Drug “allergy” is no allergy a view back, with some surprising conclusions

Werner J. Pichler, MD [email protected] [email protected]

Division Allergology, Inselspital, ADR-AC GmbH , Adverse Drug Reactions – CH 3010 Bern Analysis and Consulting Holligenstr 91, CH 3008 Bern, Switzerland

Drug allergy

allergy an immune response to an innocous

substance which leads to symptoms

an immune response to a drug or drug- metabolite which leads to clinical symptoms

activation of innate and adaptive immunity, immune response stimulation of specific T-cells and specific antibodies, inflammatory reaction

drug often < 1000 D; small molecules are per se not immunogenic, covalent binding to carrier protein (hapten-carrier complex) required for «immunogenicity»

Adverse Drug Reaction (type B): Drug Allergy

• anaphylaxis, urticaria • maculo-papular exanthema • bullous exanthema • acute generalized exanthematous pustulosis (AGEP) • Stevens-Johnson Syndrome (SJS) toxic-epidermal necrolysis (TEN) • DRESS, hepatitis , interstitial nephritis,pneumonitis • drug induced autoimmunity (SLE, pemphigus, ...)

BIZARRE CLINIC PURPOSE OF IMMUNE REACTION ? How are drugs stimulating the immune system ? Hapten/prohapten specific immunity Penicillin G SMX-NO Haptens are chemically reactive R N H O compounds able to bind H C N H CH S O S 3 H2 covalently to proteins. CH N 3 I O ON a Adduct formation gives a danger N O N N O signal to APC (CD86 A upregulation, IL-1b-secretion) T E

Adduct formation forms A neoantigenic determinants able D to induce both a T-cell and B-cell A P immune response. A hapten T I stimulates both, T- cells and B- processing cells !! V E Adverse drug reactions: Drug allergy is a type B reaction

Type A reaction Type B reaction Pharmacological action 1 Not a pharmacological action Predictable 2 Not predictable Dose dependent 3 Not dose dependent Rational 4 Not explainable, bizarre

MD Rawlins, JW Thompson: IR Edwards, JK Aronson: Pathogenesis of adverse drug reactions; Adverse drug reactions: definitions, Textbook of adverse drug reactions, diagnosis, and management Oxford (1977 & 1981), LANCET (2000) 356: 9237; 1255-1259 DM Davies (Ed.), Oxford Univ. Press, p. 10 & p11

Adverse drug reactions

Type A reaction Type B reaction Pharmacological action 1 Not pharmacological action Predictable 2 Not predictable Dose dependent 3 Not dose dependent Rational 4 Not explainable, bizarre

The consequence of this «not» classification (type B) are: 1. ADR Type B were considered to be due to the patient: «idiosyncrasy»: the drug is OK, the patient is weird 2. Type B reactions cannot be analysed in usual risk assessments: «fatalism» 3. Dose reduction not considered in avoidance of type B 4. The clinical picture was not explainable, it is «bizarre»

Type B reaction as an excuse to do nothing ??

Adverse drug reactions

Type A reaction Type B reaction Pharmacological action 1 Not pharmacological action Predictable 2 Not predictable AlainDose de Weck:dependent don`t build your scientific3 Not dose career dependent on research about drug hypersensitivity ! Rational 4 Not explainable, bizarre

It isT ahe neglected consequence field of, not this very «not» popular classification with money (type givers B) are (grants: ) no 1.animalADR models Type B, weremainly considered case records to be, seldom due to systematic the patient : investigations«idiosyncrasy, since »:the the topic drug is bizarreis OK, the – your patient research is weird may be considered2. Type bizarreB reactions ...... cannot be analysed in usual risk assessments: «fatalism» 3. Dose reduction not considered in avoidance of type B 4. The clinical picture was not explainable, it is «bizarre»

Adverse drug reactions

“As more becomes known about the mechanisms of Drug allergy specific adverse drug effects, 1 Not pharmacological action this classification will be revised 2 Not predictable further and classification of currently unclassifiable 3 Not dose dependent ? reactions will become easier” 4 Not explainable, bizarre Edwards & Aronson: LANCET, 356: 1255, 2000

Drug allergy: maculopapular drug eruption(MPE) 1984

45

40

35

30

25

20

15

10

5

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Lymphocyte Transformation Test

In PBMC: Stimulation 3H-Thymidine vitro by drug, Incorporation expansion cytotoxicity In vitro

PBMC: Stimulation 3H-Thymidine by drug, Incorporation expansion cytotoxicity

Lymphocyte transformation test (LTT): To stimulate T cells in vitro with drugs and to conclude to a drug allergy..... ?? Delirant isti helvetii LTT: stimulation of PBMC by drugs 1986

Konz. Autol. Plasma AB-Serum Medikament/Substanz microg SI SI

Phenytoin RS 1 16.2 18.7 10 55.1 81.7 50 93.3 72.9 100 63.5 63.8 Lamotrigin RS 0.1 1.4 1.5 1 1.6 1.9 10 2.6 4.4 100 2.9 19.6 Pantoprazol RS 0.1 1.2 1.1 1 2.0 1.0 10 1.1 1.8 50 0.2 0.3 Positive LTT for phenytoin and lamotrigine (DRESS) dose dependent, reaction in PBMC, no liver (metabolism) LTT: stimulation of PBMC by drugs 1986

Konz. Autol. Plasma AB-Serum Medikament/Substanz microg SI SI

Phenytoin RS 1 16.2 18.7 10 55.1 81.7 50 93.3 72.9 100 63.5 63.8 LamotriginA RS clear result should0.1 be repeated1.4 , and1.5, 1 1.6 1.9 if confirmed, should10 not and2.6 can not 4.4 be ignored100 2.9 19.6 Pantoprazol RS 0.1 1.2 1.1 Delirant isti1 helvetii2.0 1.0 10 1.1 1.8 50 0.2 0.3 Positive LTT for phenytoin and lamotrigine (DRESS) dose dependent, reaction in PBMC, no liver (metabolism) 1992-1999 Tony Wyss-Coray Christian Brander Daniela Mauri-Hellweg Benno Schnyder Karin Frutig-Schnyder Martin Zanni Salomé von Greyerz

Drug specific T-cell clones (TCC) Immunohisto- from blood and tissue chemistry cell culture with drug Expansion of a single cell T-cells Nikhil Yawalkar T-cells TCC phenotype function: cytokines, cytotoxicity proliferation interaction with APCs, tissue cells Perforin - CD4 drug recognition cross-reactivity red - brown Drug stimulated T cells are peculiar 2002

HLA-DR-allele Binding from the independent drug outside to the MHC- APC peptide complex recognition

peptide MHC-molecule

Additional specificity: MHC class II CD3-TCR drug complex restricted CD8+ high incidence of T cells alloreactivity T cell

CD4 and CD8 Oligoclonal TCR Vb T cells stimulated expansion

T cell activation How are drugs stimulating T cells ? starts within seconds of drug exposure.

Immediate Ca++ influx to oxypurinol (OXP) Hapten Proteasome is

formation is not required. not required. . The best control for your experiments is questioning it by a new PhD student (or by Dean Naisbitt): he/she will not believe what has been done previously. But when she/he is convinced, your data are solid !

2000 - 2006 Christoph Burkhardt Dean Naisbitt Nikhil Yawalkar Jan Depta Markus Britschgi Urs Steiner Andreas Beeler Yvonne Hari

Oliver Engler Petra Kuechler

Antigenicity of a drug DOGMA: “a small compound (<1000D) is not a complete antigen”

It is necessary that a drug binds covalently (as a “hapten”) to a carrier molecule to be able to stimulate the immune system ?

or

can a drug directly stimulate certain immune cells as well, without covalently binding to a carrier molecule ? (p-i-concept) Alain de Weck: be progressive in research but conservative in clinic 2002- 2004 «Pharmacological interaction with immune receptors» = p-i concept

1. Non-covalent binding of drugs to proteins functioning as immune receptors (TCR, HLA)

2 p-i explains an immune stimulation by a drug without postulating antigen-features of a drug ! 3 Landsteiner et al. were right: small molecules do not act as antigen, but they have other means of stimulating the immune system

TCR progress means also to change dogma: drugs can stimulate T cells without covalent binding (hapten characteristics)

pharmacological interaction with immune receptors

p-i concept (2002)

? the drug binds to HLA immune receptor proteins (HLA or TCR)?? (by non-covalent bonds, independent on processing, metabolism, immediately)

HLA-peptide-TCR complex Adverse drug reactions

Type B reaction Type A reaction immune mediated Pharmacological action 1 Not pharmacological action Predictable 2 Not predictable Dose dependent 3 Not dose dependent Rational 4 Bizarre (immune-mediated = complex diseases !!)

Some immune mediated reactions to drugs are «off target» (pharmacological) activities of the drug on immune receptors.

TCR transfectionsT11.1 )

-3 10 NRFX (20 g/ml) 2006-2010 8 6

4

2

0 H-thymidine incorporation (cpm x 10 x (cpm incorporation H-thymidine Monika Keller 3 0 250 500 1000 2500 5000 6250 10000 1250025000 50000 105 2.105 4.105 APC (cells/well) Zoi Spanou Interstitial nephritis Cytotoxic mechanism Marianne Lerch Simone Schmid Daphne Schmid Basil Gerber Anna Zawodniak IL-5 IFN-g Precursor Priska Lochmatter analysis

phenytoin phenytoin Progress from an unexpected site far away:

A U S T R A L I A

Lancet 2002: 359:727-737

T A marker for Stevens-Johnson Syndrome A Wen-Hung Chung*, Shuen-Iu Hung†, Hong-Shang Hong*, I Mo-Song Hsih‡, Li-Cheng Yang*, Hsin-Chun Ho*, Jer-Yuarn Wu†§, W Yuan-Tsong Chen†¶ Departments of *Dermatology and ‡Neurology, A Chang Gung Memorial Hospital, Taipei, †Institute of Biomedical Sciences, N Academia Sinica, Taipei, Taiwan NATURE|VOL 428 | 1 APRIL 2004 |www.nature.com/nature p-i concept (p-i HLA) TCR a drug fits into a particular HLA molecule the drug binds to an allele-typic region in the HLA by van der Waals forces; it does not bind to the peptide; the {HLA-peptide-drug} complex is then recognized by the TCR

HLA-B*5701: binding groove HLA for abacavir

Illing et al, Nature 2012 HLA-peptide-TCR complex Ostrov D et al, PNAS 2012 HLA polymorphism in population ca 14 HLA alleles / individual ca. 9900 HLA alleles in human population some alleles are frequent, some are rare Risk for Mister Meyer Carbamazepine HLA - A*02:01, A* 31:01 hypersensitivity - B*07:01; B*57:01 Abacavir - C* 02:01; C*06:01 hypersensitivity - DR* B1 01:01; 04:02 - DR* B5 01:01 - DP* 04:04; 08:01 - DQ* 01:05; 05:01

Mister Wang HLA - A*02:01, A* 32:01 Risk for - B* 15:02; B*58:01 Allopurinol - C* 02:01; C*06:01 hypersensitivity - DR* B1-01:01; 04:02 - DR* B5-01:01 Risk for - DP* 03:04; 07:01 Carbamazepine - DQ* 02:04; 03:02 hypersensitivity HLA polymorphism in population ca 14 HLA alleles / individual ca. 9900 HLA alleles in human population some alleles are frequent, some are rare Risk for Mister Meyer Carbamazepine HLA - A*02:01, A* 31:01 hypersensitivity Adverse side effects- B*07:01; to drugs B*57:01 belong to the most Abacavirpredictable - C* 02:01; C*06:01 side effects and personalized medicine is alreadyhypersensitivity used for - DR* B1 01:01; 04:02 abacavir (B*57:01) -and DR* carbamazepinB5 01:01 (B*15:02 USA, Taiwan and- DP* HLA 04:04;-A*31:01 08:01 in CH) - DQ* 01:05; 05:01

Mister Wang HLA - A*02:01, A* 32:01 Risk for - B* 15:02; B*58:01 Allopurinol - C* 02:01; C*06:01 hypersensitivity - DR* B1-01:01; 04:02 - DR* B5-01:01 Risk for - DP* 03:04; 07:01 Carbamazepine - DQ* 02:04; 03:02 hypersensitivity Adverse drug reactions

Type B reaction Type A reaction immune mediated Pharmacological action 1 Not pharmacological action Predictable 2 Not predictable Dose dependent 3 Not dose dependent Rational 4 Bizarre (immune-mediated = complex diseases !!)

Some immune mediated reactions are «off target» (pharmacological) activities of the drug on immune receptors. - In the time of personalized medicine they are, at least partly predictable and avoidable; - dose dependence? Role of drug concentration

Pharmacology Immunology initiation > < elicitation no effect naive > < effector humoral > < celluar immune response effect Con- cen- T cells and B cells react diferently toxic effect tra-

tion The IgE/Fc-IgE-RI system is geared to react to extremely small doses

Affinity maturation for B-cells Simple, one cell model

Complex, > 1 cell involved, naive>< memory, IgE, etc Type B reaction: dose independent ?

IgE mediated reactions can occur after very Dose Dose Therapeutic efficacy small concentrations, S for for far below the Y de- Anaphylaxis concentration M sensiti for therapy P zation T looks «dose- Induction For O independently» of T and B DRESS M Dose for cells anaphyl But even these reactions are dose dependent, as revealed by tolerance of femtomolar picomolar nanomolar molar even smaller concentrations concentration differ for induction and elicitation in frame of desensitizations Dose dependence for eliciting and causing allopurinol hypersensitivity a) In vivo: Stamp LK, et al, 2012 Starting dose is a risk factor for allopurinol hypersensitivity syndrome: a proposed safe starting dose of allopurinol. Arthritis Rheum. 2012; 64(8): 2529-36.

b) In vitro (Yun J, Clin Exp. All 2013):

HLA-B*58:01+ HLA-B*58:01- TCL HD HD

ALP/OXP 100μg/ml 7/7 1/7

ALP/OXP 10μg/ml 1/6 0/6

ALP/OXP 1μg/ml 0/5 0/6

ALP/OXP-specific TCL: high drug concentration work synergistically with B*58:01 to induce ALP/OXP-TCL 2010-2014 James Yun Barbara Daubner Stephen Watkins Jacqueline Adam Daniel Yerly Natascha Wuillemin Oliver Hausmann

Open questions 2010-2014

1. Are DH indeed dose independent ? - dose dependent √ 2. Is p-i more dangerous than hapten stimulations ? - Fluxacillin induced B*57:01 linked hepatitis > < exanthema (not HLA-allele linked) 3. Are drugs binding to TCR, and is it stimulatory ? - model of SMX stimulation 4. What is stimulating T cells in p-i stimulations ? - drug binding to HLA make it look like an alloallele

Drug binds covalently to the drug binds non covalently «antigen», the peptide to HLA-protein p i ≠ hapten !

Hapten = Exanthem

p-i= hepatitis

It is the F-pocket in B*57:01 What does it mean: p-i vs hapten?

B* 58:01... B* 57:01...

E H X E A APC FLUX P APC Flux-peptide N presentation via p-i presentation (hapten) A T T H I E T M B* 58:01... I A S

Amount of hapten-peptide Amount of presentation depends on presented is limited: amount of HLA-B*57:01: ca. 10 – 200/cell  50.000/cell !

pharmacological interaction with TCR immune receptors p-i concept: a) the drug binds to the TCR (by non-covalent bonds; not restricted to a HLA-allele) or

b) the drug binds to the

HLA HLA molecule (NOT to the presented peptide); the {HLA-drug + peptide complex} is recognized by the TCR

HLA-peptide-TCR complex

Sulfamethoxazole and 11 other sulfanilamides:

Comparing - In vitro stimulation to 12 SA

- two TCR

- Blocking SMX stimulation by other sulfanilamides

- docking

- dynamic modelling SMX is bound to TCRVb2 of TCR «H13», outside the HLA-peptide interaction site!

TCC 1.3 binding to CDR3 Va

TCC H13 binding to CDR2 Vb

St Watkins & WJ Pichler: Sulfamethoxazole Induces a Switch Mechanism in T-Cell Receptors Containing TCRVβ20-1, by Altering pHLA Recognition; PLOS One, 2013 Drug (SMX) binding to the TCR-Vb CDR 2 loop of TCC H13

Only SMX and 5 of 11 other sulfanilamides fit into the pocket formed by the CDR2 region (TCR H13) The same 6 SA were stimulatory

Stephan Watkins & Werner J. Pichler: Activating Interactions of Sulfanilamides with T Cell Receptors, Open J Immunology, 2013 : Allosteric effect of SMX binding to CDR2-Vb pocket of TCR H13

Stephan Watkins & Werner J. Pichler: Sulfamethoxazole Induces a Switch Mechanism in T cell Receptors Containing TCRVb-20-1 Altering pHLA Recognition, PLOS ONE, 2013 Visualizing the H13 Binding Process

SMX No SMX

MD simulations of TCR H13 and HLA-DR*10:01 with or without SMX binding

Humphrey W, Dalke A, Schulten K (1996) VMD: visual molecular dynamics. J Mol Graph 14: 33-38, 27-38. Gibbs Free Energy, ΔG

- Free energy change is the most straightforward of the parameters

- For H13 it was shown SMX caused a 7 fold increase in affinity, from -24 to -140 kcal/mol.

- This translates from a 2 μmol to a 0.79 μmol affinity.

* Normal TCR affinities are in the range of 5-1 μmol, however we know the H13 T cell only proliferates with SMX present. A rapid walk through 30 years of drug allergy research:

Drug allergy is no allergy How is a drug stimulating T-cells via «Drug allergy» is 1 Not pharmacological action p-i ? -2 a Not pharmacological predictable 3off Not-target dose dependent activity or/and -4 predictable Not explainable , bizarre - dose dependent why is the clinical - bizarre picture bizarre ? T cells activated by {drug + peptHLA} are activated similar to an alloantigen. Such a T cell activation is „abnormal“, „not educated“ = bizarre !

Tolerant Allo-specific T cell T cell

No exchange TCR TCR of peptide necessary! same peptide B*57:01 different B*57:01 + + peptide configuration abacavir

Abacavir (drug) turns a ‚self‘ HLA molecule into a foreign ‚allo‘ HLA molecule Adam J, Yerly D, PlosOne, 2014 simply speaking In certain forms of drug «allergy» the drug binding - modifies the self HLA allele and makes it to an allo-allele! - the T cell immune response to a allo-allele is strong and misbehaving (not thymus-educted) - it may cause - transiently mild symptoms (MPE) (as long as treatment lasts) - sometimes only local symptoms (high local drug concentration) - sometimes causing systemic, persisting & deletereous reactions (SJS/TEN, DRESS?) Drug allergy* is not an allergy it is a type of allo-transplantat reaction !

allotransplantation Relation of the immune system to drug hypersensitivity

neither transplantation nor the use of millions of novel chemicals was foreseen in the evolution of the immune system

both are man made diseases

both result in bizarre clinical pictures (no biological sense)

both illustrate the limitations of medicine when medicine interferes with natural evolution

Thanks to the ADR-AC and drug allergy research team

2013

Antonia Klara Dario Tatjana James Daniel Natascha Heidi Thanks to Christiane

Jule Florian Lisa The sunshine of my life 30 years of clinic and drug allergy research at Inselspital

In addition to previously mentioned collaborators Franziska Mitton ALLI Team (Arthur Helbling, Oliver Hausmann, Benno Schnyder, Michael Fricker, Ulrich Müller, the MPAs etc....) RIA Team (Peter Villiger, Michael Seitz, Christa Mermod etc..) Collaborators: Dean Naisbitt, Andreas Cerny, Stephan Krähenbühl Teachers and mentors: Hans-Hartmut Peter, Alain de Weck Supporters in industry: Cathrine Christiansen, Thomas Kawabata many patients, friends, fair reviewers, etc......