TEN patients at the acute phase of the disease correlated with the final clinical severity (as defined by the maximal percentage of skin detachment).
[0281] Massive Expansion of Unique TCR Clonotypes in TEN Patients
[0282] The most striking observation of our study is certainly the demonstration that there is a dramatic expansion of unique polycytotoxic CD8+ T cell clones in TEN patients, which largely outnumbers the frequency of clonotypes expanding in less severe MPE patients. A few studies have already described oligoclonal expansion in TEN (or in the less severe Stevens-Johnson syndrome (SJS)). These studies focused on in vitro T cell (re)activation experiments, or used samples which were isolated from individuals with restricted HLA genotype (for instance HLA-B*15.02 ((4) (29)) and reactive to a limited number of compounds (mainly allopurinol and carbamazepine) (4) (30) (29) (31). They showed preferential usage of TRBV subtypes, clonal expansion of specific CDR3 and less TCR diversity, in comparison to data obtained from healthy or drug-tolerant donors. Similarly, the infiltration of predominant T cell clones has already been reported in many benign inflammatory skin diseases such as psoriasis, atopic dermatitis and contact dermatitis (32) (33) (and in MPE, as shown in our study (data not shown)). Here, novelty then resides in the demonstration that the strength of clonal expansions reached levels (both in blisters and blood) that have only been described in skin neoplasic disorders, such as cutaneous T cell lymphomas (CTCLs) (33). Additionally, the fact that our results can be generalized to patients expressing highly diverse HLA genotypes and reactive to very different drugs (Table 1), thus reinforces the idea that a massive clonal bias is a major immunological hallmark of TEN disease. Of note, as expected, we failed to detect any shared TCR sequences in our HLA diverse cohort, except for patients TEN-6 and TEN-10, exposed respectively to Norfloxacin and Ciprofloxacin quinolones, and who both expressed HLA-B*73:03 (unfortunately, due to low sampling, it was not possible to compare TCR sequences from TEN-1 & TEN-3 patients, harbouring the HLA-B*58 risk allele and exposed to allopurinol).
[0283] It will then be crucial to determine in the future the reasons for such clonal expansion in TEN disease compared to less severe MPE. (i) The massive production of inflammatory mediators noticed in the sera and the blister fluid of TEN patients (14) (34), or the reported defective Treg functions (34), certainly participates to enlarge the proliferation of drug-specific cells, but whether it is a consequence, a cause or both remains to be clarified. (ii) T regulatory cells (Treg) are critical regulators of CTLs causing TEN in mouse models (35). In this context, the reported defective functions of TEN circulating Tregs as well as their decreased ability to infiltrate the skin (36) (37), may explain the uncontrolled expansion and skin migration of drug specific CTLs. Interestingly, our data showed a differing CD4/CD8 ratio between TEN (ratio=0.5) and MPE skin (ratio=2) with MPE having a ratio similar to healthy skin (
[0284] Immunophenotype of TCR Clonotypes in TEN Patients
[0285] Another important point of the present study is the extended characterization of the expanded clonotypes, which mostly comprise CD8+ T cells endowed with a polycytotoxic phenotype. We observed that the dominant skin TCRV+ CTLs mainly expressed the cluster C phenotype, which was assigned to T.sub.EM cells. As expected (40) (26), this subset expressed high levels of Granzyme A, Granzyme B and especially Granulysin markers, and it was the only subset (with cluster D, poorly represented in skin samples) to express the CD38 protein, which is classically associated with T cell activation and/or diapedesis (41). By contrast, it lacked the expression of the senescence marker CD57 (classically assigned to T.sub.EMRA subsets), indicating that the expanded CTL clones correspond to recently activated T cells.
[0286] By comparison, CD8+ T cells infiltrating the skin of MPE or healthy donors displayed a distinct functional phenotype, as shown both at the total population level (data not shown) and after multidimensional analysis (
[0287] Drug Specificity
[0288] A major finding of our study is the antigenic specificity of the highly expanded clones found in TEN patients. Indeed, we were able to demonstrate that some of our engineered transductants (produced from TEN-3 and TEN-7) responded to their putative culprit drugs in vitro. Interestingly, potential drug reactivity was also recorded with transductants, generated from the rearranged pairs of TCRbeta and TCRalpha genes detected in the unusual dominant clone found in patient TEN-9, who was exposed to multiple drugs (Table 1). Nevertheless, as no clear culprit drug was identified for this patient, it was not possible to validate the relevance of our findings (data not shown). In contrast, transductants generated from sequences identified in patients TEN-10 and TEN-15 failed to respond to the tested drugs (Ceftriaxone, Ciprofloxacin, Levofloxacin, Metronidazole; Table 2). Various reasons might explain these TEN-10 and TEN-15 results. The simplest hypothesis is that we did not transfect the appropriate pathogenic TCR sequences. Alternatively, in keeping with the results obtained with TEN-3 transductants, which confirmed that T cells from allopurinol allergic patients are reactive to its metabolite (oxypurinol), but not to the parent molecule (4), it is possible that our in vitro drug exposure conditions (during Skw3/EBV-transformed B cell cultures) did not generate enough metabolites or drug-induced epitopes necessary to activate the transductants (in particular for Ciprofloxaxin, Levofloxacin or Metronidazole). Similarly, we cannot exclude that a specific mode of drug-epitope presentation (using peculiar non-conventional HLA-presentation (42)) or the involvement of an altered peptide repertoire (12) (13), govern T cell expansion from patients TEN-10 or -15.
[0289] Correlation with Disease Severity
[0290] The identification of early biomarkers, which predict final severity, is a highly desirable goal to improve clinical management of TEN patients. Our data confirm and extend the recent study reported by Xiong et al., which compared TCR repertoire diversity in patients suffering from SJS or TEN and showed that TCR repertoire metrics correlate with disease severity (31). So far, it is still debated whether SJS is an early stage of TEN (SJS is a bullous cADRs characterized by <10% of skin detachment) or a different pathology (both at the etiological and mechanistic levels). Here, we enrolled patients with progressing but established TEN phenotype only (with 40-100% of skin detachment at the peak of disease; except for patient TEN-2 who displayed an SJS/TEN intermediate phenotype with 20% of skin detachment). Despite extensive clonal expansion in TEN blisters at disease onset, we failed to detect any correlation between blister TCR repertoire diversity (or the percentage of top skin clones, data not shown) and final skin severity (
[0291] In conclusion, our results demonstrate that the quantity and quality of skin-recruited CTLs conditions the clinical presentation of cADRs. Importantly, they open major opportunities for the development of new prognostic markers in TEN.
TABLE-US-00001 TABLE 1 Patient demographics, clinical features and HLA genotype (Part1) Demographics Clinical Characterisitics Patient ID Sex/Age Ethnicity Underlying diseases Comorbidities TEN-1 M/48 East Asian Hyperuricemia None TEN-2 M/39 European Urine tract infection None TEN-3 F/40 European Hyperuricemia None TEN-4 M/74 European Melanoma None TEN-5 M/32 North African Pneumocystis HIV+ prophylaxis TEN-6 F/83 European Urine tract infection Cardiac insufficiency TEN-7 M/50 European Gastritis Cirrhosis TEN-8 F/33 European Bipolar disease None TEN-9 F/34 African Chronic pain None American TEN-10 F/63 European Severe angina None TEN-11 M/58 European Infectious osteoarthritis Diabetes, renal insufficiency TEN-12 F/27 European Cirrhosis Autoimmune hepatitis TEN-13 F/75 European Post-surgery infection Bladder adenocarcinoma TEN-14 M/41 European Myeloma None TEN-15 F/69 European Lung infection Ischemic stroke, SLE TEN-16 F/69 European Lung Infection None TEN-17 H/50 European Infection None TEN-18 H/58 European Liver cancer HCV+ MPE-1 M/18 European ENT infection None MPE-2 M/61 European ENT infection None MPE-3 F/68 European Breast infection None MPE-4 F/78 European Myeloma None MPE-5 F/71 European Cardiac insufficiency None MPE-6 F/62 European Infectious osteoarthritis None MPE-7 M/61 North African Pulmonary infection None MPE-8 F/24 East Asian Chronic pain None MPE-9 F/94 European Urine tract infection None MPE-10 F/62 European Graft versus Host Disease Bone marrow transplant MPE-11 F/39 North African Hypertension SLE MPE-12 F/62 European Hypertension, Gout None MPE-13 H/52 North African Myeloma None MPE-14 F/67 European Dermatomyositis None (Part 2) Clinical Characterisitics Drug exposure % of skin % & date of before onset Date & nature of first SCORTEN (TEN)/ detachment maximal skin Culprit drug (days) symptoms Severity (MPE) at day 0 detachment Allopurinol 8 day-2/fever 3 2% 100% at day 2 Sulfamethoxazole/Trimethoprim 7 day-2/fever 1 6% 20% at day 5 Allopurinol 15 day-3/fever + skin rash 2 20% 80% at day 2 Vemurafenib 22 day-4/skin rash 5 30% 100% at day 1 Sulfamethoxazole/Trimethoprim 15 day-2/fever 3 10% 80% at day 2 Norfloxacin 8 day-2/fever/skin rash 3 20% 50% at day 5 Pantoprazole 10 day-1/fever + skin rash 3 20% 100% at day 2 Lamotrigine 12 day-3/fever + eye stinging 2 10% 40% at day 5 * 2 day-2/fever 3 10% 50% at day 3 Ceftriaxone, Ciprofloxacin 8 day-4/skin rash 2 15% 30% at day 3 Ceftriaxone 15 day-1/skin rash 4 10% 60% at day 2 Furosemide 21 day-3/fever + skin rash 3 40% 40% at day 3 Cefixime 4 day-1/fever + skin rash 4 30% 30% at day 2 Revlimid 15 day-1 + fever + skin rash 2 5% 25% at day 3 Levofloxacin, Metronidazole 5 day-2 + fever + skin rash 3 10% 50% at day 3 Pristinamycin 1 day 0/fever + skin rash 2 10% 38% at day 2 Azithromycin, paracetamol 5 day-2/fever + skin rash 4 20% 80% at day 5 Sorafenib 10 day-3/skin rash 5 5% 48% at day 7 Amoxicillin 2 day-1/skin rash mild na na Amoxicillin 3 day-1/skin rash mild na na Vancomycin 28 day-4/skin rash severe na na Bortezomid 5 day-4/skin rash severe na na Diltiazem 15 day-3/skin rash severe na na Vancomycin 2 day-1/skin rash severe na na Vancomycin 42 day-4/skin rash mild na na Ibuprofen 9 day-3/skin rash severe na na Clindamycin 3 day-3/skin rash mild na na Tazocillin, contrast material 2 day-1/skin rash mild na na Macrogol, Urapidil, Amlodipine 14 day-3/skin rash moderate na na Allopurinol, Fibrate 28 day-2/skin rash mild na na Revlimid, Bortezomid 15 day-2/skin rash severe na na Hydroxychloroquine 15 day-4/skin rash mild na na (Part 3) HLA genotype Treatment Locus A Locus B Systemic corticosteroid + A*02; A*33 B*38; B*58 G-CSF A*30; A*30 B*13; B*18 A*02; A*03 B*27; B*58 Maintenance of existing A*03; A*23 B*44; B*51 corticosteroid therapy + G-CSF Systemic corticosteroid + A*02; A*24 B*44; B*45 G-CSF G-CSF A*03; A* B*18; B*73: 01 G-CSF A*02; A*11 B*15; B*44 A*02; A*30 B*08; B*44 G-CSF A*02; A*02 B*15; B*53 A*01: 03; A*68 B*08; B*73: 01 G-CSF A*02; A*29 B*44; B*45 Maintenance of existing A*01; A* B*08; B*51 corticosteroid therapy + G-CSF G-CSF A*02; A* B*44; B*57 Systemic corticosteroid A*02; A*02 B*15; B*27 A*03; A*30 B*18; B*40 A*02; A*03 B*35; B*51 G-CSF A*02; A*03 B*07; B*51 G-CSF A*03; A*11 B*35; B*40 Topical corticosteroid A*01; A*02 B*40; B*51 Topical corticosteroid A*02; A* B*08; B*40 Topical corticosteroid A*24; A*25 B*15; B*18 Topical corticosteroid A*29; A*31 B*35; B*44 Topical corticosteroid A*02; A* B*51; B* Topical corticosteroid A*01; A*02 B*40; B*57 Topical corticosteroid A*02; A*32 B*49; B*51 Topical corticosteroid A*24; A* B*15; B*38 Topical corticosteroid A*23; A*31 B*39; B49 Topical corticosteroid A*02; A*03 B*15: 16; B*39 Topical corticosteroid A*32; A*34 B*39; B*44 Topical corticosteroid A*23; A*68 B*44; B53 Systemic corticosteroid A*01; A*02 B*07; B*51 Topical corticosteroid A*01; A*29 B*08; B*44
[0292] Alden's algorithm was used to determine culprit drugs for TEN patients. For MPE patients, the main putative drugs are also indicated.
[0293] Disease severity for TEN patients was evaluated by the SCORTEN at day 0 (arrival at hospital and diagnosis). The SCORTEN predicts the risk of death. The SCORTEN scale consists in 7 independent factors for high mortality, and varies from 0 or 1 (low mortality rate) to 5 or more (very high mortality rate). Disease severity was appreciated by calculating percentages of skin detachment (using E-Burn application). The peak of disease was appreciated as the date at which TEN patients displayed maximal percentage of skin detachment.
[0294] Disease severity for MPE patients was estimated based on the extent of skin rash and the presence of systemic and/or visceral symptoms. None of the MPE patients exhibit symptoms suggestive of Drug Reaction and Eosinophilia Systemic Symptoms (DRESS)/Drug-Induced Hypersensitivity Syndrome (DIHS) and the Kardaun score was <3 for all the patients.
[0295] M=Male. F=Female. ENT=Ear Nose Throat. SLE=Systemic lupus erythematosus. HIV+=Human Immunodeficiency virus positive. HCV+=Hepatitis C virus positive, na=not applicable.
[0296] * no culprit drug was identified for patient TEN-9, using ALDEN algorithm. The patient received ibuprofen, doxycyclin, sulfamethoxazole-trimethoprime, tetracyclin, isoniazid, rifampicin in the days before TEN onset.
TABLE-US-00002 TABLE 2 Drug-induced activation of TCR Skw3 transfectants % of CD69 expression in CD3+ transfectants Patient ID SKW3 transfectant ID Drug concentrations (g/ml)) No drug Concentration 1 Concentration 2 Concentration 3 TEN-3 C1 Allopurinol (62.5/250) 2.3 2.2 3.05 Oxypurinol (62.5/250) 2.3 22.9 39.6 Sulfamethoxazole (100/200) 2.3 1.3 1.4 TEN-7 C2 Pantoprazole (10/50) 31.7 40.1 47.4 TEN-10 C3 Ceftriaxone (50/100/200) 12.2 11.9 12.0 13.8 Ciprofloxacin (12.5/25/50) 10.9 10.1 11.8 10.6 TEN-15 C4 Levofloxacin (25/50/100) 6.0 5.6 5.4 4.4 Metronidazole (25/50/100) 6.3 5.7 5.4 6.0 Control-1 17D Abacavir (1/10/20) 1.4 93.2 88.9 93.1 Pantoprazole (12.5/25/50) 1.4 1.8 1.8 1.7 Control-2 AnWe A1 Allopurinol (62.5/250) 4.3 17.1 26.4 Oxypurinol (62.5/250) 4.3 5.0 5.75 Control-3 UNO H13 Ibuprofen (20/100/200) 5.2 4.6 10.5 12.4
[0297] Skw3 cell lines engineered for the expression of TCRs bearing V and V chains from top clones found in patients TEN-3,-7,-10 and -15 were stimulated in vitro with EBV-transformed B cells in presence of graded doses of different drugs, or left unpulsed. Table 2 depicts the percentage of CD69 expression in CD3+ transductants measured by FACS after 24 h stimulation. Results from control transductants generated from Abacavir- (17D), Allopurinol-(AnWeAl) or Sulfamethoxazole- (UNO H13) allergic donors (53) (51) (7) are also shown. Bold and underlined values indicate >2 or >1.5 CD69 expression fold increase versus unpulsed cultures.
[0298] Transductant ID are from Table S10.
[0299] Autologous EBV-transformed B cells were used for all the patients, except for patient TEN-7, for whom we did not have any autologous PBMCs available; hence we performed the same analysis with heterologous PBMCs from different healthy donors. Heterologous EBV-transformed B cells were also used to stimulate control transductants.
Example 2 Preclinical Assessment of Anti-CD38 Monoclonal Antibody
[0300] The treatment of Toxic epidermal necrolysis (TEN), a rare but life-threatening cutaneous adverse drug reaction, is characterized by a rapidly progressing epidermal necrosis (1-2). TEN is associated with an important mortality rate of approximately 25-40%, and nearly constant and invalidating sequelae (blindness, respiratory disturbance . . . ), which are responsible for profound loss of quality of life in surviving patients.
[0301] To date, there is no efficient curative treatment for TEN, just palliative cares to relieve symptoms. TEN etiopathogenesis involves the recruitment and the activation into the skin of drug-specific polycytotoxic CD8+ T cells (3-6).
[0302] We have previously demonstrated that these cells express high levels of the activation marker CD38 (Example 1).
[0303] Goal and Method of the Therapeutic Approach
[0304] Therapeutic injection of anti-CD38 monoclonal antibody (mAb) to deplete the drug-specific cytotoxic CD8+CD38+ T cells as soon as the patient arrives to the clinic, in order to prevent/limit skin detachment and fatal outcome or invalidating sequelaes.
[0305] To make the proof of concept of the efficacy of anti-CD38 mAbs in a new TEN preclinical mouse model (i.e. humanized NGS mice transferred with CD8+ T cells collected from TEN patients at acute phase, and reactivated by the infusion of culprit drug(s)). We used this new preclinical model to assess the ability of a marketed anti-CD38 mAb, daratumumab, in depleting pathogenic cytotoxic CD8+CD38+ T cells.
[0306] Result in a Model of Graft Versus Host Versus Disease (GVHD).
[0307] We have assessed the efficacy of a marketed anti-CD38 mAb (daratumumab) to deplete human CD38+CD8+ T cells in a model of graft versus host versus disease (GVHD).
[0308] To this end, NGS mice were reconstituted with 1010.sup.6 peripheral blood mononuclear cells (PBMCs) from a healthy donor, and treated by two-weekly injections of daratumumab (at 100 or 300 microg/mouse). Control group received PBS. Reconstitution is generally assessed by measuring the ratio of humanization (calculated by dividing the % of human blood CD45+ cells/the % of mouse blood CD45+ cells). A high ratio of humanization (>50-60%) classically correlates with the appearance of GVHD symptoms, approximately 1 month after transfer.
[0309] In preliminary experiments, we observed that daratumumab depleted CD38+ cells (
[0310] Nevertheless, interestingly, we recorded that daratumumab transiently inhibited the expansion of human cells (measured by calculating the ratio of humanization) at day 12 (that is 7 days after the initial daratumumab injection) (
[0311] Of note, higher daratumumab regimen (300 microg/mouse, 2 times a week) also failed to prevent GVDH development, but transiently inhibited T cell expansion (not shown).
[0312] Besides, FACS analysis demonstrated that expanding cells poorly expressed CD38 marker (approximately 5% of CD38+CD8+ T cells) at day 12 (
[0313] It is probable that daratumumab failed to hamper GVDH development because pathogenic cells poorly expressed CD38+ in this model.
[0314] Therefore, to make the proof of concept of the efficacy of daratumumab, it is important to design a more relevant model, using CD38+ T cells collected from TEN patients.
[0315] Result in TEN Preclinical Model (NGS Mice).
[0316] Aims: (i) Determine engraftment upon cell transfer (CD8+ T cells isolated from TEN patients at acute phase). (ii) Characterize the immune response (lymphoid organs and peripheral tissues (skin, liver)), after drug administration. (iii) Characterize the clinical reaction: organ inflammation and cytokine production in the sera) induced by CD8+ T cell reactivation.
[0317] Deliverables: (i) Expansion of patient's cells before drug administration (i.e. percentage humanization=percentage of human versus mouse CD45+ cells). (ii) Percentage of proliferating (Ki67+) and activated (CD38+; Granulysine+ or Granzyme B+) CD8+ T cells in lymphoid organs, liver or skin. (iii) Skin, liver or kidney histology after drug administration. (iv) Main inflammatory cytokines/mediators in the sera (IL-1, IL-6, IL-15, TNF-, IFNg, Granulysin, Granzyme B).
[0318] Proof of Concept for Anti-CD38 mAb Efficacy
[0319] Objective: 1To demonstrate that anti-CD38 mAb injections deplete CD38+CD8+ T cells. 2To demonstrate that anti-CD38 mAb injections prevent the development of the clinical reaction induced by the reactivation of drug-specific CD8+ T cells.
[0320] Deliverables: In the two groups of mice injected or not with the anti-CD38 mAbs=(i) Percentage of proliferating (Ki67+) and activated (CD38+; Granulysine+ or Granzyme B+) CD8+ T cells in lymphoid organs, liver or skin (ii) Skin, liver or kidney histology. (iii) Main inflammatory cytokines/mediators in the sera (IL-1, IL-6, IL-15, TNF-, IFNg, Granulysin, Granzyme B).
[0321] Model description: We generated a surrogate model of TEN disease in mouse, by reconstituting NSG animals with 1.10.sup.6 millions of peripheral blood mononuclear cells (PBMCs) collected from a TEN patient, 1 year after disease recovery. The patients' T cells were then reactivated with the offending drug delivered to animals by oral gavage. In this model, T cells progressively expanded in response to the xenogenic environment, as well as well to drug addition. Hence, as shown in
[0322] These data thus indicate that our surrogate model, in which NSG mice are reconstituted with PBMCs from TEN patients and then administered with the culprit drug, recapitulates some of the key immune parameters of the disease.
[0323] Drug evaluation: We then capitalized on this new preclinical model to make the proof of concept of the efficacy of daratumumab in depleting pathogenic cytotoxic CD8+CD38+ T cells.
[0324] The effects of daratumumab were evaluated according to two administration regimens: (i) in a preventive mode, i.e. daratumumab was injected by intraperitoneal route (i.p.) very early (from day 4) after PBMC transfer, when cytotoxic CD8+ T cells have not yet proliferated, and (ii) in a curative mode, i.e., daratamumab was injected lately, from day 29 after PBMC transfer, once cytotoxic CD8+ T cells have proliferated.
[0325] Daratumumab injections from day 4 efficiently and extensively prevented the formation of CD4+CD38+ and CD8+CD38+ T cells in the blood and the spleen of transferred NSG recipients throughout the protocol (
[0326] Finally, by injecting daratumumab in curative mode from day 29, we demonstrated that daratumumab acutely depleted the CD38+ cells that have already expanded in the model (
[0327] Collectively, our results thus demonstrate the efficacy of daratumumab to deplete clonally expanded pathogenic T cells in a surrogate model of TEN disease.
[0328] Importantly, those data open new avenues for a new proof of concept study in TEN patients. After demonstrating that a single injection of daratumumab is well tolerated, and that it does not generate any side-effects (e.g. cytokine release syndrome), we will search to prove that it depletes the clonally expanded drug-specific cytotoxic CD8+CD38+ T cells. Ultimate objectives will consist to determine whether it alters the course of the disease and prevents/limits skin detachment, fatal outcome and/or invalidating sequelaes in TEN patient.
TABLE-US-00003 TABLE3 Usefulnucleotideandaminoacidsequencesforpracticingtheinvention SEQID NO Nucleotideoraminoacidsequence 1(CD38 MANCEFSPVSGDKPCCRLSRRAQLCLGVSILVLILVVVLAVVVPRWRQ AA QWSGPGTTKRFPETVLARCVKYTEIHPEMRHVDCQSVWDAFKGAFIS sequence KHPCNITEEDYQPLMKLGTQTVPCNKILLWSRIKDLAHQFTQVQRDM human FTLEDTLLGYLADDLTWCGEFNTSKINYQSCPDWRKDCSNNPVSVFW isoform1) KTVSRRFAEAACDVVHVMLNGSRSKIFDKNSTFGSVEVHNLQPEKVQ TLEAWVIHGGREDSRDLCQDPTIKELESIISKRNIQFSCKNIYRPDKFLQ CVKNPEDSSCTSEI 2(CD38 gcagtttcagaacccagccagcctctctcttgctgcctagcctcctgccggcctcatcttcgcccagccaacccc nucleic gcctggagccctatggccaactgcgagttcagcccggtgtccggggacaaaccctgctgccggctctctagga acid gagcccaactctgtcttggcgtcagtatcctggtcctgatcctcgtcgtggtgctcgcggtggtcgtcccgaggt sequence ggcgccagcagtggagcggtccgggcaccaccaagcgctttcccgagaccgtcctggcgcgatgcgtcaag human tacactgaaattcatcctgagatgagacatgtagactgccaaagtgtatgggatgctttcaagggtgcatttatttc isoform1) aaaacatccttgcaacattactgaagaagactatcagccactaatgaagttgggaactcagaccgtaccttgcaa caagattcttctttggagcagaataaaagatctggcccatcagttcacacaggtccagcgggacatgttcaccct ggaggacacgctgctaggctaccttgctgatgacctcacatggtgtggtgaattcaacacttccaaaataaactat caatcttgcccagactggagaaaggactgcagcaacaaccctgtttcagtattctggaaaacggtttcccgcag gtttgcagaagctgcctgtgatgtggtccatgtgatgctcaatggatcccgcagtaaaatctttgacaaaaacagc acttttgggagtgtggaagtccataatttgcaaccagagaaggttcagacactagaggcctgggtgatacatggt ggaagagaagattccagagacttatgccaggatcccaccataaaagagctggaatcgattataagcaaaagga atattcaattttcctgcaagaatatctacagacctgacaagtttcttcagtgtgtgaaaaatcctgaggattcatcttg cacatctgagatctgagccagtcgctgtggttgttttagctccttgactccttgtggtttatgtcatcatacatgactc agcatacctgctggtgcagagctgaagattttggagggtcctccacaataaggtcaatgccagagacggaagc ctttttccccaaagtcttaaaataacttatatcatcagcatacctttattgtgatctatcaatagtcaagaaaaattattg tataagattagaatgaaaattgtatgttaagttacttcactttaattctcatgtgatccttttatgttatttatatattggtaa catcctttctattgaaaaatcaccacaccaaacctctcttattagaacaggcaagtgaagaaaagtgaatgctcaa gtttttcagaaagcattacatttccaaatgaatgaccttgttgcatgatgtatttttgtacccttcctacagatagtcaa accataaacttcatggtcatgggtcatgttggtgaaaattattctgtaggatataagctacccacgtacttggtgcttt accccaacccttccaacagtgctgtgaggttggtattatttcattttttagatgagaaaatgggagctcagagaggt tatatatttaagttggtgcaaaagtaattgcaagttttgccaccgaaaggaatggcaaaaccacaattatttttgaac caacctaataatttaccgtaagtcctacatttagtatcaagctagagactgaatttgaactcaactctgtccaactcc aaaattcatgtgctttttccttctaggcctttcataccaaactaatagtagtttatattctcttccaacaaatgcatattg gattaaattgactagaatggaatctggaatatagttcttctggatggctccaaaacacatgtttttcttcccccgtctt cctcctcctcttcatgctcagtgttttatatatgtagtatacagttaaaatatacttgttgctggtactggcagcttatatt ttctctcttttttcatggattaaccttgcttgagggctttaacaattgtattactttttcaaagaactaagctttagcttcat tgatttttttctatttaattgggttttgctcttctctttagcattggaaacatagaaatgctttctgatttctttgggtagattt acgtattcagcttcttgagatggaagtttagatcactgatccttcagcttgttttcttttttgtatacatagattttaggac gatatattttcccttgagttctgctttagctgcagctcttatgttttgatatgcctctctttattatccttcagttaaaaatat ctttcaattcattgttatataaaaatatgtgcctagtttttaacatctggagattttctagttttgaaaaaaacataagcc aggcatggtggctcacacctgtatccccagcactttgggaggccgagacgggaggatcgcctgagctcagga gtttttacaccagcctgggaataacagtgagacattatctccaaaaaaattacctgggtatggtgttgtgcacctgt agtcccagctactctggagactgaggtgggaggattgtttgagcttgggaggttgaggctgcagggagctgtg atcacaccactgcactctggcctgagtgacagattgagaccctgtctcaataaaagcaaaaataaagaaaataa accatatgtgttgaacaaaggattaataaattaatttgagactccttcagggaatgaccacaatttattgaaaatagc ctaaatgttggagtcaggcatttctggattcatattttgacatcatgctgtcatcttgaacaaaatgcctaacctttctg aacttcaacttccttgccactcaaataaggattacaaaacttaaaatgtggtaagtactaaagacgacagcaaaaa ttgagtccagcacagagcttcctaaataagcaagcactcaacagagttggttcctttcttcctcccctgcttgacaa tccagtttcccacaggagcctttgtagctgtagccaccatggtcagtccagggattcttcactagccccttctcccc tggcagacatccttgtgggagtttagtcttggctcgacatgaggatgggggtttgggaccagttctgagtgagaa tcagacttgccccaagttgccattagctccccctgcagaatgtcttcagaatcggggcccggtcagtctcctggg tgacctgctgttttcctcttaagatcctttccactttggttgctgctttcgggactcatcgagtccttgctcaacaggat accccttgaagtggctgcctgggccacatccccttccaaacaagaaatcaaaatattagaaatcaatttttgaaatt tcccctaggaagactcatttgagtgttcaagttcagagccagtggagaccttaggggagggtggtcacaaggat tttgcacagtgctttagagggtcccagggagccacagaggtggtgaggggctgggtgctcttttctccgtgcatg accttgtgtgtctatcttcattaccacaatgcctcatctctacctcctttccccctgtagttccaacgtgggtatctttg ccatctctggcccgaaggactttctgacctacatgtataaataccccctcacaatatatattacttttcctataagtga cttctctactggattactggttgctcatacacctcatattttactcgtaaatctactactccctgtctgcctactccattct catttgctgtagaaaattctcttaccatcccaactttcacccaccatcatgcttacccaaaggctgtgggaatgacc tgggccctaatgccccttttctaaattcctaaggctcaccattttcctattgtaatggttcttgaccttataatgtttgag gcaccttttcaaatatagtcctttgatttcagactgaatacttgaaaggacacacacacacatacgtaagtgcatat gactgcatacacccacacacacacacgtgcctgtatacagtcatatgatacatacacaaacacacgcacacaag cctgcatacatcatatgccaacagtggggatatgttctgagaaatgcatcattagatgattttgtcattgtgtgaaca tcatagagtgtacttacactaacctagatggtctaacctactacacacccaggctacatggtatcacctattcctcct aggctacaagcctgtacagcgtgtgtctgtactaaatgctgtgggcaattttaacctgatggtaaatgtttgtgtatc taaacatatctaaacatagaaaaggtacagtaaacatgcagtattataatcttatgagaccgtcatcatatatgtggt ccactgtttgggccatcattggctgaaaagtggttatgcgacacatgactgtatatatactttcctgttacaacaaca gtgtctctcaatccacagtaattgcagcatccagtaggtcttactttagccctgagtcaccatttgtgtcaacgtgttt agtgccatgtccacgtctctcatgtaactggcagagctatcaaatattttggcaaaacacattgtttctttggctttgc cttggtaactttctgtgccttttgtagctcttgtttggaagaagctcaacccatgtctgcacactgtgatacaagggg gacagcatcgacatcgacttacttcttggtgccttattcctccttagaacaattcctaaatctgtaacttaagtttctca ggaagattccatactgcacagaaaactgcttttgtgggtttttaaaaggcaagttgttatatgtgctggatagttttta agtatgacataaaaattgtataaagtaaaatattaaaatacacctagaatactgtataactttaagtcattttatcaac acattgctaatccagatattttcccgcagttttttttgaataacagagcaattaatttacttttactatgaagagtcatc attttagtatgtattttaagcaatccaccaagaactcagtaggcagctgagaggtgctgcccagagaagtggtgat tagcttggccttagctcacccacacaaagcacaacaggctttgaactattccctaacggggcatttattctttttttttt ttttttttgggagacggagtctcgctgtcgcccaggctagagtgcagtggcgcgatctcggctcactgcaggctc caccccctggggttcacgccattctcctgcctcagcctcccaagtagctgggactgcaggcgcccgccatctcg cccggctaattttttgtatttttagtagagacggggtttcaccgtgttagccaggatagggcatttattcttgaacttg attcagagaggcacacattaccattctctaatcagaatgcaagtagcgcaaggcggtggaaactatggaattcg gaggcaggtgatgcattgggcgagtttattaacatctgtgactctctagtttgaaatttatttgtaacagacaaaaat gaattaaacaaacaataaaagtataataaagaa
REFERENCES
[0329] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure. [0330] 1. A. P. Villani et al., Ncrolyse pidermique toxique: physiopathologie et avances thrapeutiques Rev Fr Allergol 57, 418-425 (2017). [0331] 2. T. A. Duong, L. Valeyrie-Allanore, P. Wolkenstein, O. Chosidow, Severe cutaneous adverse reactions to drugs. Lancet 390, 1996-2011 (2017). [0332] 3. C. B. Chen et al., An Updated Review of the Molecular Mechanisms in Drug Hypersensitivity. J Immunol Res 2018, 6431694 (2018). [0333] 4. W. H. Chung et al., Oxypurinol-Specific T Cells Possess Preferential TCR Clonotypes and Express Granulysin in Allopurinol-Induced Severe Cutaneous Adverse Reactions. J Invest Dermatol 135, 2237-2248 (2015). [0334] 5. A. Nassif et al., Toxic epidermal necrolysis: effector cells are drug-specific cytotoxic T cells. The Journal of allergy and clinical immunology 114, 1209-1215 (2004). [0335] 6. A. Nassif et al., Drug specific cytotoxic T-cells in the skin lesions of a patient with toxic epidermal necrolysis. J Invest Dermatol 118, 728-733 (2002). [0336] 7 J. Adam et al., Abacavir induced T cell reactivity from drug naive individuals shares features of allo-immune responses. PloS one 9, e95339 (2014). [0337] 8. W. Hoetzenecker et al., Toxic epidermal necrolysis. F1000Res 5, (2016). [0338] 9. W. J. Pichler, D. J. Naisbitt, B. K. Park, Immune pathomechanism of drug hypersensitivity reactions. The Journal of allergy and clinical immunology 127, S74-81 (2011). [0339] 10. M. Vocanson, D. J. Naisbitt, J. F. Nicolas, Current perspective of the etiopathogenesis of delayed-type, and T-cell-mediated drug-related skin diseases. The Journal of allergy and clinical immunology, (2020). [0340] 11. W. J. Pichler, O. Hausmann, Classification of Drug Hypersensitivity into Allergic, p-i, and Pseudo-Allergic Forms. International archives of allergy and immunology 171, 166-179 (2016). [0341] 12. P. T. Illing et al., Immune self-reactivity triggered by drug-modified HLA-peptide repertoire. Nature 486, 554-558 (2012). [0342] 13. D. A. Ostrov et al., Drug hypersensitivity caused by alteration of the MHC-presented self-peptide repertoire. Proceedings of the National Academy of Sciences of the United States of America 109, 9959-9964 (2012). [0343] 14. W. H. Chung et al., Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nature medicine 14, 1343-1350 (2008). [0344] 15 I. Viard-Leveugle et al., TNF-alpha and IFN-gamma are potential inducers of Fas-mediated keratinocyte apoptosis through activation of inducible nitric oxide synthase in toxic epidermal necrolysis. J Invest Dermatol 133, 489-498 (2013). [0345] 16. C. Schlapbach et al., NKp46+ cells express granulysin in multiple cutaneous adverse drug reactions. Allergy 66, 1469-1476 (2011). [0346] 17. E. de Araujo et al., Death ligand TRAIL, secreted by CD1a+ and CD14+ cells in blister fluids, is involved in killing keratinocytes in toxic epidermal necrolysis. Experimental dermatology 20, 107-112 (2011). [0347] 18. N. Saito et al., An annexin A1-FPR1 interaction contributes to necroptosis of keratinocytes in severe cutaneous adverse drug reactions. Science translational medicine 6, 245ra295 (2014). [0348] 19. M. Hertl et al., Predominance of epidermal CD8+ T lymphocytes in bullous cutaneous reactions caused by beta-lactam antibiotics. J Invest Dermatol 101, 794-799 (1993). [0349] 20. Y. Hari et al., T cell involvement in cutaneous drug eruptions. Clinical and experimental allergy: journal of the British Society for Allergy and Clinical Immunology 31, 1398-1408 (2001). [0350] 21. N. Yawalkar, W. J. Pichler, Pathogenesis of drug-induced exanthema. International archives of allergy and immunology 124, 336-338 (2001). [0351] 22. F. Skowron et al., Comparative histological analysis of drug-induced maculopapular exanthema and DRESS. Journal of the European Academy of Dermatology and Venereology: JEADV 30, 2085-2090 (2016). [0352] 23. S. Singh et al., Assessment of histopathological features of maculopapular viral exanthem and drug-induced exanthem. J Cutan Pathol 44, 1038-1048 (2017). [0353] 24. M. Mockenhaupt, Epidemiology of cutaneous adverse drug reactions. Allergol Select 1, 96-108 (2017). [0354] 25. Y. K. Jee, S. Kim, J. M. Lee, H. S. Park, S. H. Kim, CD8(+) T-cell activation by methazolamide causes methazolamide-induced Stevens-Johnson syndrome and toxic epidermal necrolysis. Clinical and experimental allergy: journal of the British Society for Allergy and Clinical Immunology 47, 972-974 (2017). [0355] 26. P. C. Kuechler et al., Cytotoxic mechanisms in different forms of T-cell-mediated drug allergies. Allergy 59, 613-622 (2004). [0356] 27. J. Dean et al., Annotation of pseudogenic gene segments by massively parallel sequencing of rearranged lymphocyte receptor loci. Genome Medicine 7, (2015). [0357] 28. H. G. Drexler, G. Gaedicke, J. Minowada, Isoenzyme studies in human leukemia-lymphoma cell lines1. Carboxylic esterase. Leuk Res 9, 209-229 (1985). [0358] 29. R. Y. Pan et al., Identification of drug-specific public TCR driving severe cutaneous adverse reactions. Nat Commun 10, 3569 (2019). [0359] 30. T. M. Ko et al., Shared and restricted T-cell receptor use is crucial for carbamazepine-induced Stevens-Johnson syndrome. The Journal of allergy and clinical immunology 128, 1266-1276 e1211 (2011). [0360] 31. H. Xiong et al., Comprehensive assessment of T cell receptor beta repertoire in Stevens-Johnson syndrome/toxic epidermal necrolysis patients using high-throughput sequencing. Mol Immunol 106, 170-177 (2019). [0361] 32. J. L. Harden, D. Hamm, N. Gulati, M. A. Lowes, J. G. Krueger, Deep Sequencing of the T-cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis. F1000Res 4, 460 (2015). [0362] 33. I. R. Kirsch et al., TCR sequencing facilitates diagnosis and identifies mature T cells as the cell of origin in CTCL. Science translational medicine 7, 308ra158 (2015). [0363] 34. S. C. Su et al., Interleukin-15 Is Associated with Severity and Mortality in Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis. J Invest Dermatol 137, 1065-1073 (2017). [0364] 35. H. Azukizawa, Animal models of toxic epidermal necrolysis. Journal of Dermatology 38, 255-260 (2011). [0365] 36. R. Takahashi et al., Defective Regulatory T Cells In Patients with Severe Drug Eruptions: Timing of the Dysfunction Is Associated with the Pathological Phenotype and Outcome. Journal of immunology 182, 8071-8079 (2009). [0366] 37. N. Yoshioka et al., Disturbed balance in three subpopulations of CD4(+)Foxp3(+) regulatory T cells in Stevens-Johnson syndrome and toxic epidermal necrolysis patients. Clinical immunology 148, 89-91 (2013). [0367] 38. M. K. Jenkins, H. H. Chu, J. B. McLachlan, J. J. Moon, On the composition of the preimmune repertoire of T cells specific for Peptide-major histocompatibility complex ligands. Annual review of immunology 28, 275-294 (2010). [0368] 39. C. A. Almeida et al., Virus-specific T-cell clonotypes might contribute to drug hypersensitivity reactions through heterologous immunity. The Journal of allergy and clinical immunology, (2019). [0369] 40. S. C. Su et al., HLA Alleles and CYP2C9*3 as Predictors of Phenytoin Hypersensitivity in East Asians. Clin Pharmacol Ther 105, 476-485 (2019). [0370] 41. S. Deaglio et al., Human CD38 (ADP-ribosyl cyclase) is a counter-receptor of CD31, an Ig superfamily member. Journal of immunology 160, 395-402 (1998). [0371] 42. R. J. Betts et al., Contact sensitizers trigger human CD1-autoreactive T-cell responses. Eur J Immunol 47, 1171-1180 (2017). [0372] 43. S. H. Kardaun et al., Variability in the clinical pattern of cutaneous side-effects of drugs with systemic symptoms: does a DRESS syndrome really exist? The British journal of dermatology 156, 609-611 (2007). [0373] 44. A. Auquier-Dunant et al., Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study. Arch Dermatol 138, 1019-1024 (2002). [0374] 45. B. Sassolas et al., ALDEN, an algorithm for assessment of drug causality in Stevens-Johnson Syndrome and toxic epidermal necrolysis: comparison with case-control analysis. Clin Pharmacol Ther 88, 60-68 (2010). [0375] 46. S. Bastuji-Garin et al., SCORTEN: a severity-of-illness score for toxic epidermal necrolysis. J Invest Dermatol 115, 149-153 (2000). [0376] 47. R. Finck et al., Normalization of mass cytometry data with bead standards. Cytometry A 83, 483-494 (2013). [0377] 48. S. Van Gassen et al., FlowSOM: Using self-organizing maps for visualization and interpretation of cytometry data. Cytometry A 87, 636-645 (2015). [0378] 49. R. V. Bruggner, B. Bodenmiller, D. L. Dill, R. J. Tibshirani, G. P. Nolan, Automated identification of stratifying signatures in cellular subpopulations. Proceedings of the National Academy of Sciences of the United States of America 111, E2770-2777 (2014). [0379] 50. H. G. Drexler, G. Gaedicke, J. Minowada, Isoenzyme Studies in Human Leukemia-Lymphoma Cell-Lines 0.1. Carboxylic Esterase. Leukemia Res 9, 209-& (1985). [0380] 51. D. Yerly et al., Design, expression, and processing of epitomized hepatitis C virus-encoded CTL epitopes. Journal of immunology 181, 6361-6370 (2008). [0381] 52. C. Segouffin-Cariou, G. Farjot, A. Sergeant, H. Gruffat, Characterization of the epstein-barr virus BRRF1 gene, located between early genes BZLF1 and BRLF1. J Gen Virol 81, 1791-1799 (2000). [0382] 53. D. Mauri-Hellweg et al., Activation of drug-specific CD4+ and CD8+ T cells in individuals allergic to sulfonamides, phenytoin, and carbamazepine. Journal of immunology 155, 462-472 (1995).