Genetic susceptibility in pharmacodynamic and pharmacokinetic pathways underlying drug-induced arrhythmia and sudden unexplained deaths

Forensic Science International: Genetics 42 (2019) 203–212

Contents lists available at ScienceDirect

Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsigen

Research paper

Genetic susceptibility in pharmacodynamic and pharmacokinetic pathways underlying drug-induced arrhythmia and sudden unexplained deaths

T

M. Martinez-Matillaa,b, , A. Blanco-Vereaa,b, M. Santorib, J. Ansede-Bermejob,c, E. Ramos-Luisa,b, R. Gila,b, AM. Bermejod, F. Lotufo-Netoe, MH. Hirataf,g, F. Brisighellih, M. Paramoi, A. Carracedob,c, M. Briona,b,c ⁎

a

Xenética Cardiovascular, Instituto de Investigación Sanitaria de Santiago de Compostela, Complexo Hospitalario Universitario de Santiago, Santiago de Compostela (A Coruña), Spain b Grupo de Medicina Xenómica, Universidade de Santiago de Compostela, Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER), Santiago de Compostela, Spain c Centro Nacional de Genotipado-CeGen-USC-PRB3-ISCIII, Santiago de Compostela, Spain d Instituto de Ciencias Forenses "Luis Concheiro" (INCIFOR), Universidade de Santiago de Compostela, Santiago de Compostela, Spain e Psiquiatry Institute - Faculty of Medicine at University of São Paulo, São Paulo, Brazil f Institute Dante Pazzanese of Cardiology, São Paulo, Brazil g School of Pharmaceutical Sciences, University of São Paulo, São Paulo, Brazil h Institute of Public Health, Section of Legal Medicine, Università Cattolica del Sacro Cuore, Rome, Italy i Servizo de Psiquiatría, Complexo Hospitalario Universitario de Santiago de Compostela (CHUS), Servizo Galego de Saúde (SERGAS), Santiago de Compostela, Spain

ARTICLE INFO

ABSTRACT

Keywords: SUD Drug-induced arrhythmia Genetic susceptibility Pharmacodynamics Pharmacokinetics

Drug-induced arrhythmia is an adverse drug reaction that can be potentially fatal since it is mostly related to drug-induced QT prolongation, a known risk factor for Torsade de Pointes and sudden cardiac death (SCD). Several risk factors have been described in association to these drug-induced events, such as preexistent cardiac disease and genetic variation. Our objective was to study the genetic susceptibility in pharmacodynamic and pharmacokinetic pathways underlying suspected drug-induced arrhythmias and sudden unexplained deaths in 32 patients. The genetic component in the pharmacodynamic pathway was studied by analysing 96 genes associated with higher risk of SCD through massive parallel sequencing. Pharmacokinetic-mediated genetic susceptibility was investigated by studying the genes encoding cytochrome P450 enzymes using mediumthroughput genotyping. Pharmacodynamic analysis showed three probably pathogenic variants and 45 variants of uncertain significance in 28 patients, several of them previously described in relation to mild or late onset cardiomyopathies. These results suggest that genetic variants in cardiomyopathy genes, in addition to those related with channelopathies, could be relevant to drug-induced cardiotoxicity and contribute to the arrhythmogenic phenotype. Pharmacokinetic analysis showed three patients that could have an altered metabolism of the drugs they received involving CYP2C19 and/or CYP2D6, probably contributing to the arrhythmogenic phenotype. The study of genetic variants in both pharmacodynamic and pharmacokinetic pathways may be a useful strategy to understand the multifactorial mechanism of drug-induced events in both clinical practice and forensic field. However, it is necessary to comprehensively study and evaluate the contribution of the genetic susceptibility to drug-induced cardiotoxicity.

1. Introduction Some beneficial drugs for the majority of population may be ineffective or produce adverse drug reactions (ADRs) in patients with similar clinical characteristics often in an unpredictable manner, even

using therapeutic dosages. The study of these ADR becomes especially relevant when they are related with sudden death. Drug-induced channelopathies are a rare ADR that can be potentially fatal since they are mostly related to Torsades de Pointes (TdP), a life-threatening polymorphic ventricular

⁎ Corresponding author at: Xenética Cardiovascular, Instituto de Investigación Sanitaria de Santiago de Compostela (IDIS), Complexo Hospitalario Universitario de Santiago. Edif. Consultas planta -2, Rúa Choupana s/n, 15706. Santiago de Compostela (A Coruña), Spain. E-mail address: [email protected] (M. Martinez-Matilla).

https://doi.org/10.1016/j.fsigen.2019.07.010 Received 23 February 2019; Received in revised form 14 July 2019; Accepted 14 July 2019 Available online 18 July 2019 1872-4973/ © 2019 Elsevier B.V. All rights reserved.

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

tachycardia which can lead to ventricular fibrillation and sudden cardiac death (SCD). For example, the drug-induced long QT Syndrome (diLQTS) is an acquired reversible pathologic prolongation of the QT interval following exposure to drugs. Prolongation of the QT interval is a well-known risk factor for ventricular arrhythmias and one of the leading causes of drug withdrawal from the market, even though the correlation of QT interval with the risk of TdP is not well established. [1,2] Likewise, a considerable number of drugs have been reported to increase the risk of ventricular arrhythmias and SCD by inducing the type-1 ECG and/or arrhythmias in Brugada Syndrome (BrS) patients, many of them previously asymptomatic. [3] A great variety of drugs have been reported as proarrhythmic and/ or QT prolonging agents, such as (paradoxically) antiarrhythmics and non-cardiac drugs like antibiotics, antihistamines, antipsychotics or antidepressants. [4–7] The websites https://crediblemeds.org/ [8] and http://www.brugadadrugs.org/ [3] compile an extensive list of these drugs regularly updated. These drug-induced channelopathies are thought to be a multifactorial mechanism. To date, several risk factors have been described to be related to acquired channelopathies in addition to drugs, such as bradycardia, electrolyte imbalance, pre-existent structural cardiac disease (e.g. dilated or hypertrophic cardiomyopathy), specific conditions (e.g. liver disease), gender, drug interactions or genetic variation. [9–12] Genetic factors that predispose drug-induced arrhythmias and SCD have been widely studied, but how they contribute to the manifestation of these ADRs is still unknown [13]. Previous studies in both pharmacodynamic and pharmacokinetic pathways suggest that genetic defects associated with congenital forms of channelopathies or metabolism of drugs may be involved in such drug-induced events. The study of the genetic susceptibility in pharmacodynamic pathways has been focused mainly on genetic variants related to congenital forms of cardiac channelopathies [14] [15–17] since drug-induced alteration of IKr current is the most frequent cause of acquired LQTS and TdP [18–20]. The genetic susceptibility to drug induced arrhythmia in pharmacokinetic pathways has been investigated mainly focused on genetic variants in genes encoding cytochrome P450 enzymes and some transporters because it is thought that intracellular drug concentrations can produce variability in Kv11.1 (also known as HERG) block [21–23]. Both congenital and acquired channelopathies are associated with higher risk of SCD, an even more tragic event when sudden death is the first manifestation of the disease. Unfortunately, a significant proportion of these sudden deaths remain as unexplained cases without a conclusive cause of death after a conventional autopsy, called sudden unexplained deaths (SUD). In the forensic field, the study of genetic factors in SUD cases through the performance of a molecular autopsy may suggest possible causes of death. It is therefore likely that both the molecular autopsy in SUDs and the genetic study in clinical practice would provide new insights into these drug-induced events. In this complex framework, our objective was to describe the genetic factors in both pharmacodynamic and pharmacokinetic pathways underlying drug-induced channelopathies or SUD in a cohort of 32 individuals.

induced channelopathy or sudden unexplained death were recruited at the participating centres in Brazil (n = 8), Italy (n = 3) and Spain (n = 21) (Supplementary table S1). For this study, the individuals with suspected drug-induced channelopathy were collected by examining medical records of patients referred by cardiologists from psychiatry and cardiology centres or departments. They were selected on the basis of the following criteria: having suffered TdP (n = 1), borderline QTc interval (n = 2), long QTc interval (n = 12) or Brugada Syndrome (n = 2) after the administration of a drug previously reported as proarrhythmic/QT prolonging drug or after the administration of a dose change of one of those drugs. Patients were included if they had prolonged QT interval corrected according to the Bazett’s formula. Borderline QTc interval was considered 440 ms for male or 450 ms for women, while prolonged QT was considered ≥450 ms for male or ≥470 ms for female. In SUD cases (n = 15), autopsy reports were investigated. Full medico-legal autopsy and toxicological analysis were performed. Individuals were selected if they had therapeutic levels of at least one proarrhythmic/QT prolonging drug in post-mortem peripheral blood. We excluded SUD cases related with coronary artery disease or drug abuse. Toxic levels were considered as those established in Clarke's analysis of drugs and poisons book [24] and http://busca-tox.com. The drugs were selected according to the CredibleMeds (last access May 2019) and BrugadaDrugs (last access May 2019) registers and published literature. 2.2. Pharmacodynamic analysis through massive parallel sequencing (MPS) For the study of putative pathogenic variants in pharmacodynamic pathways, a previously designed custom panel [25] was analysed including 78 genes related to primary arrhythmogenic syndromes and cardiomyopathies associated with higher risk of SCD (SCD panel v1&2). Since new genes were described related with higher risk of SCD, an updated release of the custom panel was designed targeting a total of 96 genes (SCDv3) [26]. Table 1 shows the list of analysed genes included in each panel. For the targeted resequencing, DNA templates were obtained from peripheral blood. Targeted genes were enriched using SureSelect Custom Target Enrichment System kit (Agilent Technologies Inc., Santa Clara, CA) and subsequently sequenced following a paired-end sequencing approach in 5500xl SOLiD™ System (panel v1&2) and Ion Proton™ System (panel v3) (ThermoFisher, Carlsbad, CA). For data analysis, the pipeline comprised: data pre-processing or clean-up (sequence reads mapping, duplicates marking and base quality scores recalibration), variant discovery, variant annotation and variant prioritisation. Data pre-processing and variant discovery and annotation were carried out as previously described in Brion et al., 2014. [27] Detected variants were prioritised according to their possible pathogenicity, following recommendations of the American College of Medical Genetics (ACMG) [28], by:

• Functional impact on the protein: frameshift and non-frameshift •

2. Methods This study was developed according to the recommendations of the Declaration of Helsinki. Ethical committees of participating institutions approved the study and all participants or their relatives signed an informed consent form.

•

2.1. Study population

•

A total of 32 polymedicated patients suspected of having drug204

indels, non-sense and non-synonymous changes in exonic or splicing regions were prioritised. Population data: minor allele frequency (MAF) of the variants was consulted in Exome Variant Server (http://evs.gs.washington.edu/ EVS/), ExAC, [29] gnomAD (http://gnomad.broadinstitute.org/) and 1000 Genomes. [30] Variants whose MAF was greater than published prevalence of disease were filtered out (1/500 for cardiomyopathies related genes; 1/2000 for arrhythmogenic diseases related genes). In-house database was consulted in order to exclude frequent polymorphic variants or false positive changes due to systematic sequencing errors. Computational data: in silico pathogenicity prediction was assessed with SIFT (http://sift.jcvi.org/), Polyphen2 (http://genetics.bwh.

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

Table 1 (continued)

Table 1 Genes screened in each panel. Gene

Transcript

SCD panel v1&2 (78 genes)

SCD panel v3 (96 genes)

Associated cardiac phenotype

ABCC9 ACTA1 ACTC1

NM_020297.3 NM_001100.3 NM_005159.4

X X X

X

ACTN2

NM_001103.3

X

X

AKAP9 ANK2 ANKRD1 BAG3 BRAF CACNA1C

NM_005751.4 NM_001148.4 NM_014391.2 NM_004281.3 NM_004333.4 NM_001129833.1

X X X X

X X X X X X

CACNA2D1 CACNB2 CALM1 CALM2 CALR3 CASQ2 CAV3 CRYAB CSRP3 CTF1 CTNNA3 DES

NM_000722.2 NM_000724.3 NM_006888.4 NM_001743.5 NM_145046.4 NM_001232.3 NM_033337.2 NM_001885.2 NM_003476.4 NM_001330.3 NM_013266.3 NM_001927.3

X X

DMD DSC2 DSG2 DSP EMD EYA4 FHL1

NM_004006.2 NM_024422.3 NM_001943.3 NM_004415.2 NM_000117.2 NM_004100.4 NM_001159702.2

X X X X

FKTN FLNC

NM_001079802.1 NM_001458.4

X

GATA4 GJA1 GJA5 GLA GPD1L HCN4 JPH2 JUP KCND3 KCNE1 KCNE1L KCNE2 KCNE3 KCNH2 KCNJ2

NM_002052 NM_000165.4 NM_005266.6 NM_000169.2 NM_015141.3 NM_005477.2 NM_020433.4 NM_002230.2 NM_004980.4 NM_000219.5 NM_012282.2 NM_172201.1 NM_005472.4 NM_000238.3 NM_000891.2

KCNJ5 KCNJ8 KCNQ1 KRAS LAMP2 LDB3

NM_000890.3 NM_004982.3 NM_000218.2 NM_033360.3 NM_013995.2 NM_007078.2

X X X X X

X X X X X X

LMNA

NM_170707.3

X

X

MYBPC3

NM_000256.3

X

X

MYH6

NM_002471.3

X

X

MYH7

NM_000257.2

X

X

MYL2

NM_000432.3

X

X

MYL3 MYLK2

NM_000258.2 NM_033118.3

X X

X X

DCM, BRS HCM, DCM RCM, NCCM, HCM, DCM RCM, NCCM, HCM, DCM LQTS LQTS, CPVT HCM, DCM RCM, DCM HCM LQTS, DCM, BRS, SQTS BRS, SQTS BRS, SQTS LQTS, CPVT LQTS HCM HCM, CPVT LQTS, HCM, DCM RCM, DCM HCM, DCM DCM ARVC RCM, HCM, DCM, ARVC NCCM, DCM DCM, ARVC DCM, ARVC DCM, ARVC DCM DCM RCM, NCCM, HCM DCM RCM, HCM, DCM, ARVC DCM Atrial fibrillation Atrial fibrillation RCM, HCM BRS NCCM, BRS HCM DCM, ARVC BRS LQTS BRS LQTS BRS LQTS, SQTS LQTS, SQTS, CPVT LQTS BRS LQTS, SQTS HCM HCM, DCM RCM, NCCM, HCM, DCM RCM, NCCM, DCM, ARVC RCM, NCCM, HCM, DCM NCCM, HCM, DCM RCM, NCCM, HCM, DCM RCM, NCCM, HCM RCM, HCM HCM

X

X X X X X

X X

X X X X X X X X X X X X

X

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

Gene

Transcript

SCD panel v1&2 (78 genes)

SCD panel v3 (96 genes)

Associated cardiac phenotype

MYOZ2 MYPN NEBL NEXN NKX2.5 NOS1AP PKP2

NM_016599.4 NM_032578.3 NM_006393.2 NM_144573.3 NM_004387.3 NM_014697.2 NM_004572.3

X

X X X X X

PLN

NM_002667.3

X

X

PRKAG2 PSEN1 PSEN2 PTPN11 RAF1 RANGRF RBM20 RYR2

NM_016203.3 NM_000021.3 NM_000447.2 NM_002834.3 NM_002880.3 NM_016492.4 NM_001134363.2 NM_001035.2

X X X

X

SCN10A SCN1B SCN2B SCN3B SCN4B SCN5A

NM_006514.3 NM_199037.3 NM_004588.4 NM_018400.3 NM_174934.3 NM_198056.2

SDHA SGCD SLC25A4 SLMAP SNTA1 SOS1 TAZ

NM_004168.2 NM_000337.5 NM_001151.3 NM_007159.3 NM_003098.2 NM_005633.3 NM_000116.3

TBX5 TCAP TGFB3 TMEM43 TMPO TNNC1 TNNI3 TNNT2

HCM RCM, HCM, DCM DCM HCM, DCM NCCM LQTS DCM, ARVC, BRS, CPVT NCCM, HCM, DCM, ARVC HCM DCM DCM HCM HCM BRS RCM, NCCM, MC NCCM, DCM, ARVC, CPVT BRS LQTS, BRS BRS BRS LQTS LQTS, SQTS, DCM, ARVC, BRS DCM HCM, DCM HCM BRS LQTS HCM RCM, NCCM, HCM, DCM DCM HCM, DCM ARVC ARVC DCM RCM, HCM, DCM RCM, HCM, DCM RCM, NCCM, HCM, DCM RCM, NCCM, HCM, DCM LQTS, CPVT BRS RCM, HCM, DCM, ARVC RCM, HCM, DCM HCM, DCM

X X X X

X X X X X X X X X X

X

X X X X X X X X X X X X

X

X X X X

NM_000192.3 NM_003673.3 NM_003239.2 NM_024334.2 NM_003276.2 NM_003280.2 NM_000363.4 NM_000364.3

X X X X X X X

X X X X X X X X

TPM1

NM_000366.5

X

X

TRDN TRPM4 TTN

NM_006073.3 NM_017636.3 NM_133437.3

X

X X X

TTR VCL

NM_000371.3 NM_014000.2

X

X X

X

LQTS: Long QT syndrome; BRS: Brugada syndrome; SQTS: Short QT syndrome; CPVT: Catecholaminergic polymorphic ventricular tachycardia; HCM: Hypertrophic cardiomyopathy; DCM: Dilated cardiomyopathy; RCM: Restrictive cardiomyopathy; NCCM: non-compaction cardiomyopathy; ARVC: Arrhythmogenic right ventricular cardiomyopathy; RCM: Restrictive cardiomyopathy.

•

harvard.edu/pph2/), Human Splicing Finder (http://www.umd.be/ HSF3/HSF.shtml), MaxEnt (http://genes.mit.edu/burgelab/ maxent/Xmaxentscan_scoreseq.html) and NetGene2 (http://www. cbs.dtu.dk/services/NetGene2/); conservation scores were also consulted. Published literature review.

Prioritised putative pathogenic variants resulting from massive parallel sequencing were also confirmed by Sanger sequencing.

205

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

3.2. Pharmacodynamic study

Table 2 List of polymorphisms analysed for the pharmacokinetic study. CYP3A5

CYP2C19

rs10264272 rs41279854 rs41303343 rs55965422 rs776746 rs12248560 rs28399504 rs3758581 rs41291556 rs4244285 rs4986893 rs55640102 rs56337013 rs72552267 rs72558186

CYP3A4

CYP2C9

CYP2D6 rs35599367 rs4646438 rs55785340 rs67666821 rs1057910 rs1799853 rs2256871 rs28371685 rs28371686 rs56165452 rs72558187 rs1304490498 rs72558190 rs7900194 rs9332130 rs9332131 rs9332239

For the study of putative pathogenic variants in pharmacodynamic pathways, the panel v1&2 was analysed in 22 patients while the new release, v3, was studied in 10 patients. A total of 48 variants were prioritised in 28 individuals (87.5%) and classified according to ACMG recommendations as: three probably pathogenic variants: p.(Asn4539*) in TTN and p.(V1125 M) in MYBPC3 in two patients with suspected drug-induced LQTS, and p.(D982 V) in KCNH2 in one SUD case; and 45 variants of uncertain significance (VUS) (Table 3; Fig. 1). Among the 48 variants, there were 43 missense, one nonsense, three splicing variants, and one insertion. When the Human Gene Mutation Database (HGMD) [33] was consulted, we found 14 previously described variants: seven as disease-causing mutation (DM) and seven with an unclear classification (represented as DM?). Thirteen variants were considered novel as they were not described in any of the databases consulted. Variants classified as benign or likely benign were filtered out in the prioritisation process. Supplementary Table S2 shows the 48 prioritised variants and their bioinformatic annotations relevant for their classification of pathogenicity. Among the 28 patients harbouring the 48 prioritised variants, 15 (48.4%) carried one prioritised variant and 13 (41.9%) were carriers of at least two variants. The remaining four patients (˜12.5%; APs_02, APs_13, APs_30 and APs_32) were negative for any prioritised variant. With regard to variant distribution among diseases, 13 variants (25.5%) were detected in genes related to congenital arrhythmogenic syndromes including one probably pathogenic variant, whereas a total of 35 (72.9%) were identified in genes associated with cardiomyopathies, including two variants classified as probably pathogenic. Finally, we explored common genetic variants regarded as modifiers of sudden cardiac death. [34] Three of these variants were present in our cohort. The polymorphism D85 N in KCNE1 gene, previously associated with both congenital and drug-induced long QT syndrome [35,36,16], was identified in one patient who also harboured a novel missense variant in PKP2 gene (P472R) classified as VUS. The variant K897 T in KCNH2, previously associated with aggravation of symptoms in LQT2 and an increase of QT interval in LQT1-G589D founder population, was present in a patient who carried a probably pathogenic variant in KCNH2. The variant H558R previously described as a genetic modifier in LQTS type 3 was not present in the only carrier of a variant in SCN5A gene (Table 4). [34]

rs1065852 rs1080985 rs28371706 rs28371725 rs35742686 rs3892097 rs5030655 rs5030656 rs5030862 rs201377835 rs5030865 rs5030867 rs72549346 rs72549347 rs72549349 rs72549352 rs72549353 rs72549354 rs72549356 rs774671100

2.3. Pharmacokinetic study through medium-throughput genotyping Pharmacokinetic assay was performed through medium-throughput SNP genotyping using MassARRAY System (Agena Bioscience Inc., San Diego, CA) with MALDI-TOF mass spectrometry analysis at the Spanish National Genotyping Centre (CeGen-Santiago de Compostela). This method allowed us to genotype the iPLEX PGx Pro Panel, a commercial pharmacogenetic panel covering 191 polymorphisms, insertion-deletions (INDELs) and copy number variations (CNV) in 36 pharmacogenetically relevant genes involved in the absorption, distribution, metabolism and excretion (ADME) of a wide range of xenobiotics. Genotyping assay was performed following the manufacturer’s recommendations; sample analysis and data acquisition were carried out on the MassARRAY Analyzer 4 (MA4). MassARRAY Typer 4.0 software generated the diplotype and CNV reports. The SNPator tool was used to manage, transform and analyse the results obtained by genotyping [31]. The genetic susceptibility in pharmacokinetic pathways was studied by analysing the polymorphisms in genes previously related to pharmacokinetics on diQT interval prolongation shown in Table 2. The haplotypes were assigned by analysing SNPs described in Table 2. The alleles were defined by the Pharmacogene Variation Consortium (PharmVar www.PharmVar.org) and the genotype to phenotype translation was consulted in PharmGKB and published literature. To establish the functional status of CYP2D6, the activity score of the diplotype was assigned by totaling the score of both alleles through the scoring system established by Gaedigk et al., 2008. [32] The terms to describe the functionality of alleles and the phenotype were determined according to CPIC (https://cpicpgx.org/resources/termstandardization/).

3.3. Pharmacokinetic study: haplotype characterisation and analysis Genetic variants in the genes encoding for the enzymes of cytochrome P450 (CYPs) can contribute to drug-induced channelopathies by altering the metabolism of drugs. These genetic defects in the phamacokinetics could modify drug concentrations (in plasma or tissue) and drug effects, and thus take part in the proarrhythmic action [4]. As a large number of QT prolonging drugs are metabolised by these CYPs, we analysed the genetic variants in the genes encoding the CYPs that metabolize the drugs on each subject in the moment of the arrhythmogenic event. For example, if an individual was on venlafaxine, CYP2D6 was investigated because poor metabolizers may have increased levels of venlafaxine as compared to CYP2D6 normal metabolizers. Therefore, the genotypes of CYP2C19, CYP2D6, CYP2C9, CYP3A4 and CYP3A5 were investigated (Supplementary Table S3). The genotype of CYP2C19 was investigated in 23 patients. The analysis found two individuals ultra-rapid metabolizers (UM), carrying two increased function alleles (*17/*17), four rapid metabolizers (RM) with one normal function allele and one increased function allele (*17/ *1B;C) and 12 normal metabolizers (NM) with two functional alleles (*1/*1 or *1/*1B;C or *1B;C/*1B;C). The analysis also found three intermediate metabolizers (IM): patient APs_13 carrying one normal function allele and one no function allele (*1B;C/*2), probably

3. Results 3.1. Patients Among the 32 patients, there were 20 males (62.5%) and 12 females (37.5%) aged from 7 to 60 years old, taking at least one proarrhythmic or QT prolonging drug. In particular, 12 patients were taking one of these drugs, while 20 patients were treated with two or more of them (Supplementary Table S1). Since pre-existent cardiomyopathy could be a risk factor of druginduced cardiotoxicity, it is important to point out that one patient (APs_09) had atrial enlargement and three patients (APs_07, APs10 and APs_12) had normal echocardiograms, but had a family history of dilated cardiomyopathy (DCM) (Supplementary Table S1). 206

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

Table 3 List of subjects with their arrythmogenic events and prioritised variants. Patient

Arrhythmogenic event

APs_01 APs_02 APs_03

SUD BrS SUD

APs_04 APs_05 APs_06

SUD SUD SUD

APs_07

LQTS

APs_08

Limit QTc

APs_09

Limit QTc

APs_10

LQTS

APs_11 APs_12 APs_13 APs_14

LQTS LQTS TdP | LQTS | CRA SUD

APs_15 APs_16

SUD SUD

APs_17 APs_18 APs_19

SUD SUD SUD

APs_20 APs_21 APs_22 APs_23 APs_24

SUD SUD SUD SUD CRA1 | CRA2| LQTS | TACHYCARDIA

APs_25 APs_26 APs_26 APs_27

LQTS LQTS LQTS | LQTS | Limit QTc

APs_28 APs_29

LQTS LQTS | Limit QTc

APs_30 APs_31 APs_32

LQTS BrS LQTS

Gen

Prioritized variant

AKAP9 – TMEM43 EYA4 GLA MYBPC3 MYBPC3 MYBPC3 MYLK2 MYL3 AKAP9 DMD SCN5A MYBPC3 SCN4B SCN3B AKAP9 MYH6 DMD RBM20 MYBPC3 PKP2 TTN – JPH2 CACNA1C SLMAP MYH6 MYH6 SCN10A FLNC FLNC CTNNA3 DSC2 KCNH2 DSC2 VCL LDB3 SDHA NEBL CACNA1C CACNA1C MYH6 ANK2 MYBPC3 DSG2 MYBPC3 PKP2 DMD – SDHA –

c.4825_4826delAGinsCA (p.Arg1609Gln) – c.164G > A (p.G55D) c.1087C > A (p.P363T) c.427G > A (p.A143T) c.1814A > G (p.D605G) c.1321G > A (p.E441K) c.836G > C (p.G279A) c.260C > T (p.A87V) c.133G > C (p.E45Q) c.10229C > A (p.S3410Y) c.11026C > A (p.P3676T) c.3532T > C (p.C1178R) c.713G > A (p.R238H) c.97G > T (p.V33L) c.585-4A > G c.9022G > A (p.E3008K) c.2486G > A (p.W829X) c.7988C > G (p.T2663R) c.3169C > T (p.R1057W) c.133G > A (p.G45R) c.1415C > G (p.P472R) c.13614dupT (p.Asn4539*) – c.5G > A (p.S2N) c.5996C > T (p.T1999I) c.2140C > T (p.R714W) c.1328G > A (p.R443H) c.292G > A (p.E98K) c.4552G > A (p.V1518I) c.4991C > T (p.T1664M) c.6802G > A (p.E2268K) c.614C > T (p.A205V) c.1A > G (p.M1V) c.2945A > T (p.D982V) c.1165T > A (p.W389R) c.1040C > A (p.P347Q) c.1075G > A (p.D359N) c.1919A > G (p.E640G) c.1227+5G > A c.6116G > A (p.S2039N) c.5155G > A (p.G1719S) c.831G > T (p.Q277H) c.10274A > G (p.E3425G) c.3373G > A (p.V1125M) c.283A > G (p.T95A) c.2179G > A (p.V727M) c.293T > C (p.V98A) c.5550A > C (p.K1850N) – c.C1979T (p.A660V) –

Variant classification VUS – VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS PROB. PATH – VUS VUS VUS VUS VUS VUS VUS VUS VUS VUS PROB. PATH VUS VUS VUS VUS VUS VUS VUS VUS VUS PROB. PATH VUS VUS VUS VUS – VUS –

SUD: Sudden unexplained death; TdP: Torsades de Pointes; LQTS: Long QT Syndrome ; BrS: Brugada Syndrome; CRA: Cardiorespiratory arrest; NA: not applicable; VUS: Variant of uncertain significance; PROB.PATH: Probably pathogenic.

affecting metabolism of escitalopram, and two patients (APs_03, APs_32) carrying one increased function allele and one no function allele (*17/*2). The predicted IM phenotype for the *2/*17 diplotype is a provisional classification. The evidence so far indicates that the no function of CYP2C19*2 allele is not completely compensated by the CYP2C19*17 increased function allele. Whether the phenotype of patient APs_32 results in an altered systemic exposure of formoterol has not been sufficiently investigated; however, it has been suggested that CYP2C19 genotype has not a significant role in the valproic acid metabolism (APs_03). Furthermore, one patient was poor metabolizer (PM; APs_25) carrying two no function alleles (*2/*2), which could alter the metabolism of clomipramine (and maybe fluoxetine). Finally, the analysis found one patient (APs_24) with an uncharacterised allele (homozygous for rs12248560 and heterozygous for rs3758581, i.e. *17/unknown). The allele *9 was not considered in this study due to its

low frequency in the general population and because the genetic variant rs17884712, present in this haplotype, was not included in the panel used in the present study. Since CYP2D6 is the major contributor to the metabolism of a wide range of clinically used drugs, the predicted phenotype of this enzyme played an important role in the metabolism of the drugs taken by the majority of the patients of this cohort. Thus, the activity status of CYP2D6 was investigated in 30 patients. According to the scoring system of Gaedigk et al. for the predicted phenotype of CYP2D6, we found 26 patients who were NM, two who were IM (APs_10 and APs_23), one PM (APs_32) and one patient who was unclear if she was NM or PM (*2A;*31;*51/*4; APs_13). In APs_10, who was IM, CYP2D6 was studied in relation to the metabolism of quetiapine, but since this patient was wild type for CYP3A4, the major contributor to the metabolism of this drug, the role of the CYP2D6 may not be relevant. 207

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

Fig. 1. Summary of pharmacodynamic study results.

However, the patient APs_23, also IM for CYP2D6, could have a decreased metabolism of venlafaxine, as CYP2D6 isozyme plays the most important role in the metabolism of this drug. Finally, APs_32 was PM for CYP2D6, but whether this phenotype results in an elevated exposure to formoterol has not been widely studied. The genotype of CYP2C9 was analysed in eight patients: six patients had *1/*1 genotype, associated with normal metabolism, while the

remaining two patients had *1/*2 and *1/*3;*18 genotypes (APs_07 and APs_24, respectively), both previously associated with increased toxicity and ADRs when exposed to some drugs such as phenytoin. In the individual APs_24, the genotype of this gene was investigated in relation to phenobarbital, however, it was unclear whether the effect of the genotype 1/*3;*18 caused such a clinically significant increase of plasma concentrations of phenobarbital to contribute to the

Table 4 Genetic variants considered modifiers of sudden cardiac death detected in our cohort. Patient

KCNH2_K897T ↑LQT2

KCNE1_D85N ↑QT ↑risk

SCN5A_H558R normal LQT3

KCNH2_D982V PP

TTN_N4539* PP

SCN5A_p.C1178R VUS

APs_01 APs_02 APs_03 APs_04 APs_05 APs_06 APs_07 APs_08 APs_09 APs_10 APs_11 APs_12 APs_13 APs_14 APs_15 APs_16 APs_17 APs_18 APs_19 APs_20 APs_21 APs_22 APs_23 APs_24 APs_25 APs_26 APs_27 APs_28 APs_29 APs_30 APs_31 APs_32

– Het Het Het Het – Het – – – Het Het Het Het Het Hom – Het – Het Het – – Het Het – Het – Hom Hom – –

– – – – – – – – – – Het – – – – – – – – – – – – – – – – – – – – –

– – Het Het – Het Het – – – – Het – – Het – – – Het – Het – Het Het Het – – – – Het – Het

– – – – – – – – – – – – – – – – – – – Het – – – – – – – – – – – –

– – – – – – – – – – – Het – – – – – – – – – – – – – – – – – – – –

– – – – – – – Het – – – – – – – – – – – – – – – – – – – – – – – –

Het: Heterozygous; Hom: Homozygous; LQT: long QT syndrome; PP: probably pathogenic; VUS: variant of uncertain significance. 208

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

arrhythmogenic phenotype. In the patient APs_07, the CYP2C9 genotype was studied in relation to valproic acid but according to what was recently suggested by Smith et al (2016) [37], the genotype *1/*2 may not be relevant for the variability of valproic acid exposure, and consequently, may not play an important role in the arrhythmogenic event in this patient. CYP3A5 was assessed in six patients: five of them (APs_09, APs_10, APs_15, APs_26, APs_27) had a *3/*3 genotype and one (APs_24) had a *1/*3 genotype. The *3 allele has no function and it has been previously shown that it modifies the metabolism of some drugs such as risperidone, present in APs_09, APs_24, APs_26 and APs_27, or quetiapine, in APs_10, APs_15 and APs_26. However, these patients were NM for the major metabolizers of these drugs (CYP2D6 for risperidone and CYP3A4 for quetiapine). The CYP3A4 genotype was investigated in 24 patients and all of them were normal metabolizers (*1/*1). Finally, the patient APs_32 was homozygous for UGT2B17 gene deletion. This gene has been associated with glucuronidation of drugs such as steroids and formoterol. Formoterol, taken by this patient, is metabolised through direct glucuronidation by the glucuronosyltransferases (UGTs; UGT2B17, 1A9, 1A1, 1A7, 2B7, 1A3, 2B4) and through O-demethylation by CYP2D6, 2C19, 2C9 and 2A6. However, it is unknown whether the homozygous deletion of UGT2B17 inhibiting just one UGT isozyme and the *1B;C/*2 genotype resulting in a IM phenotype of CYP2C19 could be responsible for altered exposure of formoterol. On balance, three patients (APs_23, APs_13 and APs_25) could have an altered metabolism of the drugs they received, probably contributing to the arrhythmogenic phenotype. Furthermore, seven patients (APs_09, APs_10, APs_15, APs_24, APs_26, APs_27 and APs_32) might carry a risk genotype, although published evidence was so far not clear.

channelopathies together with drugs. In this cohort of 32 patients, 28 were carriers of rare variants in genes associated with inherited cardiomyopathies and channelopathies with higher risk of SCD. These prioritised variants were detected in 28 genes of the 96 genes of the panel v3 and in 25 genes of our panel v1&2. Only one variant was detected in SDHA gene that was not in panel v3, while five variants would have been lost with panel v1&2 (detected in SLMAP, SCN10A, FLNC and CTNNA3). Comparing our panel v3 to other molecular autopsy panels such as the Genetic Heart Disease (GHD) panel published last year by Tester et al comprising 90 genes [43], the majority of genes (74 in total) were in common, including major channelopathy and cardiomyopathy genes: KCNQ1, KCNH2, SCN5A, MYBPC3 and MYH7. The two panels differ mainly in minor genes associated with channelopathies or cardiomyopathies, and those associated with mitochondrial diseases and rasopathies. Among the 28 individuals carriers of prioritised variants, twelve of them carried 13 rare variants in genes regulating the cardiac ion channels and one polymorphism previously related to drug-induced LQTS (patient APs_11). These variants were in genes modulating the intracellular trafficking of K+ channels (AKAP9, KCNH2 and KCNE1) in four patients (including the polymorphism D85 N in KCNE1), genes regulating Na+ channels (SCN5A, SCN10A, SCN3B, SCN4B, SLMAP) in three patients, genes related with Ca2+ current (CACNA1C) in three patients and in ANK2 gene that has been shown to regulate voltagegated Na+ channels and ATP-gated K+ channels in one patient. The high proportion of rare variants in genes linked to cardiomyopathies is also remarkable. It is noteworthy that several of these variants have been previously described associated to mild or late-onset forms of cardiomyopathies, such as p.(A143 T) in GLA in the patient APs_04, p.(E441 K) in MYBPC3 and p.(A87 V) in MYLK2 in the patient APs_06 or p.(V1125 M) in MYBPC3 in the patient APs_27; also the variant p.(R238 H) in MYBPC3 (patient APs_08) has been suggested as a variant that may increase the severity of the cardiomyopathy. [44–48] Likewise, it is important to highlight the variant p.(Asn4539*) in TTN classified as probably pathogenic in the patient APs_12. Truncating variants in TTN gene are the most common genetic cause of idiopathic DCM and, although this patient didn’t have any known structural cardiomyopathy, her paternal uncle suffered DCM. We speculate that patients with genetic variants in cardiomyopathy genes may be vulnerable to drug induced arrhythmia even in the absence of the phenotype of the disease. Several authors have proposed molecular mechanisms for both cardiomyopathies and channelopathies involving the same genes, such as PKP2 or SCN5A [49]. For example, pathogenic mutations in cardiomyopathy genes could alter the contractile responses to Ca+2 and myofilament Ca+2 sensitivity triggering for arrhythmias and also modifying drug response. [50–53] Furthermore, several authors have reported QT alterations in HCM patients even in genetically affected individuals without hypertrophy [54–56]. However, the presence of these rare variants could be randomly in a carrier state in these individuals and may not be related with the arrhythmogenic event. Also with regard to SUD cases, when the only finding detected is the genetic variant in a cardiomyopathy gene in an individual with structurally normal heart, it does not mean that the individual developed the phenotype or that the variant was the cause of the death [57–60]. A total of 24 patients (75%) had at least one variant in genes related to cardiomyopathies, 7 of them also carrying variants in genes regulating cardiac ion channels. Among genes related with cardiomyopathies, 12 patients carried genetic variants in those encoding sarcomere proteins (LDB3, MYBPC3, MYH6, MYL3, TTN), four patients in the Z-disc proteins (VCL, NEBL and FLNC), five in desmosome proteins (DSC2, DSG2, PKP2) and 12 patients carried variants in genes encoding proteins with different locations such as junctional membrane complexes (JPH2), inner nuclear membrane (TMEM43), dystrophin-associated protein complex (DMD), etc. (Supplementary Table S2). Genetic defects in pharmacokinetic pathways have been also deeply studied in relation to drug-induced arrhythmias. In this study,

4. Discussion Drug induced channelopathies are thought to be a multifactorial event. Many approaches have been focused on mechanisms involving the depolarisation and repolarisation of cardiomyocytes and congenital forms of channelopathies. Mainly genetic defects altering the hERG channel, which regulates Ikr current, have been investigated. In addition, some studies have explored other subunits of potassium and sodium channels, as well as the key role of inward currents, like the late sodium current or the L-type calcium current. [38–40] Regarding these approaches, the pharmacodynamic component was investigated through massive parallel sequencing (MPS). In this cohort, only patient APs_20 was carrier of a novel genetic variant (D982 V) in KCNH2 gene, which encodes hERG channel. This patient was a 51-yearold man with schizophrenia and nephrogenic diabetes insipidus who suffered sudden collapse and SUD. Toxicological analysis revealed a concentration of levomepromazine of 0.05ug/ml and valproic acid of 4.1ug/ml in post-mortem blood, both therapeutic levels. The pharmacokinetic genetic analysis did not show any noteworthy finding. The KCNH2 gene is related to congenital LQTS, SQTS, BrS and also to the acquired form of LQTS. The variant D982 V, carried by this patient, has not been reported before, however another change (c.2944 G > A:p.D982 N) in the same codon was registered in HGMD as pathogenic associated with SUDs. [41] This patient also carried the common variant K897 T in KCNH2, a modifier of SCD associated with aggravation of symptoms in LQT2. The presence of both variants in KCNH2 might have had a subclinical effect in cardiac repolarisation until the drug (environmental factor) eventually triggered the arrhythmogenic phenotype. Roden and Viswanathan proposed in 2005 [42] the concept of “repolarisation reserve” to explain how the multiple redundant mechanisms of the myocardium to keep normal cardiac repolarisation can make up for defects in one of them. Therefore, several subclinical defects would be necessary to produce the arrhythmia, such as mutations or polymorphisms in genes related to congenital 209

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

pharmacokinetic approach was focused on genetic variants in CYPs that may alter the intracellular concentration of the drug and produce cardiotoxicity. In this cohort, there were some examples in which the pharmacokinetic study provided revealing results of an increased risk of arrhythmogenic phenotype. At least three patients could have an altered metabolism of the drugs they were taking involving CYP2C19 and/or CYP2D6:

pharmacokinetics of some antipsychotics such as clozapine, olanzapine or chlorpromazine, the effect of CYP1A2 genotype was not analysed in these patients. The strong influence of environmental factors such as smoking and caffeine and the still unknown effect of CYP1A2 genotype on antipsychotics metabolism make their interpretation complex and difficult [63,64]. 5. Conclusion

- Patient APs_13: a 44-year-old woman treated with escitalopram, olanzapine, and salbutamol (also called albuterol) suffered cardiac arrest and QT interval prolongation showing QTc of 500 ms. Two days later, she experienced Torsades de Pointes arrhythmia. At this time, the serum potassium was 3 mEq/L and serum magnesium was 1.3 mEq/L. After that, two electrocardiogram records showed QTc = 516 ms and QTc = 486 ms. The study of genes involved in pharmacokinetics showed an IM or PM CYP2C19 genotype, suggesting a reduced metabolism of escitalopram, and consequently, higher plasma concentrations of this drug increasing the probability of side effects such as QT interval prolongation or Torsades de Pointes. Furthermore, the co-administration of olanzapine with escitalopram in this patient with hypokalemia and hypomagnesaemia could increase the risk of an arrhythmogenic phenotype. The analysis of genes involved in pharmacodynamics did not reveal any putative pathogenic variant. In this case, risk factors as electrolyte imbalance, genetic alteration of CYP2C19 function, female gender and co-administration of olanzapine and escitalopram showed the multifactorial nature of this drug induced event. - Patient APs_23: a 49-year-old woman with coronary risk factors in her cardiac history, suffered SUD at home. Toxicological analysis revealed in postmortem blood 0.50ug/ml of venlafaxine. Pharmacokinetic analysis showed that this patient was IM for CYP2D6, presenting only one copy of the gene, and IM or NM for CYP2C19. The genotypes of both genes could contribute to the cardiovascular toxicity altering the metabolism of venlafaxine and its metabolites. [61,62] The analysis of genes involved in pharmacodynamic pathways detected one VUS (D359 N) in LDB3 gene, related with myofibrillar myopathy, DCM, HCM, heart block, atrial fibrillation and ventricular fibrillation. The risk factors for this patient may be female gender, coronary risk factors, venlafaxine, IM for CYP2D6 and the genetic VUS in LDB3. - Patient APs_25: a cumulative risk involving the co-administration of some drugs and genetic variants in both pharmacodynamic and pharmacokinetic pathways could lead to the arrhythmogenic phenotype. This patient was a 44-year-old man from Brazil with refractory obsessive-compulsive disorder who developed long QT interval when treated with clomipramine (75 mg/day), clonazepam (2 mg/day), fluoxetine (80 mg/day), lithium (1200 mg/day) and olanzapine (25 mg/day) showing QTc of 483 ms. Seven years later, the patient was taking clomipramine 75 mg/day, clonazepam 4 mg/ day, fluoxetine 80 mg/day and olanzapine 20 mg/day. When the dose of clomipramine was increased from 75 mg/day to 125 mg/ day, while the rest of the medication remained with the same daily dose, the patient referred tremors and anxiety and developed QTc of 473 ms. The pharmacokinetic analysis revealed that this patient was PM for CYP2C19 and, consequently, he could have a reduced metabolism of tertiary amines (clomipramine). Furthermore, fluoxetine is substrate and strong blocker of CYP2D6, an important enzyme for the metabolism of fluoxetine and clomipramine. The pharmacodynamic analysis detected one VUS (S2039 N) in CACNA1C. This variant could take part in the arrhythmogenic phenotype as CACNA1C gene encodes an alpha-1 subunit of a voltage-dependent L-type calcium channel, important for contraction, secretion, excitation and electrical signalling.

These results suggest that the study of genetic variants in pharmacodynamic and pharmacokinetic pathways may be a useful strategy to understand the multifactorial mechanism of drug-induced events in individuals with proarrhythmic/QT prolonging drugs. The study of pharmacodynamic genetic component through MPS including genes associated with arrhythmogenic syndromes and cardiomyopathies with higher risk of SCD allowed us to explore possible subclinical inherited cardiac conditions present in these patients. Our results propose that genetic variants in cardiomyopathy genes, in addition to those related with channelopathies, could be relevant to druginduced cardiotoxicity and contribute to the arrhythmogenic phenotype even before an overt structural disease. The results of the pharmacokinetic genetic analysis were consistent with the previously published key role of the cytochrome P450 enzymes in drug-induced events. The alteration of intracellular drug concentrations can increase the risk of ADRs. Furthermore, the concomitant intake of some medications was also an important risk factor in these patients as it could modify the kinetics of the drugs. Finally, the results of this study indicate that either genetic cardiac conditions, such as channelopathies or cardiomyopathies, or an altered drug metabolism may underlie these drug-induced events, sometimes with a possible cumulative risk. The molecular autopsy in SUDs and the genetic screening in clinical practice could be helpful tools to explore the causes and prevent new drug-induced events. Our results support that the molecular autopsy may be a useful tool in SUD cases to investigate a possible genetic cardiac condition underlying a fatal drug-induced event in the presence of proarrhythmic/ QT prolonging drugs (even in therapeutic concentrations) and provide a possible explanation of death. Drugs can unmask a genetic defect altering ion channels, even in therapeutic doses. In clinical practice, drug-induced arrhythmias may warn about the possible presence of inherited channelopathies, supporting the link between congenital and acquired forms of disease. Attending to our results, we speculate that genetic variants in cardiomyopathy genes, even in the absence of the phenotype of the disease, could make an individual more vulnerable to drug-induced arrhythmia. Also an altered drug metabolism possibly modifying drug concentrations in plasma or tissue could contribute to drug-induced cardiotoxicity. Therefore the screening of genetic variants in CYPs could help to identify at-risk patients and to choose a more effective and safer drug or therapeutic dose for the patient and to find a possible explanation of cause of death in SUDs. However, it is necessary to comprehensively study and evaluate the contribution of the genetic susceptibility to drug-induced cardiotoxicity. The present study has some limitations, such as the small sample size. It was very difficult to recruit individuals meeting the criteria of this project. However, we included 32 unrelated patients what allowed us to know even more about the genetic landscape of drug-induced cardiotoxicity, to better understand the multifactorial mechanism of these drug-induced events and to explore the presence of putative pathogenic variants in channelopathy and cardiomyopathy associated genes. We were also unable to study the variant cosegregation within families to elucidate the pathogenicity of some of the detected variants.

Although the CYP1A2 enzyme also plays an important role in the 210

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

Funding acknowledgements

[18] M.C. Sanguinetti, N.K. Jurkiewicz, Two components of cardiac delayed rectifier K+ current. Differential sensitivity to block by class III antiarrhythmic agents, J. Gen. Physiol. 96 (1) (1990) 195–215 7. [19] H. Numaguchi, J. Johnson, C.I. Petersen, J.R. Balser, A sensitive mechanism for cation modulation of potassium current, Nat. Neurosci. 3 (5) (2000) 429–430 1 5. [20] C.M. McBride, A.M. Smith, J.L. Smith, A.R. Reloj, E.J. Velasco, J. Powell, C.S. Elayi, D.C. Bartos, D.E. Burgess, B.P. Delisle, Mechanistic basis for type 2 long QT syndrome caused by KCNH2 mutations that disrupt conserved arginine residues in the voltage sensor, J. Membr. Biol. 246 (5) (2013) 355–364 5. [21] A. Puangpetch, N. Vanwong, N. Nuntamool, Y. Hongkaew, M. Chamnanphon, C. Sukasem, CYP2D6 polymorphisms and their influence on risperidone treatment, Pers. Med. 9 (2016) 131–147. [22] D.M. Roden, Pharmacogenetics of potassium channel blockers, Card. Electrophysiol. Clin. 8 (2) (2016) 385–393 6. [23] G.A. Ford, S.M. Wood, A.K. Daly, CYP2D6 and CYP2C19 genotypes of patients with terodiline cardiotoxicity identified through the yellow card system, Br. J. Clin. Pharmacol. 50 (1) (2000) 77–80 7. [24] T. Kelly, Clarke’s analysis of drugs and poisons. Fourth edition, Aust. J. Forensic Sci. 44 (2) (2012) 213–214 6. [25] M. Santori, A. Blanco-Verea, R. Gil, J. Cortis, K. Becker, P.M. Schneider, A. Carracedo, M. Brion, Broad-based molecular autopsy: a potential tool to investigate the involvement of subtle cardiac conditions in sudden unexpected death in infancy and early childhood, Arch. Dis. Child. 100 (10) (2015) 952–956 10. [26] S. Scheiper, E. Ramos-Luis, A. Blanco-Verea, C. Niess, B.-M. Beckmann, U. Schmidt, M. Kettner, C. Geisen, M.A. Verhoff, M. Brion, S. Kauferstein, Sudden unexpected death in the young — value of massive parallel sequencing in postmortem genetic analyses, Forensic Sci. Int. 293 (2018) 70–76 12. [27] M. Brion, A. Blanco-Verea, B. Sobrino, M. Santori, R. Gil, E. Ramos-Luis, M. Martinez, J. Amigo, A. Carracedo, Next generation sequencing challenges in the analysis of cardiac sudden death due to arrhythmogenic disorders, ELECTROPHORESIS 35 (21-22) (2014) 3111–3116. [28] S. Richards, N. Aziz, S. Bale, D. Bick, S. Das, J. Gastier-Foster, W.W. Grody, M. Hegde, E. Lyon, E. Spector, K. Voelkerding, H.L. Rehm, Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology, Genet. Med. 17 (5) (2015) 405–423. [29] M. Lek, et al., Analysis of protein-coding genetic variation in 60,706 humans, Nature 536 (7616) (2016) 285–291 18 8. [30] R.A. Gibbs, et al., A global reference for human genetic variation, Nature 526 (7571) (2015) 68–6874 1 10. [31] C. Morcillo-Suarez, J. Alegre, R. Sangros, E. Gazave, R. de Cid, R. Milne, J. Amigo, A. Ferrer-Admetlla, A. Moreno-Estrada, M. Gardner, F. Casals, A. Perez-Lezaun, D. Comas, E. Bosch, F. Calafell, J. Bertranpetit, A. Navarro, SNP analysis to results (SNPator): a web-based environment oriented to statistical genomics analyses upon SNP data, Bioinformatics 24 (14) (2008) 1643–1644 15 7. [32] A. Gaedigk, S. Simon, R. Pearce, L. Bradford, M. Kennedy, J. Leeder, The CYP2D6 activity score: translating genotype information into a qualitative measure of phenotype, Clin. Pharmacol. Ther. 83 (2) (2008) 234–242 31 2. [33] P.D. Stenson, M. Mort, E.V. Ball, K. Evans, M. Hayden, S. Heywood, M. Hussain, A.D. Phillips, D.N. Cooper, The Human Gene Mutation Database: towards a comprehensive repository of inherited mutation data for medical research, genetic diagnosis and next-generation sequencing studies, Hum. Genet. 136 (6) (2017) 665–677 27 6. [34] P.J. Schwartz, L. Crotti, A.L. George, Modifier genes for sudden cardiac death, Eur. Heart J. 39 (44) (2018) 3925–3931 11 9. [35] A.D.C. Paulussen, R.A.H.J. Gilissen, M. Armstrong, P.A. Doevendans, P. Verhasselt, H.J.M. Smeets, E. Schulze-Bahr, W. Haverkamp, G. Breithardt, N. Cohen, J. Aerssens, Genetic variations of KCNQ1, KCNH2, SCN5A, KCNE1, and KCNE2 in drug-induced long QT syndrome patients, J. Mol. Med. 82 (3) (2004) 182–188 1 3. [36] S. Kääb, D.C. Crawford, M.F. Sinner, E.R. Behr, P.J. Kannankeril, A.A.M. Wilde, C.R. Bezzina, E. Schulze-Bahr, P. Guicheney, N.H. Bishopric, R.J. Myerburg, J.J. Schott, A. Pfeufer, B.-M. Beckmann, E. Martens, T. Zhang, B. Stallmeyer, S. Zumhagen, I. Denjoy, A. Bardai, I.C. Van Gelder, Y. Jamshidi, C. Dalageorgou, V. Marshall, S. Jeffery, S. Shakir, A.J. Camm, G. Steinbeck, S. Perz, P. Lichtner, T. Meitinger, A. Peters, H.E. Wichmann, C. Ingram, Y. Bradford, S. Carter, K. Norris, M.D. Ritchie, A.L. George, D.M. Roden, A large candidate gene survey identifies the KCNE1 D85N polymorphism as a possible modulator of drug-induced torsades de pointes, Circ. Cardiovasc. Genet. 5 (1) (2012) 91–99. [37] R.L. Smith, T. Haslemo, H. Refsum, E. Molden, Impact of age, gender and CYP2C9/ 2C19 genotypes on dose-adjusted steady-state serum concentrations of valproic acid—a large-scale study based on naturalistic therapeutic drug monitoring data, Eur. J. Clin. Pharmacol. 72 (9) (2016) 1099–1104 29 9. [38] C.C. Veerman, A.O. Verkerk, M.T. Blom, C.A. Klemens, P.N. Langendijk, A.C. van Ginneken, R. Wilders, H.L. Tan, Slow delayed rectifier potassium current blockade contributes importantly to drug-induced long QT SyndromeClinical perspective, Circ. Arrhythm. Electrophysiol. 6 (5) (2013). [39] L. Johannesen, J. Vicente, J. Mason, C. Erato, C. Sanabria, K. Waite-Labott, M. Hong, J. Lin, P. Guo, A. Mutlib, J. Wang, W. Crumb, K. Blinova, D. Chan, J. Stohlman, J. Florian, M. Ugander, N. Stockbridge, D. Strauss, Late sodium current block for drug-induced long QT syndrome: results from a prospective clinical trial, Clin. Pharmacol. Ther. 99 (2) (2016) 214–223 2. [40] T. Yang, D.F. Meoli, J. Moslehi, D.M. Roden, Inhibition of the α-Subunit of phosphoinositide 3-Kinase in heart increases late sodium current and is arrhythmogenic, J. Pharmacol. Exp. Ther. 365 (3) (2018) 460–466 1 6. [41] D. Wang, K.R. Shah, S.Y. Um, L.S. Eng, B. Zhou, Y. Lin, A.A. Mitchell, L. Nicaj, M. Prinz, T.V. McDonald, B.A. Sampson, Y. Tang, Cardiac channelopathy testing in

This work was supported by the Plan Estatal de I+D+i 2008-2011 and 2013-2016 of the Subdirección General de Evaluación y Fomento de la Investigación (ISCIII-SGEFI) from Instituto de Salud Carlos III (ISCIII) and Fondo Europeo de Desarrollo Regional (FEDER) (grant numbers RD12/0042/0037, PI16/00903) and Brazilian CNPq (01097/ 2014-5). Declaration of Competing Interest None. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.fsigen.2019.07.010. References [1] B. Darpo, The thorough QT/QTc study 4 years after the implementation of the ICH E14 guidance, Br. J. Pharmacol. 159 (1) (2010) 49–57. [2] H. Food, Drug Administration, International conference on harmonisation; guidance on E14 clinical evaluation of QT/QTc interval prolongation and proarrhythmic potential for non-antiarrhythmic drugs; availability. Notice, Fed. Regist. 70 (202) (2005) pp. 61134-5, 20 10. [3] P.G. Postema, C. Wolpert, A.S. Amin, V. Probst, M. Borggrefe, D.M. Roden, S.G. Priori, H.L. Tan, M. Hiraoka, J. Brugada, A.A. Wilde, Drugs and Brugada syndrome patients: review of the literature, recommendations, and an up-to-date website (www.brugadadrugs.org), Heart Rhythm 6 (9) (2009) 1335–1341 9. [4] S. Sicouri, C. Antzelevitch, Mechanisms underlying the actions of antidepressant and antipsychotic drugs that cause sudden cardiac arrest, Arrhythm. Electrophysiol. Rev. 7 (3) (2018) 199–209 8. [5] G. Sarganas, E. Garbe, A. Klimpel, R.C. Hering, E. Bronder, W. Haverkamp, Epidemiology of symptomatic drug-induced long QT syndrome and torsade de pointes in Germany, Europace 16 (1) (2014) 101–108 1 1. [6] S. Sicouri, C. Antzelevitch, Sudden cardiac death secondary to antidepressant and antipsychotic drugs, Expert Opin. Drug Saf. 7 (2) (2008) 181–194 3. [7] E. Vandael, T. Marynissen, J. Reyntens, I. Spriet, J. Vandenberghe, R. Willems, V. Foulon, Frequency of use of QT-interval prolonging drugs in psychiatry in Belgium, Int. J. Clin. Pharm. 36 (4) (2014) 757–765 8 8. [8] W.D.a.R.K. R.L. Woosley, C.W. Heise, T. Gallo, J. Tate, “www.CredibleMeds.org,” [Online]. Available: https://crediblemeds.org/. [9] A.L. Coughtrie, E.R. Behr, D. Layton, V. Marshall, A.J. Camm, S.A.W. Shakir, Drugs and life-threatening ventricular arrhythmia risk: results from the DARE study cohort, BMJ Open 7 (10) (2017) e016627 16 10. [10] E.M. Kallergis, C.A. Goudis, E.N. Simantirakis, G.E. Kochiadakis, P.E. Vardas, Mechanisms, risk factors, and management of acquired long QT syndrome: a comprehensive review, Sci. World J. 2012 (2012) 1–8. [11] A. Pezhouman, N. Singh, Z. Song, M. Nivala, A. Eskandari, H. Cao, A. Bapat, C.Y. Ko, T.P. Nguyen, Z. Qu, H.S. Karagueuzian, J.N. Weiss, Molecular basis of hypokalemia-induced ventricular FibrillationCLINICAL PERSPECTIVE, Circulation 132 (16) (2015) 1528–1537 20 10. [12] S. Wyles, S. Hrstka, S. Reyes, A. Terzic, T. Olson, T. Nelson, Pharmacological modulation of calcium homeostasis in familial dilated cardiomyopathy: an In vitro analysis from an RBM20 patient-derived iPSC model, Clin. Transl. Sci. 9 (3) (2016) 158–167 6. [13] M.N. Niemeijer, M.E. van den Berg, M. Eijgelsheim, P.R. Rijnbeek, B.H. Stricker, Pharmacogenetics of drug-induced QT interval prolongation: an update, Drug Saf. (2015). [14] H. Itoh, L. Crotti, T. Aiba, C. Spazzolini, I. Denjoy, V. Fressart, K. Hayashi, T. Nakajima, S. Ohno, T. Makiyama, J. Wu, K. Hasegawa, E. Mastantuono, F. Dagradi, M. Pedrazzini, M. Yamagishi, M. Berthet, Y. Murakami, W. Shimizu, P. Guicheney, P.J. Schwartz, M. Horie, The genetics underlying acquired long QT syndrome: impact for genetic screening, Eur. Heart J. 37 (18) (2016) 1456–1464 7 5. [15] V.S. Williams, C.J. Cresswell, G. Ruspi, T. Yang, T.C. Atak, M. McLoughlin, C.D. Ingram, A.H. Ramirez, D. Roden, M. Armstrong, Multiplex ligation-dependent probe amplification copy number variant analysis in patients with acquired long QT syndrome, Europace 17 (2015) 635–641. [16] P. Weeke, J.D. Mosley, D. Hanna, J.T. Delaney, C. Shaffer, Q.S. Wells, S. Van Driest, J.H. Karnes, C. Ingram, Y. Guo, Y. Shyr, K. Norris, P.J. Kannankeril, A.H. Ramirez, J.D. Smith, E.R. Mardis, D. Nickerson, A.L. George Jr, D.M. Roden, Exome sequencing implicates an increased burden of rare potassium channel variants in the risk of drug-induced long QT interval syndrome, J. Am. Coll. Cardiol. 63 (14) (2014) 1430–1437. [17] P. Yang, H. Kanki, B. Drolet, T. Yang, J. Wei, P.C. Viswanathan, S.H. Hohnloser, W. Shimizu, P.J. Schwartz, M. Stanton, K.T. Murray, K. Norris, A.L. George, D.M. Roden, Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes, Circulation 105 (16) (2002) pp. 1943-8, 23 4.

211

Forensic Science International: Genetics 42 (2019) 203–212

M. Martinez-Matilla, et al.

[42] [43]

[44]

[45]

[46] [47]

[48]

[49]

[50]

[51] [52]

274 ethnically diverse sudden unexplained deaths, Forensic Sci. Int. 237 (2014) 90–99 4. D.M. Roden, P.C. Viswanathan, Genetics of acquired long QT syndrome, J. Clin. Invest. 115 (8) (2005) 2025–2032. D.J. Tester, L.C. Wong, P. Chanana, A. Jaye, J.M. Evans, D.R. FitzPatrick, M.J. Evans, P. Fleming, I. Jeffrey, M.C. Cohen, J. Tfelt-Hansen, M.A. Simpson, E.R. Behr, M.J. Ackerman, Cardiac genetic predisposition in sudden infant death syndrome, J. Am. Coll. Cardiol. 71 (11) (2018) 1217–1227 20 3. J. Lukas, A.-K. Giese, A. Markoff, U. Grittner, E. Kolodny, H. Mascher, K.J. Lackner, W. Meyer, P. Wree, V. Saviouk, A. Rolfs, Functional characterisation of alpha-galactosidase a mutations as a basis for a new classification system in fabry disease, PLoS Genet. 9 (8) (2013) e1003632. P. Gajendrarao, N. Krishnamoorthy, S. Selvaraj, F. Girolami, F. Cecchi, I. Olivotto, M. Yacoub, An investigation of the molecular mechanism of double cMyBP-C mutation in a patient with end-stage hypertrophic cardiomyopathy, J. Cardiovasc. Transl. Res. 8 (4) (2015) 232–243 14 6. J.D.C. Marsiglia, M.d.C.P. Batitucci, Fd. Paula, C. Barbirato, E. Arteaga, A.Qd. Araújo, [Study of mutations causing hypertrophic cardiomyopathy in a group of patients from Espirito Santo, Brazil], Arq. Bras. Cardiol. 94 (1) (2010) 10–17 1. R. Frank-Hansen, S.P. Page, P. Syrris, W.J. McKenna, M. Christiansen, P.S. Andersen, Micro-exons of the cardiac myosin binding protein C gene: flanking introns contain a disproportionately large number of hypertrophic cardiomyopathy mutations, Eur. J. Hum. Genet. 16 (9) (2008) 1062–1069 12 9. J.S. Davis, S. Hassanzadeh, S. Winitsky, H. Lin, C. Satorius, R. Vemuri, A.H. Aletras, H. Wen, N.D. Epstein, The overall pattern of cardiac contraction depends on a spatial gradient of myosin regulatory light chain phosphorylation, Cell 107 (5) (2001) 631–641 30 11. D.J. Tester, J.P. Ackerman, J.R. Giudicessi, N.C. Ackerman, M. Cerrone, M. Delmar, M.J. Ackerman, Plakophilin-2 truncation variants in patients clinically diagnosed with catecholaminergic polymorphic ventricular tachycardia and decedents with exercise-associated autopsy negative sudden unexplained death in the young, JACC Clin. Electrophysiol. 5 (1) (2019) 120–127 1 1. A. Stöhr, F.W. Friedrich, F. Flenner, B. Geertz, A. Eder, S. Schaaf, M.N. Hirt, J. Uebeler, S. Schlossarek, L. Carrier, A. Hansen, T. Eschenhagen, Contractile abnormalities and altered drug response in engineered heart tissue from Mybpc3targeted knock-in mice, J. Mol. Cell. Cardiol. 63 (2013) 189–198. A.P. Landstrom, D. Dobrev, X.H. Wehrens, Calcium signaling and cardiac arrhythmias, Circ. Res. 120 (12) (2017) 1969–1993 9 6. S. Huke, Linking myofilaments to sudden cardiac death: recent advances, J. Physiol.

595 (12) (2017) 3939–3947 15 6. [53] M. Zhang, A. Xue, Y. Shen, J.B. Oliveira, L. Li, Z. Zhao, A. Burke, Mutations of desmoglein-2 in sudden death from arrhythmogenic right ventricular cardiomyopathy and sudden unexplained death, Forensic Sci. Int. 255 (2015) 85–88 10. [54] X. Jouven, A. Hagege, P. Charron, L. Carrier, O. Dubourg, J.M. Langlard, S. Aliaga, J.B. Bouhour, K. Schwartz, M. Desnos, M. Komajda, Relation between QT duration and maximal wall thickness in familial hypertrophic cardiomyopathy, Heart (British Cardiac Society) 88 (2) (2002) 153–157 8. [55] K. Uchiyama, K. Hayashi, N. Fujino, T. Konno, Y. Sakamoto, K. Sakata, M.-A. Kawashiri, H. Ino and M. Yamagishi, “Impact of QT variables on clinical outcome of genotyped hypertrophic cardiomyopathy, Ann. Noninvasive Electrocardiol. 14 (1) (2009) 65–71 1. [56] G. Captur, L.R. Lopes, V. Patel, C. Li, P. Bassett, P. Syrris, D.M. Sado, V. Maestrini, T.J. Mohun, W.J. McKenna, V. Muthurangu, P.M. Elliott, J.C. Moon, Abnormal cardiac formation in hypertrophic cardiomyopathy, Circ. Cardiovasc. Genet. 7 (3) (2014) 241–248 6. [57] J. Das K, J. Ingles, R.D. Bagnall, C. Semsarian, Determining pathogenicity of genetic variants in hypertrophic cardiomyopathy: importance of periodic reassessment, Genet. Med. 16 (4) (2014) 286–293 10 4. [58] A.K. Manrai, B.H. Funke, H.L. Rehm, M.S. Olesen, B.A. Maron, P. Szolovits, D.M. Margulies, J. Loscalzo, I.S. Kohane, Genetic misdiagnoses and the potential for health disparities, N. Engl. J. Med. 375 (7) (2016) 655–665 18 8. [59] R. Walsh, K.L. Thomson, J.S. Ware, B.H. Funke, J. Woodley, K.J. McGuire, F. Mazzarotto, E. Blair, A. Seller, Reassessment of Mendelian Gene Pathogenicity Using 7,855 Cardiomyopathy Cases and 60,706 Reference Samples, (2017). [60] B.J. Maron, Clinical course and management of hypertrophic cardiomyopathy, N. Engl. J. Med. 379 (7) (2018) 655–668 16 8. [61] A. Bavle, Venlafaxine induced QTc interval prolongation in a therapeutic dose, Asian J. Psychiatr. 16 (2015) 63–64 8. [62] S. Garcia, M. Schuh, A. Cheema, H. Atwal, P.S. Atwal, Palpitations and asthenia associated with venlafaxine in a CYP2D6 poor metabolizer and CYP2C19 intermediate metabolizer, Case Rep. Genet. (2017) 1–4. [63] B.B. Rasmussen, T.H. Brix, K.O. Kyvik, K. Brøsen, The interindividual differences in the 3-demthylation of caffeine alias CYP1A2 is determined by both genetic and environmental factors, Pharmacogenetics 12 (6) (2002) 473–478 8. [64] V. Perera, A.S. Gross, T.M. Polasek, Y. Qin, G. Rao, A. Forrest, J. Xu, A.J. McLachlan, Considering CYP1A2 phenotype and genotype for optimizing the dose of olanzapine in the management of schizophrenia, Expert Opin. Drug Metab. Toxicol. 9 (9) (2013) 1115–91137 4 9.

212