Genetics and Therapeutic Response PDF

Genetics and therapeutic response Tony Marinaki Purine Research Laboratory Viapath Guy’s and St Thomas’ Hospitals  What will be covered An overview of pharmacogenetics Examples of pharmacogenetic associations which have not translated into...

Genetics and therapeutic response Tony Marinaki Purine Research Laboratory Viapath Guy’s and St Thomas’ Hospitals  What will be covered An overview of pharmacogenetics Examples of pharmacogenetic associations which have not translated into clinical practice – ibuprofen, warfarin, allopurinol What makes a good pharmacogenetic test? Clinically useful pharmacogenetic associations: thiopurine and fluoropyrimidine drugs  The cost of adverse drug reactions 18,820 patients admitted over six months and assessed for cause of admission, 1,225 (6.5%) admissions related to drug toxicity, The median bed stay was eight days, accounting for 4% of hospital bed capacity, projected annual cost to the NHS was £466m The overall fatality was 0.15%. Most reactions were avoidable. Drugs - aspirin, diuretics, warfarin, and non-steroidal anti- inflammatories The most common reaction - gastrointestinal bleeding. Pirmohamed, et al BMJ VOLUME 329 3 JULY 2004 What is pharmacogenetics ? Term first used in 1959 … individual variation in the handling of drugs in the body according to genetically determined factors. Butyrylcholinesterase (BCHE) – an early example of pharmacogenetics in clinical practice Succinylcholine muscle relaxants developed in the 1950’s and are still widely used for rapid short term skeletal muscle paralysis - eg intubation Some patients experience life-threatening prolonged apnea – due to genetic variation in the BCHE gene. Kalow et al (1956) developed phenotypic assays (the dibucaine number) to detect “Atypical” or the A variant, assay still used today. Subsequently, genetic basis of the Atypical (A) and Kalow (K) variants defined. More than 60 variants now described, mostly very rare and associated with the Silent (S) or deficient phenotype Phenotype and genotype are used to characterise patients, warning cards issued where BCHE activity is low, cascade testing is done in at risk family members Note: testing is done after toxicity has occurred  Aims of pharmacogenetic testing Benefit patient – reduced morbidity and mortality Optimise use of proven first line therapies Cost saving to the NHS by reducing adverse drug reactions and hence hospital admissions Why is pharmacogenetics important? Lifetime risk of cancer, UK (source CRUK)  Terminology Genetic terminology is used inconsistently – “allele” vs “genetic variant” vs “mutation” vs “polymorphism” vs “SNP”, … Nomenclature of genetic variants not consistent – A, K, J, S variants for butyrylcholinestrase, *1, *2 etc for CYPs and TPMT, cDNA and amino acid numbering for others All genetic variant are given a unique rsNUMBER  Pharmacogenetic variation Normal genetic polymorphism – only revealed when there is a drug challenge Coding region single nucleotide polymorphisms (SNPs) and indels which change an amino acid sequence. Variants affecting splicing – new exons created, exon skipping Repeat variation in regulatory elements. Gene duplications and deletions. May be multiple rare variants within a single gene Markers may be in disequillibrium with a causative variant Will a pharmacogenetic marker be clinically useful? Ibuprofen Martinez et al 2004  Ibuprofen – is a cause of serious adverse reactions Not cost effective to test entire population to personalise therapy CYP2C9*3 genotyping may identify subgroups of persons who potentially are at increased risk of gastroduodenal bleeding when treated with NSAIDs Further studies are needed before genotyping is introduced into clinical practice.. Focus on drugs with serious toxicity or a clear therapeutic benefit in optimising therapy Warfarin  Warfarin Anticoagulant used for the prevention of thrombosis Major side effect – life threatening bleeding Vitamin K epoxide reductase subunit 1 (VKORC1) polymorphisms explain 30% of the dose variation CYP2C9 poor metabolises explain 10% of the dose variation  Pharmacogenetics of warfarin Kimmel et al. N Engl J Med. 2013 Dec 12;369(24):2283-93. Genotype-guided dosing of warfarin did not improve anticoagulation control during the first 4 weeks of therapy Pirmohamed et al N Engl J Med. 2013 Dec 12;369(24):2294-303. Pharmacogenetic-based dosing was associated with a higher percentage of time in the therapeutic INR range than was standard dosing during the initiation of warfarin therapy. Allopurinol  Allopurinol Used for the treatment of gout. Considered a safe drug, tonnes prescribed every year Active metabolite is oxypurinol, a potent xanthine oxidase inhibitor preventing the formation of uric acid Toxicity HLA-B*5801 allele is associated with toxic epidermal necrolysis in Han Chinese populations (odds ratio 580, 95% confidence interval 34.4 to 9780.9; PT (Glu98Stop), 460G>A (Ala154Thr), 719A>G (Tyr240Cys) Point mutations: -91 A>G, -168 T>G TPMT*3A : 460G>A (Ala154Thr), 719A>G (Tyr240Cys) promoter 1 2 3 4 5 6 7 8 9 10 TPMT*14: 1 A>G (Met1Val) TPMT*3C: 719A>G (Tyr240Cys) TPMT*13: 83A>T (Glu28Val) TPMT*7: 681T>G (His227Gln) TPMT*17: 124C>G (Gln42Glu) TPMT*8: 644G>A (Arg215His) TPMT*5: 146T>C (Leu49Ser) TPMT*4: -1G>A (splicing defect) TPMT*18: 211G>A (Gly71Arg) TPMT*6: 539A>T (Tyr180Phe) TPMT*2: 238G>C (Ala80Pro) TPMT*15: -1G>A (splicing defect) TPMT*9: 356A>C (Lys119Thr) TPMT*19: 365A>C (Lys122Thr) TPMT*10: 430G>C (GLy144Arg) TPMT*3B: 460G>A (Ala154Thr) TPMT*11:395G>A (Cys132Tyr) TPMT*12: 374C>T (Ser125Leu) TPMT*16: 488G>A (Arg163His) Thiopurine methyltransferase activity in the population 700 600 500 Frequency (no. patients) 400 300 200 100 0 0 10 20 30 40 50 60 70 80 -100 TPMT 1:250-300 Avoid or very low High TPMT – 1:12 Use a dose: Usual dosing epiphenomenon? 50% dosing High risk of strategy strategy fatal toxicity  TPMT activity vs. AZA withdrawal in IBD : 189 patients London IBD Forum prospective study n=12/15=80% p= 0.002 % AZA n=63/174= 36% withdrawal Intermediate Normal  TPMT phenotype as a guide to dosing TPMT zero Use alternative therapy Carrier range: Initiate at 1/3rd standard dose, dose increment to a target dose of 50% standard dose Normal Range Initiate at standard dose Monitor FBC and LFT as per usual as majority of side effects are not explained by TPMT deficiency !  Case study 48 yr (M), Steroid-dependent CD. Azathioprine (2mg/kg). Severe neutropenia. Died 10 days after admission. Post-admission TPMT phenotyping = 0  Case study 27 year old female Crohn's diagnosed 2 years earlier. Treated with mesalazine, elemental diet. Repeated courses of steroids. AZA 2mg/kg started because of steroid dependency. Presented with sore throat, fever, nausea and ankle swelling. Neutropenia.  continued In view of low TPMT, the patient was persuaded to re-try AZA at a dose of 25 mg daily (0.5 mg/kg). This was well-tolerated. § AZA increased to 1mg/kg 25mg 8 25/25/50mg 7 6 5 Patient is well wbc Cells 4 neut x109 3 lymph 2 1 0 Pre 1 3 5 12 Weeks  Genotype vs phenotype - which is better? Main purpose of TPMT testing – to detect all completely deficient patients. Added value, detect carriers and dose adjust accordingly It is not currently cost effective to test for all possible TPMT variants – testing is done for the most common variants TPMT*3C/*3A/*2) Discordance in the TPMT carrier range – genotyping better Rare and private TPMT variants – phenotyping better at picking up all possible mutations Blood transfusions – will mask carrier and complete TPMT deficiency, genotyping better  Genotype vs phenotype - compromises Ideally all patients should be genotyped and phenotyped Phenotyping costs pennies, genotyping costs pounds Children and young adults with acute lymphoblastic leukaemia – poor concordance between genotype and phenotype, national guidelines say genotype Lennard L, Br J Clin Pharmacol. 2013 Aug 21. doi: 10.1111/bcp.12226. [Epub ahead of print] Blood transfusions, if disclosed – genotype For most patients - use therapeutic drug monitoring to further optimise dosing  Other loci HLA-DQA1-HLA-DRB1 variants confer susceptibility to pancreatitis induced by thiopurine immunosuppressants. Nat Genet. 2014 Oct;46(10):1131-4. Patients heterozygous at rs2647087 have a 9% risk of developing pancreatitis after administration of a thiopurine, whereas homozygotes have a 17% risk. A common missense variant in NUDT15 confers susceptibility to thiopurine-induced leukopenia. Nat Genet. 2014 Sep;46(9):1017-20 …a nonsynonymous SNP in NUDT15 (encoding p.Arg139Cys) that was strongly associated with thiopurine-induced early leukopenia (odds ratio (OR) = 35.6; P(combined) = 4.88 × 10(-94)). In Koreans, this variant demonstrated sensitivity and specificity of 89.4% and 93.2%, respectively, for thiopurine-induced early leukopenia (in comparison to 12.1% and 97.6% for TPMT variants). Multiple other loci with small effects sizes, some from pathway studies, others from whole genome studies 01/04/2021  LESSON OF THE WEEK Life threatening myelotoxicity secondary to azathioprine in a patient with atopic eczema and normal thiopurine methyltransferase activity Jamie S Wee, Anthony Marinaki, Catherine H Smith BMJ 2011;342:d1417 A 24 year old Filipino man with a lifelong history of generalised atopic eczema that was uncontrolled with superpotent topical corticosteroids. Started azathioprine (1 mg/kg) TPMT activity of 37 pmol/h/mg Hb (normal range (26-50 pmol/h/mg Hb). He was discharged after three weeks, with blood counts only returning to normal four months after admission. Genotyping: homozygous for the NUDT15 p.Arg139Cys variant  Individualising azathioprine therapy TPMT testing is accepted in clinical practice and guides initial dosing strategies NUDT15 variants are more important than TPMT in Asian ethnicities and should be included in genotyping panels There are other multiple pharmacogenetic markers with small effect sizes influencing drug metabolism TGN levels summarise genetic and epigenetic effects – further optimise azathioprine dosing strategies Fluoropyrimidines  5-Fluorouracil Fluoropyrimidine drugs developed in the 1950s. First line treatment for solid tumours including colorectal and breast cancer Approximately 2 million patients pa receive these antimetabolites worldwide 10-20% of patients will develop severe (Common Terminology Criteria for Adverse Events (CTCAE) grade ≥3) toxicity Leads to treatment interruption, delays of subsequent cycles and termination of treatment altogether. Impacts negatively on efficacy and prognosis as well as posing a significant cost burden to health care providers. Cost in bed days of 5FU grade 3-4 toxicity to Guy’s and St Thomas’ Hospitals % suffering Hospital stay % admitted to Intensive care toxicity (days) ITU stay (days) Diarrhea 10 5 1 2 Neutropenia 5 5 rare Hand foot syndrome Rare, except none none capecitabine (40-50%) Mucositis 5 5 1 2 Cardiac toxicity 1, probably 7 none under- recognised Nausea, vomiting 2 5 rare 2  Is there a case for pharmacogenetic testing? Dihydropyrimidine dehydrogenase deficiency (DPD) recognised as a cause of 5FU toxicity – first case of severe toxicity due to DPD deficiency reported in 1985 (Tuchman et al N Engl J Med 313(4), 245-249 (1985). The FDA warned in 2003 that 5FU and the pro-drug capecitabine are contra-indicated in patients with DPD deficiency Predicting and avoiding severe toxicity would benefit patients Fluoropyrimidine metabolism and DPD Dihydropyrimidine dehydrogenase (DPD) deficiency 5FU has a narrow therapeutic window with 80-90% of 5FU catabolised through DPD Complete DPD deficiency is a rare metabolic disorder characterised by increased thymine and uracil in urine – usually fatal toxicity if treated with 5FU. Partial (heterozygous) DPD deficiency - 4-6% of the population - is associated with severe Grade 3-4 toxicity to 5FU Partial or heterozygous DPD deficiency – thymine and uracil not elevated in urine, carriers are asymptomatic -  Why no enzyme DPD assay? DPD cannot be assayed in red cells White cell radiochemical assay not suitable for high throughput screening or referred samples Assay is not linear at low protein concentrations and is affected by the white cell differential count Genotype-phenotype correlation in the carrier range is very poor.  Genetics of DPD deficiency Complete DPD deficiency - many, mostly rare mutations scattered across 23 exons The ‘common’ DPYD IVS14+1G>A splice junction variant was the first “common” DPD deficiency to be associated with toxicity and has a heterozygous genotype frequency of 1.8%. Four DPYD variants are generally accepted as a panel to predict toxicity and have a combined genotype frequency of ~6%.  Replication, replication, replication! Variant Meta analysis Source DPYD*2A IVS14+1G>A, 1905+1G>A, Terrazzino S, et al OR=5.42; 95% CI: 2.79–10.52; Splice junction destroyed Pharmacogenomics. 2013 pT, p.D949V OR:8.18; 95% CI: 2.65–25.25; Pharmacogenomics. 2013 (rs67376798 pG, p.I560S RR: 4·01, 95%CI: 2·27–7·10, Lancet Oncology 2015 rs55886062 pG deep intronic variant Meulendijks et al RR: 1·59, 95%CI: 1·29–1·97, Splice junction created Lancet Oncology 2015 pG (p.I560S)  Pharmacogenetic directed dosing (CPIC) Variant % of standard dose IVS14+1G>A, c.1905+1G>A 50%, c.2846A>T (p.D99V) 50-75% c.1679T>G (p.A560S) 50% c.1236G>A/HapB3 c.1129-5923C>G deep 50-75% intronic variant Compound heterozygous genotypes: effect of variant alleles is additive  Changing clinical practice Service commissioned by DPYD Requests NHSE 18000 Viapath promotion 16000 Mail Online 14000 CPIC update 12000 10000 Deenen 8000 CPIC Meulendijks Terrazzino 6000 Loganayagam Loganayagam 4000 2000 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020  2020 changes to guidelines EMA recommends testing for four DPYD variants prior to therapy NHSE commissions testing for four DPYD variants in August UK Chemotherapy Board issues dosing guidelines for DPYD variants  DPD and 5FU toxicity summary Four DPYD variants predict toxicity and are present 6% of the population Patients with these variant genotypes are at high risk for early Grade 3-4 toxicity These DPYD variants explained ~25% of cases of grade 3-4 toxicity Testing saves lives, prevents prolonged hospital stays due to extreme toxicity and is cost effective.  Genomics and the future …. Personal genomic information (exomes, whole genome, chips) linked to prescription of therapy and guides dosing. As therapy changes in evolving conditions, the same genomic information is accessed to guide new therapies. Barriers: cost, tools for accessing genomic information and poor pharmacogenetic evidence for many drugs  Conclusions Pharmacogenetics is a ‘new’ diagnostic area with considerable potential for cost savings to the NHS. Patients should be tested prior to therapy to enable dose reduction/alternative therapy rather than seeking an explanation for side effects after these have occurred – this is starting to happen The main barrier to uptake of a service has been the need to change clinical practice. The end