Implementing Pharmacogenomics across Europe. Prof. Henk-Jan Guchelaar Dept. of Clinical Pharmacy & Toxicology. Survey physicians and pharmacists

Transcription

1 Implementing Pharmacogenomics across Europe Prof. Henk-Jan Guchelaar Dept. of Clinical Pharmacy & Toxicology Leiden University Medical Center University of Leiden 47 th ESCP Symposium Belfast, 24th Oct 2018 Survey physicians and pharmacists 97.6% of physicians agreed that genetic variations may influence drug response (Stanek)G 99.7% of pharmacists agreed that a patients genetic profile may influence the response on a drug (Bank) Did you order or recommend a pharmacogenetic test in the recent 6 months (Guchelaar, Swen)? Yes 4% 15% ~400 GP s ~667 pharmacists 1

2 High expectations Bank, Pharmacogenomics 2017:18(3): Challenges Providing evidence for improvement in patient care Providing information on cost-effectiveness/consequences Providing data on diagnostic test criteria Selecting clinically relevant PGx tests Developing guidelines directing clinical use of PGx testing Improving acceptance by patients & health care professionals Swen, Plos Med 2007(4): e209 2

3 RCTs in Pharmacogenetics Drug Clinical Endpoint Variant Abacavir hypersensitivity HLA-B*5701 Acenocoumarol / Fenprocoumon Warfarin Warfarin % time between therapeutic INR % time between therapeutic INR % time between therapeutic INR Mercaptopurine leucopenia TPMT Warfarin major bleeding, INR>4, venous thromboembolism VKORC1/CYP2C9 VKORC1/CYP2C9 VKORC1/CYP2C9 VKORC1/CYP2C9/CYP4F2 TOPIC TRIAL 783 IBD patients; mercaptopurine or azathioprine 1:1 randomized to screening vs no screening TPMT*2, TPMT*3A, and TPMT*3C HET: 50% dose reduction, HOM 90% dose reduction Primairy endpoint: leuko s < 3.0*10(9)/L or platelets < 100*10(9)/L) 10-fold reduction in hematologic ADRs among variant carriers without differences in treatment efficacy Personalised medicine in cardiology Intervention Control RR (95%CI) Total (n) Hematological side effect 29 (7,2%) 29 (7,8%) Thiopurine response Optimization by Pharmacogenetic testing in Inflammatory bowel disease Clinics TMPT variant 1 / 39 (2,6%) 8 / 35 (22,9%) 0,11 (0,01-0,85) No TPMT variant 29 / 360 (8,1%) 22 / 335 (6,6%) 1,2 (0,72-2,09) Coenen MJ, Gastroenterology

4 TOPIC TRIAL 783 IBD patients; mercaptopurine or azathioprine 1:1 randomized to screening vs no screening TPMT*2, TPMT*3A, and TPMT*3C HET: 50% dose reduction, HOM 90% dose reduction Primairy endpoint: leuko s < 3.0*10(9)/L or platelets < 100*10(9)/L) 10-fold reduction in hematologic ADRs among variant carriers without differences in treatment efficacy Personalised medicine in cardiology Intervention Control RR (95%CI) Total (n) Hematological side effect 29 (7,2%) 29 (7,8%) Thiopurine response Optimization by Pharmacogenetic testing in Inflammatory bowel disease Clinics TMPT variant 1 / 39 (2,6%) 8 / 35 (22,9%) 0,11 (0,01-0,85) No TPMT variant 29 / 360 (8,1%) 22 / 335 (6,6%) 1,2 (0,72-2,09) Coenen MJ, Gastroenterology Number needed to genotype How many patients do I have to screen/test to prevent one from having a Adverse Drug Reaction (grade 3-4 toxicity, death, etc.)? 4

5 NNG for TPMT testing in Topic Hematological ADR: leuko s < 3.0*10(9)/L or platelets < 100*10(9)/L) NNG= 200 Risk: 7.4% versus 7.9% In TPMT variant carriers: NNT= 5 Risk: 2.6% versus 22.9% TOPIC TRIAL Thiopurine response Optimization by Pharmacogenetic testing in Inflammatory bowel disease Clinics 783 IBD patients; mercaptopurine or azathioprine 1:1 randomized to screening vs no screening TPMT*2, TPMT*3A, and TPMT*3C HET: 50% dose reduction, HOM 90% dose reduction Primairy endpoint: leuko s < 3.0*10(9)/L or platelets < 100*10(9)/L) 10-fold reduction in hematologic ADRs among variant carriers without differences in treatment efficacy Intervention Total (n) Control RR (95%CI) Hematological side effect 29 (7,2%) 29 (7,8%) TMPT variant No TPMT variant 1 / 39 (2,6%) 8 / 35 (22,9%) 0,11 (0,01-0,85) 29 / 360 (8,1%) 22 / 335 (6,6%) 1,2 (0,72-2,09) Coenen MJ, Gastroenterology Actionable genotypes 100% PGx-panel approach 80% 60% 40% 20% 0% 95% of patients have at least 1 actionable genotype Dunnenberger, Annu Rev Pharmacol Toxicol

6 Dutch Pharmacogenetics Working Group 12 members multidisciplinary (DPWG): (clinical) pharmacists, physicians, clinical pharmacologists, clinical chemists, epidemiologist, toxicologist, primary care physician Aim: To develop pharmacogenetic (dosing)guidelines based upon systematic review of literature To integrate these guidelines in electronic prescription systems and medication surveillance systems 2018: guidelines for 94 gene-drug pairs 47 actionable interactions 11 Swen, Clin Pharmacol Ther 2008;83(5):

7 Endorsement DPWG guidelines DPWG and CPIC CPIC and the DPWG guidelines are generally similar in terms of allele classification, genotype to phenotype translations and therapeutic recommendations for most gene-drug pairs. Clin Pharmacol Ther. 2017;103(4):

8 If genotype is known Clinical Implication Score For the 47 actionable drug-gene interactions Swen, Clin Pharmacol Ther 2018:103(5):

9 Examples scores category essential drug-gene Capecitabine-DPYD Fluorouracil-DPYD Tegafur-DPYD Irinotecan-UGT1A1 beneficial Codeine-CYP2D6 * Phenytoin-HLA-B ** potentially beneficial Tramadol-CYP2D6 Lansoprazol-CYP2C19 Omeprazol-CYP2C19 Pantoprazol-CYP2C19 * At doses of 4 dd 20 mg or higher, children >12 doses 4 dd 10 mg or higher; or Patients with additional risk factors such as CYP3A4 inhibitors or decreased Renal function ** In Asians, non-japanese DPWG May 2018 A patient. 21-year old woman recenty started with 20 mg tioguanine once daily; M. Crohn; 3 weeks after start: fatigue, headache, short of breath Lab: pancytopenia TPMT-genotyping: Heterozygous TPMT *3C/*2 Kies alternatief of verlaag de startdosering tot 6-7% van de normale dosering; evt. aanpassing op geleide van toxiciteit (monitoring van het bloedbeeld) en effectiviteit. De frequentie van monitoring dient te worden verhoogd. Adviseer de patiënt om contact op te nemen bij symptomen van beenmergdepressie (zoals erge keelpijn in combinatie met koorts, regelmatig een bloedneus en snel blauwe plekken) optreden. diagnostic - reactive Ekhart, Pharm Weekblad 2017: 35 9

10 Implementation study LUMC: IP3 Implementation of Pharmacogenetics in Primary care Project 200 patients included and pre-emptively genotyped Panel of genetic variants: CYP2C9; 2C19, 2D6, 3A5, DYPD, SLCO-1B1, TPMT and VKORC1; 40 alleles 40 pharmacies (Leiden) 200 patients included 89.5% 1 actionable genotype 61.5 % % 3 9.5% 4 2.0% % of patients therapeutic recommendation; dose adjustment or monitoring Implementation in Primary Care (IP3) Pharmacist alerts, orders the PGx test and provides physician with a personalized recommendation pre-emptive - pro-active & panel 10

11 Adherence PGx guidelines >85% of the recommendations accepted Follow-up data being collected: healthcare costs Drug (%) 14,5 7,5 8 amitriptyline atorvastatin citalopram escitalopram 8,5 1, nortriptyline simvastatin venlafaxine Impact Netherlands 2016 N = (Unique pat.) First Rx* ( ) Gene Phenotype Actionable# Dose- adj. /switch** PPI s CYP2C19 UM Coumarines VKORC1 TT Clopidogrel CYP2C19 PM + IM Statines SLCO-1B1 Lage act Thiopurines TPMT IM + PM Tramadol CYP2D6 IM + PM + UM Codeine CYP2D6 IM + PM + UM TCA s CYP2D6 IM + PM + UM Venlafaxine CYP2D6 IM + PM Flecainide CYP2D6 IM + PM + UM Paroxetine CYP2D6 IM + PM + UM Tamoxifen CYP2D6 IM + PM **based on prevalence from IP3 # based on DPWG guidelines 11

12 Inpatient care DPYD testing 5-Fluorouracil/capecitabine Colorectal cancer, head-neck cancer, breast cancer Grade 3 or higher toxicity: 15-30% Drug induced lethality: 0,5-1% 5FU inactive metabolites Dihydropyrimidine dehydrogenase (DPD) DPYD gene 12

13 Prospective *2A screening (n=2,038) 50% dose reduction in patients with DPYD*2A (literature) with DPYD*2A variant (heterozygous) with DPYD*2A variant (heterozygous) no DPYD*2A variant normal dose 50% dose reduction normal dose 73% 28% 23% Deenen, J Clin Oncol 2016: LUMC Oncologist considers DPYD testing standard of care Pharmacist alerts physician if FU/CAP is prescribed with no DPYD testing. pre-therapeutic - screening 13

14 DPWG uses gene activity score Activity DPYD variant 0 DPYD*2A (IVS14+1G>A;c G>A) DPYD*13 (c.1679t>g; I560S) 0,5 c.2846a>t (D949V) c.1236a>g/hapb3 (E412E) 1 DPYD*1 (wild-type) Dosing advice for: DPYD*2A DPYD*13 c.2846a>t c.1236g>a/hapb3 Gene activity score % of normal dose 0 no 5FU/capecitabine 0, , Henricks, 2015:Pharmacogenomics: DPYD LUMC Routine pre-therapeutic DPYD screening LUMC (per april 2013) Retrospective analysis: 314patients (18 maanden) Screening: mean: 87% final: % Lunenburg, Pharmacogenomics 2015;17(7):721 14

15 Implementation DYPD screening Pre-therapeutic screening was performed in 87% of patients, reaching % in the last 6 months of the project Acceptance of dose recommendation: 90% Chemoradiaton No grade 3-4 toxicity in patients with initial dose reduction Grade 3-4 toxicity was only seen in DPYD variant carriers without a dose reduction or who received a dose increase in subsequent cycles Dose titrations possible, guided by toxicity (not too fast) DPYD screening is feasible in clinical practice Lunenburg et al. Pharmacogenomics 2015;17(7):721 Nationwide - Alpe DPD study Funded by: Up front DPYD genotyping to reduce toxicity Alpe DPD study (NCT ) More SNPs (sufficient evidence for dose-adjustments for 4 SNPs) Study aim: To determine the safety, feasibility and cost-effectiveness of DPYD genotype and DPD phenotype-directed individualized dosing of fluoropyrimidines PI s: J. Schellens, HJ Guchelaar Lancet Oncology; 2018: 19 Oct Online Update guideline colorectal cancer Medical oncologist 2017 ( ) DPYD genotyping is highly recommended prior to fluoropyrimidine treatment. 15

16 Genetic counseling Pharmacogenetics clinic LUMC Clinical pharmacist & clinical geneticist PGx screened patients are offered genetic counseling Example counseling patient I am a CYP2D6 poor metabolizer For which drugs is this relevant? Is this relevant for certain food products? Is this relevant for my children? Can I take paracetamol safely? Should my parents be tested? Should I be re-tested in 5 or 10 years? 16

17 Overall aim U-PGx: Making actionable pharmacogenomic data and effective treatment optimization accessible to every European citizen 15 million, H2020, 10 EU countries Started 1 Jan 2016, 5 yr Reduction severe ADR: 30% U-PGx consortium H.J. Guchelaar (Coordinator), J.J. Swen, M. Kriek M. Pirmohamed, R. Turner J. Stingl M. Ingelman-Sundberg C. Mitropoulou M. van Rhenen, K.C. Cheung D. Steinberger V.H.M. Deneer M. Samwald G. Sunder-Plassmann A. Cambon-Thomsen M. Karlsson S. Jo nsson G. Toffoli E. Cecchin C.L. Davila Fajardo G. Patrinos V. Dolz an M. Schwab E. Schaeffeler 17

18 Project Outline N=8,100 Development of powerful and barrier-free CDSS 18

19 Provide suitable genotyping technology for the selected panel of variants PREPARE PREemptive Pharmacogenomic testing for preventing Adverse drug REactions Not the same patients SASG, UPAT, ULMF LUMC, MUMV, CROA, PHUL 19

20 Inclusion Criteria Patients of any ethnicity receiving a first prescription for at least 1 drug for which a DPWG guideline is available will be recognized ( triggered ) by a newly developed clinical decision support tool to the clinician and will be eligible for inclusion in the U-PGx project. All patients contribute at least 1 actionable gene-drug combination + prescription data indicate 1 and 2 additional PGx guided adjustments per patient for 50% and 30% of the patients respectively. Each of the study sites has own focus but can include all eligible patients. Data collection ecrf: ProMise Research Nurse Follow-up T=2, 4, 8, 12 weeks Baseline (± 1 week)* 4 weeks (± 1 week) * 12 weeks (± 1 week)* End of Study (± 4 weeks) *For every newly prescribed drug of interest LIM Online Survey Follow-up 1 st Rx PGx drug DNA sample Record in EMR PGx informed prescribing Safety code card provided 2 weeks* 8 weeks* *For every newly prescribed drug of interest 20

21 PGx: improving efficacy or preventing toxcity? Preventing toxicity Actionable interactions Improving efficacy CYP2C9 CYP2C19 CYP2D6 phenytoin; warfarin; acenocoumarol, phenprocoumon (es)citalopram; imipramine; sertraline; voriconazole amitriptyline; clomipramine; codeine (CI); doxepine; imipramine; nortriptyline, aripiprazole CYP2C19 CYP2D6 CYP3A5 CYP3A5 tacrolimus clopidogrel; voriconazole; lansoprazole, omeprazole, pantoprazole amitriptyline; clomipramine; codeine; doxepine; imipramine; nortriptyline; paroxetine, atomoxetine DPYD HLA-B SLCO1B1 TMPT VKORC1 UGT1A1 CYP2B6 capecitabine; fluorouracil; tegafur abacavir; carbamazepine, allopurinol, phenytoin, flucloxacillin simvastatin,atorvastatin azathioprine; mercaptopurine; thioguanine warfarin, acenocoumarol, phenprocoumon irinotecan efavirenz Bank, Clin Pharmacol Ther. 2017;103(4): Primary endpoint definition 21

22 Secondary Endpoints Secondary analyses of the U-PGx project are aimed at evaluating quantitative and qualitative indicators for successful implementation strategies. Cost-effectiveness (incl QOL using the EQ5D questionnaire) Composite endpoint for all prescriptions Survey of patients and physicians regarding: multi-domains (e.g. beliefs, knowledge, usability) process indicators for implementation (e.g. guideline compliance, extent of adoption, number of tests) among patients and physicians Secondary clinical outcome measures Patient reported outcomes (PROs) Routine drug levels Health-care consumption Model-based analysis of primary endpoint PREPARE: Current Status N=4,050 N=4,050 March 2017 Now Arm Number of patients enrolled PGx guided prescribing 2171 Standard of care 1542 TOTAL 3713 Real-time inclusion monitor 22

23 Implementing Pharmacogenomics in Europe: Design and Implementation Strategy of the Ubiquitous Pharmacogenomics Consortium. van der Wouden CH, Cambon-Thomsen A, Cecchin E, Cheung KC, Lucía Dávila- Fajardo C, Deneer VH, Dolz an V, Ingelman-Sundberg M, Jönsson S, Karlsson MO, Kriek M, Mitropoulou C, Patrinos GP, Pirmohamed M, Samwald M, Schaeffeler E, Schwab M, Steinberger D, Stingl J, Sunder-Plassmann G, Toffoli G, Turner RM, van Rhenen MH, Swen JJ, Guchelaar HJ. The Ubiquitous Pharmacogenomics Consortium. Clin Pharmacol Ther Dec 27. Repurposing of whole exome sequencing data WES is routinely used for diagnosis of monogenic diseases Over 2000 individuals with WES data at the LUMC >90% form child-parent trio Entire exome is sequenced Is it feasible to use existing WES data for pharmacogenetics? Van der Lee: poster-session 15 th Oct 5:45 pm 23

24 Results 166 out of 230 individuals gave consent and have data available 94% of individuals were child-parents trio s 5 variants, located outside of the genes/ lack of coverage: 1 for HLA-A* or HLA-B* for CYP2C19*17 1 for UGT1A1*28/*37 CYP2D6 ultra-rapid cannot be detected Take home messages Implementation of PGx in clinical practice is feasible and effective. The U-PGx project will provide both quantitative and qualitative evidence for a new model of PGx guided personalized medicine. We hope that U-PGx will provide health care professionals in the EU the necessary tools allowing them to truly embrace the concept of personalized medicine. 24

25 Thank you for your attention! U-PGx Kickoff Leiden Jan 19 th, This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No