Cardiac channelopathy testing in 274 ethnically diverse sudden unexplained deaths

Forensic Science International 237 (2014) 90–99

Contents lists available at ScienceDirect

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

Cardiac channelopathy testing in 274 ethnically diverse sudden unexplained deaths Dawei Wang a, Krunal R. Shah a, Sung Yon Um a, Lucy S. Eng a, Bo Zhou a, Ying Lin a, Adele A. Mitchell b, Leze Nicaj c, Mechthild Prinz d, Thomas V. McDonald e, Barbara A. Sampson c, Yingying Tang a,* a

Molecular Genetics Laboratory, New York City Office of Chief Medical Examiner, New York, NY 10016, United States Forensic Biology Department, New York City Office of Chief Medical Examiner, New York, NY 10016, United States c Forensic Pathology Department, New York City Office of Chief Medical Examiner, New York, NY 10016, United States d Formerly Forensic Biology Department, New York City Office of Chief Medical Examiner, New York, NY 10016, United States e Departments of Medicine (Cardiology) & Molecular Pharmacology, Albert Einstein College of Medicine, 1300 Morris Park Ave. Bronx, NY 10461, United States b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 15 July 2013 Received in revised form 23 December 2013 Accepted 24 January 2014 Available online 15 February 2014

Sudden unexplained deaths (SUD) in apparently healthy individuals, for which the causes of deaths remained undetermined after comprehensive forensic investigations and autopsy, present vexing challenges to medical examiners and coroners. Cardiac channelopathies, a group of inheritable diseases that primarily affect heart rhythm by altering the cardiac conduction system, have been known as one of the likely causes of SUD. Adhering to the recommendations of including molecular diagnostics of cardiac channelopathies in SUD investigation, the Molecular Genetics Laboratory of the New York City (NYC) Office of Chief Medical Examiner (OCME) has been routinely testing for six major channelopathy genes (KCNQ1, KCNH2, SCN5A, KCNE1, KCNE2, and RyR2) since 2008. Presented here are the results of cardiac channelopathy testing in 274 well-characterized autopsy negative SUD cases, all with thorough medicolegal death investigation including complete autopsy by NYC OCME between 2008 and 2012. The cohort consisted of 141 infants (92.9% younger than six-month old) and 133 non-infants (78.2% were between 19 and 58 years old). Among the ethnically diverse cohort, African American infants had the highest risks of SUD, and African American non-infants died at significantly younger age (23.7 years old, mean age-at-death) than those of other ethnicities (30.3 years old, mean age-at-death). A total of 22 previously classified cardiac channelopathy-associated variants and 24 novel putative channelopathyassociated variants were detected among the infants (13.5%) and non-infants (19.5%). Most channelopathy-associated variants involved the SCN5A gene (68.4% in infants, 50% in non-infants). We believe this is the first study assessing the role of cardiac channelopathy genes in a large and demographically diverse SUD population drawn from a single urban medical examiner’s office in the United States. Our study supports that molecular testing for cardiac channelopathy is a valuable tool in SUD investigations and provides helpful information to medical examiners/coroners seeking cause of death in SUD as well as potentially life-saving information to surviving family members. Published by Elsevier Ireland Ltd.

Keywords: Sudden death Epidemiology Arrhythmia Ion channels Genetics testing

1. Introduction Sudden unexplained death (SUD) in apparently healthy individuals presents vexing challenges for medical examiners and still remains an important public health priority. SUD is

* Corresponding author at: Molecular Genetics Laboratory, New York City Office of Chief Medical Examiner, 421 East 26th Street, New York, NY 10016, United States. Tel.: +1 212 323 1340; fax: +1 212 323 1540. E-mail address: [email protected] (Y. Tang). 0379-0738/$ – see front matter . Published by Elsevier Ireland Ltd. http://dx.doi.org/10.1016/j.forsciint.2014.01.014

defined by a cause of death which remains unknown after complete autopsies, comprehensive laboratory testing, review of all available history, and performance of a comprehensive scene investigation. According to the Centers for Disease Control and Prevention, more than 4500 infants die suddenly each year in the United States with no immediate or obvious cause of death; half of those deaths remain unexplained after a forensic investigation. Furthermore, the prevalence of sudden unexplained deaths beyond infancy (1–22 years old) is also estimated to be greater than 2000 annually in the United States [1]. Despite a decline in SUD rates over the past decade, the rates are still disproportionately high

D. Wang et al. / Forensic Science International 237 (2014) 90–99

91

2. Materials and methods

Sanger sequencing was used to test the other genes and to perform confirmation of identified variants. For next generation sequencing method, the probes were designed by illumina design studio, and the library preparation was performed using the TruSeq Custom Amplicon Library Preparation Kit v1.5 (Illumina, San Diego, CA, USA) according to manufacturer’s instructions. The genomic DNA input for each sample was from 11 ng to 220 ng. The Illumina ACD1 and in-house positive control samples with known variants in SCN5A exons 2–11 were used as positive controls, and Milli-Q water was used as negative control for each 96-well reaction plate. 5% Phix was spiked-in to the library to monitor the quality of the cluster generation and sequencing reaction. Sequencing was performed on the MiSeq sequencer v2 with the MiSeq Reagent Nano Kit v2 (300 cycles) (Illumina, San Diego, CA, USA) for each plate of up to 96 samples according to the manufacturer’s instructions (MiSeq System user guide). Avadis-NGS commercial software (v1.5.1) was used for NGS data analysis. More than 99% target regions with coverage higher than 300 reads, which were aligned to human genome reference (UCSC hg19) using BWA aligner. Non-synonymous variants with possible damaging effects were validated by conventional Sanger sequencing method. For Sanger sequencing, the exons and intron–exon boundaries were amplified and directly sequenced using big-dye terminator chemistry and automated capillary electrophoresis system 3130xl from Applied Biosystems Life Technologies (Carlsbad, CA). Sequencing data were analyzed by Sequencher 4.9 (Gene Codes Corporation, MI). The nucleotide sequence variants were denoted using Human Genome Variation Society (HGVS)-recommended nomenclature [14].

2.1. Study cohort

2.3. Data analysis

Biological samples from a total of 340 cases of sudden unexpected natural death were submitted to the Molecular Genetics Laboratory in the New York City OCME for cardiac channelopathy testing from 2008 to 2012. Of those cases, 66 cases were excluded from this study based on the following criteria: (1) only an external examination was performed (often due to religious objection to autopsy), and (2) the causes of deaths were explained by autopsy or forensic laboratory findings. Based on this, 274 autopsy negative cases were included in this study. All 274 cases went through a comprehensive forensic investigation, including scene investigation, police investigation, full-autopsy with microscopic examinations of heart and central nervous system, ancillary studies (including toxicology, metabolic screening in all infants, microbiology), andareview of the clinical history when available. Molecular testing for six channelopathy genes (KCNQ1, KCNH2, SCN5A, KCNE1, KCNE2, and RyR2) was performed by the Molecular Genetics Laboratory (accredited by College of American Pathologists).

The study cohort of 274 cases was divided into two main age groups: infants (1 year old group), and non-infants (>1 year old group). Demographic characteristics of the cohort are described in Table 1 and Fig. 1. Previously classified cardiac channelopathy associated variants and novel putative cardiac channelopathies variants were re-evaluated utilizing the recently expanded Exome Sequencing Project (ESP) database [15] and the 1000 genomes database [16], as well as the functional prediction programs, Polyphen 2 [17] and SIFT [18] (Table 2). The roles of the putative cardiac channelopathy associated variants were evaluated by age groups (Tables 3 and 4). The distributions of the common variants in this study cohort were compared to those reported in the ESP database [15] (Table 6). Pearson’s Chi-square test or t-test was performed to assess the statistical significance of the categorical variables or continuous data. The Kaplan–Meier cumulative survival curve was used to evaluate the age-at-death by ethnicity. The Log Rank test and two-tailed t-test were used to evaluate the age-at-death differences among all ethnicities, and between two ethnicities, respectively. Statistical significance is indicated by p < 0.05. SPSS and GraphPad Prism were used for statistical analysis.

amongst certain population groups, especially in African Americans and American Indian/Alaska Natives [2]. It is estimated that 10–35% [3–8] of SUD may be explained by cardiac channelopathies, which affect heart rhythm and cardiac electrical conduction physiology. These disorders comprise of a group of inheritable cardiac arrhythmia syndromes, such as long QT syndrome, catecholaminergic polymorphic ventricular tachycardia (CPVT), Brugada syndrome, and short QT syndrome. Despite extensive attempts at developing nationalized standards in SUD investigation including the establishment of basic guidelines [9,10], currently, wide variation exists across the United States when investigating and certifying SUD [11]. Unfortunately, the use of molecular diagnostics is not commonly made as part of the medical investigation of SUD cases. The situation in European countries is also similar, with the application of genetic testing in routine forensic investigations being very limited [12]. Nevertheless, the New York City (NYC) Office of Chief Medical Examiner (OCME) has routinely integrated molecular testing of six major cardiac channelopathy genes (KCNQ1, KCNH2, SCN5A, KCNE1, KCNE2, and RyR2) in SUD investigations since 2008. This testing panel is significant for over 80% of the long QT syndromes, 25% of Brugada syndrome, and 50% of catecholaminergic polymorphic ventricular tachycardia (CPVT) [13]. The main goal of this study is to evaluate the significance of cardiac channelopathies in SUD within a large and ethnically diverse population from a single urban medical examiner’s office in the United States.

2.2. Molecular testing Molecular analyses were performed for six major cardiac channelopathy genes (KCNQ1, KCNH2, SCN5A, KCNE1, KCNE2 and RyR2) in which disease-causing sequence variants have been previously reported in SUD. Specifically, the panel contained the entire coding regions of the KCNQ1 (NM_000218.2), KCNH2 (NM_000238.2), SCN5A (NM_198056.2), KCNE1 (NM_000219.3), KCNE2 (NM_172201.1) genes, and selected exons of the RyR2 (NM_001035.2) gene (exon 8, 14, 15, 44–47, 49, 88–93, 95–97, 100–105) in 274 cases. Genomic DNA was extracted and purified from postmortem tissues or dried blood cards using standard extraction techniques on the M48 BioRobot (Qiagen, CA). Next-generation sequencing (NGS) method was used for testing exons 2–11 in SCN5A and

3. Results 3.1. Unique demographic characteristics of the study cohort The demographic characteristics of the 274 sudden unexplained deaths (SUD) investigated by the NYC OCME are summarized in Table 1. These 274 cases were divided into two age groups: infants (1 year old, 141 cases) and non-infants (>1 year old, 133 cases). In infants, 92.9% were younger than 6 months. In non-infants, 78.2% were older than 18 years old (the oldest decedent was 58 years old). African American is the leading ethnicity observed in both infants (60.3%) and non-infants (35.3%).

D. Wang et al. / Forensic Science International 237 (2014) 90–99

92

Table 1 Demographic characteristics of 274 sudden unexplained deaths. Variables

Infants (1 year old)

Sub-total 1

0–3 m

4–6 m

7–9 m

10–12 m

1–5y

6–18y

>18y

Gender (M:F) Case, n (%) Whites African Americans Hispanics Asians Others (mixed)* Grand total

1.2

1.8

1.7

1.0

1.4

1.2

1.3

1.7

10 (7.1) 56 (39.7) 21 (14.9) 5 (3.5) 11 (7.8) 103 (73.0)

0 (0) 22 (15.7) 3 (2.1) 0 (0) 3 (2.1) 28 (19.9)

0 7 1 0 0 8

0 0 0 1 1 2

10 (7.1) 85 (60.3) 25 (17.7) 6 (4.3) 15 (10.6) 141 (100)

4 4 2 1

1 (0.8) 13 (9.8) 3 (2.2) 1 (0.8)

29 30 30 15

18 (13.6)

104 (78.2)

(0) (5.0) (0.7) (0) (0) (5.7)

(0) (0) (0) (0.7) (0.7) (1.4)

Non-infants (>1 year old)

(3.0) (3.0) (1.5) (0.8)

11 (8.3)

(21.8) (22.6) (22.6) (11.3)

Sub-total 2

p-Value (sub-total 1 vs. sub-total 2)

1.6

-

34 47 35 17

(25.6) (35.3) (26.3) (12.8)

133 (100)

9.41E 07

– –

% in () is calculated as the case number in each group, divided by the total case number from either infants or non-infants, i.e. 141 for 1 year old group, or 133 for >1 year old group. Ethnicity distribution differences in sub-total 1 and sub-total 2 was evaluated by Chi square test. M, male; F, female. * There were no decedents of mixed ethnicity observed in the >1 year old group so they were not included in the statistical calculation.

While African Americans dominated the ethnic composition in infants, the ethnicity was more diversely distributed among the decedents in non-infants (Table 1, Subtotal 1 vs. Subtotal 2, p = 9.41E 07, Chi-square test). To assess the SUD risk in different ethnic groups, the ethnic composition in our cohort is compared to that of the age-matched New York City general population [19],

[(Fig._1)TD$IG]

from which our cohort was derived. Specifically, among NYC residents who are infants, Hispanic is the leading ethnicity (34%), followed by White (30%), Black (21%) and Asian (10%); among NYC residents who are adults, White (non-Hispanic) is the leading ethnicity (36%), followed by Hispanic (27%), Black (22%), and Asian (13%)[19].

Fig. 1. Kaplan–Meier survival analysis of age-at-death differences by ethnicity in > 1 year old group. Kaplan–Meier cumulative survival distribution analysis was used to compare the age-at-death differences among ethnic groups. The overall significance (p = 0.014) of the age-at-death distribution among groups was tested by Log Rank test. Age-at-death differences between two ethnicity groups were tested by two-tailed t-test, which showed significant difference for African Americans vs. other ethnicities (p = 0.006).

Table 2 Putative cardiac channelopathy variants identified in the cohort. Gene

Base changeb

Amino acid changec

Genotype

Protein location

Polyphen 2 Prediction

Sift prediction

Case IDa

(ESP, EA/AA/ALL, %)

(1000 Genome, %)

I I I I I I I I I I

SCN5A SCN5A SCN5A KCNH2 KCNH2 KCNE1 RyR2 SCN5A SCN5A SCN5A

c.2678 G>A c.2893 C>T c.4459 A>C c.2447 G>T c.2680 C>T c.199 C>T c.13291G>A c. 80 G>A c. 674 G>A c. 1502 A>G

p.R893H p.R965C p.M1487L p.G816V p.R894C p.R67C p.E4431K p. R27H p. R225Q p. D501G

G/A C/T A/C G/T C/T C/T G/A G/A G/A A/G

DII-S5/S6 DII/DIII DIII/DIV C-terminal C-terminal C-terminal S1/S2 N-term DI-S4 DI-DII

– – – – – – – – – –

– – – – – – – – – –

Probably Probably Benign Probably Probably Probably Benign Probably Probably Probably

Damaging Damaging Damaging Damaging Damaging Damaging Tolerated Damaging Damaging Tolerated

AN3H AN4A AN9B AN28W AN29B AN33H AN36W AN40H AN41A AN43B

I I I I I I I I I I

SCN5A SCN5A SCN5A SCN5A KCNQ1 KCNE2 SCN5A SCN5A SCN5A SCN5A

c.3392 C>T c.3911 C>T c.4594 G>A c.5494 C>G c.1552 C>T c.161 T>C c.6010 T>G c. 1019 G>A c.3299 C>T c.3539 C>T

p.T1131I p.T1304M p.V1532I p.Q1832E p.R518X p.M54T p.F2004V p. R340Q p.A1100V p.A1180V

C/T C/T G/A C/G C/T;T/T T/C T/G G/A C/T C/T

DII/DIII DIII-S4 DIV-S1 C-terminal C-terminal Transmembrane C-terminal DI-S5/S6 DII/DIII DII/DIII

0.0/0.0237/0.0079 0.0472/0.0239/0.0395 0.0/0.1713/0.0561 0.0/0.0702/0.0235 0.0116/0.0/0.0077 0.0349/0.0/0.0231 – – 0.0118/0.0721/0.0317 0.0/0.0227/0.0077

– – – – – – 0.13 (EUR)/0.14 (ALL) 0 (AFR), 0.05 (ALL) 0.203 (AFR)’0.09 (ALL) 0.25 (CHB  CHS)/0.04(ALL)

Benign Probably damaging Possibly damaging Possibly damaging Stop codon variant Probably damaging Benign Probably damaging Benign Benign

Tolerated Damaging Tolerated Damaging AN21 W;AN22B Damaging Tolerated Tolerated Tolerated Damaging

AN6B AN8B AN10B AN11B/H AN34W AN12W AN42B AN5B AN7A

I? I?

SCN5A KCNE2

c.1852 C>T c.170 T>C

p.L618F p.I57T

C/T T/C

DI/DII Transmembrane

0.0/0.655/0.2153 0.0233/0.0681/0.0384

0.0(EUR)/0.61(AFR)/0.14(ALL) 0.829 (AMR)/0.50(ALL)

Possibly damaging Probably damaging

Tolerated Damaging

AN1W; AN2B AN35H

II II

SCN5A KCNH2

c.5830 C>T c.1281 C>G

p.R1944X p.Y427X

C/T C/G

C-terminal S1/S2

– –

– –

Stop codon variant Stop codon variant

III III III III III III III III III III III III III III III III III III III

SCN5A SCN5A SCN5A SCN5A SCN5A SCN5A SCN5A SCN5A SCN5A KCNQ1 KCNQ1 KCNQ1 KCNQ1 KCNH2 KCNH2 RyR2 RyR2 RyR2 RyR2

c. 43 A>G c. 446 C>T c.2546 T>A c.2990 C>A c.3316 G>T c.3853 A>G c.3947 G>C c.4949T>A c.5668 G>A c.152 A>G c.493 G>A c.1396 A>G c.1779 C>G c.663 C>A c.2944G>A c.11827 G>A c.12415 A>G c.13411 G>A c.13822 C>T

p. R15G p. A149V p.I849N p.A997D p.A1106S p.S1285G p.R1316P p.L1650H p.E1890K p.Y51C p.V165M p.R466G p.N593K p.H221Q p.D982N p.A3943T p.M4139V p.G4471R p.R4608W

A/G C/T T/A C/A G/T A/G G/C T/A G/A A/G G/A A/G C/G C/A G/A G/A A/G G/A C/T

N-term DI-S1 DII-S5 DII/DIII DII/DIII DIII-S3 DIII-S4 DIV-S4/S5 C-terminal N-terminal S2 C-terminal C-terminal N-terminal C-terminal N-terminal N-terminal S1/S2 S3/S4

– – – – – – – – – – – – – – – – – – –

– – – – – – – – – – – – – – – – – – –

Probably damaging Probably damaging Probably damaging Benign Possibly damaging Possibly damaging Probably damaging Probably damaging Probably damaging Probably damaging Probably damaging Benign Probably damaging Benign Benign Possibly damaging Possibly damaging Probably damaging Benign

Tolerated Damaging Damaging Tolerated Tolerated Damaging Damaging Damaging Damaging Damaging Damaging Tolerated Tolerated Tolerated Tolerated Damaging Tolerated Tolerated Damaging

AN44H AN45B AN14W AN15A/W AN16B AN17H AN18H AN19W AN20B/H AN23B AN25H AN26B AN27B AN31B AN32B AN3H AN37W AN38W AN39W

III III III

KCNQ1 SCN5A SCN5A

c.296C>G c.1984 G>T c.3032 C>T

p.P99R p.A662S p.P1011L

C/G G/T C/T

N-terminal DI/DII DII/DIII

0.0/0.023/0.0078 0.0/0.0455/0.0154 0.0121/0.0/0.0083

– – –

Benign Possibly damaging Benign

Tolerated Damaging Tolerated

AN24B AN13B AN1W

MAF

MAF

damaging damaging damaging damaging damaging damaging damaging damaging

AN11B/H AN30A

93

MAF – minor allele frequency; ESP – Exome Sequencing Project; EA – European Americans; AA – African Americans; AFR – all African individuals; ALL – all individuals; CHB – Han Chinese in Bejing, China; CHS – Southern Han Chinese; EUR – all European individuals; AMR – all American individuals. Underlined MAF is used for ethnically matched comparison. a AN (autopsy negative). The last letter of the case ID indicates ethnicity: A, Asian; B, African American; H, Hispanic; W, White; underlined cases carried two variants. b Base change numbering is based on cDNA sequence. c Amino acid change is based on the reference transcripts cited in the method section.

D. Wang et al. / Forensic Science International 237 (2014) 90–99

Variant classification

D. Wang et al. / Forensic Science International 237 (2014) 90–99

94

Table 3 Infants who carried putative cardiac channelopathy variants. Case IDa

Gene/variant

Class

Sex

Age

Scene

Personal history

Family history

AN1W

SCN5A–L618F; SCN5A–P1011L

1; 3

M

2 Months

Prone sleeping

Negative

AN5B AN6B AN8B AN9B AN11B/H

1 1 1 1 1; 2

F F M F F

2 2 2 6 5

Months Weeks Months Weeks Weeks

Co-sleeping Co-sleeping Co-sleeping Prone sleeping Co-sleeping

Negative Negative History of ‘‘dyspnea’’ Recent fever Prematurity

AN43B AN13B AN15A/W

SCN5A–A1100V SCN5A–T1131I SCN5A–T1304M SCN5A–M1487L SCN5A–Q1832E; SCN5A–R1944X SCN5A–D501G SCN5A–A662S SCN5A–A997D

Decedent’s mother delivered twin at age 13, one stillborn and one died at 3–4 days old Negative Negative Negative Negative Negative

1 3 3

M M M

3 Months 5 Months 4 Weeks

Prone sleeping Supine In baby sling

AN16B AN17H AN45B AN20B/H AN23B AN24B AN27B AN32B AN34W

SCN5A–A1106S SCN5A–S1285G SCN5A–A149V SCN5A–E1890K KCNQ1–Y51C KCNQ1–P99R KCNQ1–N593K KCNH2–D982N KCNE2–M54T

3 3 3 3 3 3 3 3 1

F M M M M F M M M

9 Months 4 Months 11 Days 2 Months 2 Months 2 Months 2 Months 5 Months 3 Days

Prone sleeping Awake Supine sleeping Prone sleeping Co-sleeping Sleeping Co-sleeping; Co-sleeping Supine sleeping

Negative Negative Anti-seizure med (phenobarbital, morphine) detected Negative In ICU for prematurity and complication Negative Recent cold, antidepressant detected Prematurity Heart murmur Respiratory problems and murmur at birth; Fever two weeks ago Negative

AN38W

RyR2–G4471R

3

F

6 Weeks

Prone sleeping and co-sleeping

Negative

Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Decedent’s paternal grandaunt died at 1week old Nephew died suddenly at 1 month old

M, male; F, female. underlined cases carried two variants. a AN (autopsy negative). The last letter of the case ID indicates ethnicity: A, Asian; B, African American; H, Hispanic; W, White; underlined case ID carried two variants.

Furthermore, the distribution of age-at-death by ethnicity was examined using Kaplan–Meier cumulative survival analysis (Fig. 1) and the average age-at-death in different ethnic groups in noninfants was compared using t-tests. Average age-at-death for African Americans was significantly younger than for other ethnic groups (p = 0.006, two-tailed t-test): mean age-at-death for African Americans was 23.7 years old, while the age for Whites, Hispanics, and Asians was 32.2, 28.6, and 29.3 years old, respectively.

Asians 286; Europeans 379; Ad Mixed Americans 181) [16] were tabulated, when applicable, along with the function predictions from two in silico programs, Polyphen 2 and SIFT (Table 2). 3.2.1. Class I variants The 22 previously classified cardiac channelopathy causing/ associated variants were re-evaluated based on the minor allele frequencies (MAF) in the ESP and/or 1000 genome databases, and were divided into the following three groups (Table 2):

3.2. Classification of the putative cardiac channelopathy variants The clinical relevance of sequence variants was evaluated in accordance with the American College of Medical Genetics (ACMG) Recommendations for Standards for Interpretation and Reporting of Sequence Variations: Revision 2007 [20]. Variants are divided into six classes: Class I (Sequence variation is previously reported and is a recognized cause of a disorder); Class II (Sequence variation is previously unreported and is of the type which is expected to cause a disorder); Class III (Sequence variation is previously unreported and is of the type which may or may not be causative of the disorder); Class IV (Sequence variation is previously unreported and is probably not causative of disease); Class V (Sequence variation is previously reported and is a recognized neutral variant); and Class VI (Sequence variation is not known or expected to be causative of disease, but is found to be associated with a clinical presentation). For this study, Class I, II, and III variants were considered as putative channelopathyassociated. In total, 22 previously classified channelopathy-associated variants (Class I) and 24 novel putative channelopathy-associated variants (Class II and Class III) were identified in the study cohort. Ethnically matched minor allele frequencies (MAF) from the Exome Sequencing Project (ESP) database (over 2000 African Americans and 4000 European Americans) [15], and/or the 1000 genome database (1092 individuals, African Americans 246;

Not reported in the ESP and 1000 genome databases: ten variants were not reported in the ESP or 1000 genome databases. These variants are listed in the first section of Table 2. These variants are rare, and therefore, they are likely to be disease-associated. The predicted function effects from Polyphen 2 and SIFT programs do not always agree, and it is known that these prediction programs can only accurately predict less than 60% of the deleterious variants [21]. Reported but with low MAF (<0.5%) in the ESP or 1000 genome databases: ten variants were reported in either ESP or 1000 genome with MAF < 0.5% (see underlined ethnically matched population where applicable). These variants are listed in the second section of Table 2. Among these variants, one variant was a stop codon variant and the rest were missense variants. Because the MAF is very low, these variants are still suspected to be disease associated. Reported with high MAF (>0.5%) in the ESP and/or 1000 genome databases: Conventionally, a variant is considered to be a common variant when MAF is >1% in the general population [22]. Common variants are not generally considered responsible for early onset, life-threatening, autosomal dominant diseases with full penetrance. In contrast, a common variant may still be causal for a disease with reduced penetrance or variable expressivity, such as cardiac channelopathies. Here, we set a more stringent MAF cut-off at 0.5% to re-examine

D. Wang et al. / Forensic Science International 237 (2014) 90–99

95

Table 4 Non-infants who carried putative cardiac channelopathy variants. Case IDa Gene/variant

Class Sex Age (years) Scene

Personal history

Family history

AN2B AN3H

1 1; 3

F M

33 41

No witness Witnessed choking

Hypothyroidism and depression Negative

Negative Negative

Sleeping

AN4A

SCN5A–L618F SCN5A–R893H; RyR2–A3943T SCN5A–R965C

1

M

41

AN7A AN10B

SCN5A–A1180V 1 SCN5A–V1532I 1

M F

23 27

AN12W

SCN5A–F2004V 1

M

39

AN40H AN41A AN42B AN14W AN18H AN19W

SCN5A–R27H SCN5A–R225Q SCN5A–R340Q SCN5A–I849N SCN5A–R1316P SCN5A–L1650H

1 1 1 3 3 3

M M M F M F

38 41 20 21 Months 33 37

AN44H AN21W

SCN5A–R15G KCNQ1–R518X

3 1

M M

18 48

AN22B AN25H AN26B AN28W AN29B

KCNQ1–R518X KCNQ1–V165M KCNQ1–R466G KCNH2–G816V KCNH2–R894C

1 3 3 1 1

F M F F F

16 15 54 27 38

AN30A AN31B AN33H AN35H

KCNH2–Y427X KCNH2–H221Q KCNE1–R67C KCNE2–I57T

2 3 1 1

F F M M

29 3 20 39

AN36W AN37W

RyR2–E4431K RyR2–M4139V

1 3

M M

25 23

AN39W

RyR2–R4608W

3

M

34

Irregular heartbeat and bipolar disorder; on olanzapine Awake, no strenuous activity Negative Awake, no strenuous activity Unspecific psychiatric history, on bupropion, lorazepam, amphetamine, ariprazole No witness Depression on med (clozapine, metoprolol, benzotropine, d ivalproex, simvastatin) Sleeping Negative Sleeping Negative Sleeping Negative Sleeping Mild fever Awake, no strenuous activity Negative Sleeping Morbid obese; cardiac hypertrophy, HT cardiovascular Sleeping ‘‘Syncope’’ Sleeping Sleep apnea at 18 years; mild cardiac hypertrophy Sleeping Complaint of abnormal pain; asthma Playing video game History of ‘‘syncope’’ Witnessed seizure Negative Awake, no strenuous activity Obesity, gastric band adjustment Witnessed clutching throat Hypothyroidism with goiter, hypertension Sleeping Negative Sleeping Negative Sleeping Negative Sleeping Schizophrenia on med (risperoldol and benzotrophine), open wound infection Complained heart palpitation History of palpitations Complained chest pain in a concert Personal history of Galloping heart beat; treated for tachycardia;

Witnessed seizure

Past history of arrhythmia

Negative Negative Negative

Negative

Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Long history of heart irregularities on paternal side. Father was diagnosed with PFO after aphasic episode 10 years ago. He had irregular heartbeat. The paternal grandmother has a history of irregular heartbeat, as do paternal uncle and nephew. One of the decedent’s sister had irregular heartbeat and the other had heart murmur as a child Negative

M, male; F, female. Underlined cases carried two variants. a AN (autopsy negative). The last letter of the case ID indicates ethnicity: A, Asian; B, African American; H, Hispanic; W, White; underlined case ID carried two variants.

previously classified cardiac channelopathy-associated variants. Two variants were reported in the ESP and/or 1000 genome databases, with MAF > 0.5% (the ethnically matched population was used where applicable, otherwise global MAF was used). These variants are listed in the third section of Table 2. SCN5AL618F [23,24] was found in two decedents in our cohort: a 33 years old African American female and a 2-month-old white male who also carried a second putative cardiac channelopathy variant, SCN5A-P1011L. The MAF for this variant in African Americans was reported as 0.66% in ESP and as 0.61% in the 1000 genome database, but the MAF for this variant in Whites was zero in both databases. No function or family studies are available for supporting this variant’s involvement in channelopathy and conflicting function predictions (tolerate by SIFT and probably damaging by Polyphen 2) were obtained. KCNE2–I57T was found in a 39 years old Hispanic male in our cohort. This variant was reported in ESP with MAF 0.038% in all Americans (no Hispanic American population in ESP) and 1000 genome

with MAF 0.8% in Americans and 0.5% in all 1092 individuals. There is no Hispanic American classification in the 1000 genome database, but this variant has a high frequency in Mexicans (MAF = 1.5%) and Puerto Ricans (MAF = 0.9%). Functional studies [25–27] supported abnormal channel functions when this variant was present. This variant was predicted as ‘‘damaging’’ by Polyphen 2 and SIFT. Currently, there are no guidelines for reclassifying these two variants, so they remained as Class I variants in this study. 3.2.2. Class II variants Two previously unreported stop-codon variants were found in our cohort (the fourth section of Table 2). Neither was reported in the ESP or 1000 genome databases. They are expected to be disease-associated. 3.2.3. Class III variants Twenty-two previously not reported missense variants were found in this study cohort (the last section of Table 2). Nineteen variants were neither reported in the ESP nor in the 1000 genome

D. Wang et al. / Forensic Science International 237 (2014) 90–99

96

Table 6 Common variants in the cardiac ion channel genes.

database. Most variants were predicted as functionally damaging by SIFT or Polyphen 2. Three variants were reported in the ESP database with a MAF lower than 0.5%. Because there are no functional or family studies for these variants, the pathogenicity of these variants is yet to be determined and they are considered putatively disease-associated for this study.

Gene/variant

Ethnicity

MAF

SCN5A–S524Y

Total Whites African Americans

2.555 2.273 4.167

1.140 0.084 3.309

SCN5A–H558R

Total Whites African Americans

26.277 21.591 31.818

24.628 23.209 27.483

SCN5A–S1103Y

Total Whites African Americans

4.380 1.136 7.955

2.433 0.036 7.287

KCNQ1–G643S

Total Whites African Americans

1.095 0.000 1.515

0.689 0.047 1.953

KCNQ1–V648I

Total Whites African Americans

1.460 0.000 3.030

0.613 0.024 1.774

KCNH2–K897T

Total Whites African Americans

11.679 27.273 5.682

16.908 23.256 4.517

KCNE1–S38G

Total Whites African Americans

35.402 34.091 37.121

33.808 36.314 28.915

Our study (%)

3.3. Putative cardiac channelopathy associated variants in infants Among 141 infants, nineteen infants (13.5%, 19/141) carried at least one of the Class I-3 variants (Table 3). Two infants carried two variants (case ID AN1W and AN11B/H). The majority of variants were found in the SCN5A gene (68.4%, 13/19) and in African American infants (63.2%, 12/19). Co-sleeping (42.1%, 8/19) and prone-sleeping (31.6%, 6/19) were frequent. Half of the infants (52.6%, 10/19) had records of medical complaints, but only three decedents had relevant family history (15.8%, 3/19). 3.4. Putative cardiac channelopathy associated variants in noninfants Twenty-six out of 133 (19.5%, 26/133) non-infants carried at least one of the Class I-3 variants (Table 4). One 41 years old Hispanic male carried two variants (case ID AN3H). Compared to in infants, variants were more diversely distributed among the 6 genes in non-infants, even though half of the variants were found in SCN5A (13/26). None of the non-infants were undergoing any strenuous physical activity at the time of death. More than half of the decedents (61.5%, 16/26) had past medical complaints, but only one had relevant family history.

ESP (%)

MAF, minor allele frequency; ESP, Exome Sequencing Project.

MAF in Whites in our cohort than that in ESP database (no statistical analysis performed due to small sample size).

3.5. Distribution of the putative channelopathy associated variants

4. Discussion

When examining the ethnicity, gender, age, and death circumstances, interesting patterns were revealed in the distribution of the putative channelopathy associated variants (Table 5). African Americans carried more SCN5A and KCNQ1 variants than other ethnic groups, while Whites carried more RyR2 variants; KCNH2 variants were more frequently observed in non-infant females, while SCN5A and RyR2 variants were more frequently seen in non-infant males. Furthermore, the variants in the five long QT genes (SCN5A, KCNQ1, KCNH2, KCNE1, and KCNE2) were found more frequently in the decedents who were sleeping at the time of death; in contrast, the variants in CPVT gene, RyR2, were more often seen in the decedents who were active at the time of death.

We present data from a comprehensive postmortem evaluation of 274 sudden unexplained deaths in New York City. Here we discuss the unique characteristics of the cohort, principal findings from this study, the benefits of molecular testing to the at-risk family members, and the challenges of the postmortem molecular diagnostics along with the limitations of this study. 4.1. What is unique about this study cohort? We believe that our cohort constitutes the largest, most ethnically diverse SUD population for which molecular testing for all 6 cardiac channelopathy genes has been performed to be reported in one study to date. Among previously reported cardiac channelopathy tested SUD cohorts in the United States, the largest infant cohort had 93 decedents, where 63% were Whites [28]. Furthermore, in a study published by Plant et al. in 2006 [29], only one gene, SCN5A, was screened in 224-autopsy negative cases. Also, in a study conducted by Tester et al. in 2012 [8], the majority of the cohort was White (88%, 153 in total 173), and among the 173 cases, the author acknowledged that the first 49 of these 173 consecutively referred cases have been reported previously. Internationally, a study conducted by Arnestad et al. [3] included 201 Sudden Infant Death Syndrome (SIDS) cases, who were all from

3.6. Common variants in the cardiac ion channel genes Common variants are Class V variants, which are not considered as disease-causing. As in the ESP database, the distribution of three SCN5A common variants (S524Y, H558R, and S1103Y) and two KCNQ1 variants (G643S and V648I), all showed a pattern of a MAF higher for African Americans than for Whites; whereas the KCNH2 variant (K897 T) showed a pattern of a higher MAF for Whites than for African Americans (Table 6). Interestingly, we found that two variants, SCN5A–S524Y and SCN5A–S1103Y, presented higher

Table 5 Putative channelopathy associated variant distributions among 274 SUDS decedents. Variables

SCN5A

KCNQ1

KCNH2

KCNE2

KCNE1

RyR2

Ethnicity distribution Gender distribution Age distribution Death circumstances

11B, 5W, 4O, 5H, 3A 18M, 10F 15 (1 y), 13 (>1 y) 94% Sleeping in infants

5B, 1W, 1H 4M, 3F 3 (1 y), 4 (>1 y) 71% Sleeping

3B, 1W, 1A 4F, 1M 1 (1 y), 4 (>1 y) 60% Sleeping

1W, 1H 2M 1 (1 y), 1 (>1 y) 100% Sleeping

1H 1M 1 (>1 y) 100% Sleeping

4W, 1H 4M, 1F 1 (1 y), 4 (>1 y) 80% Active

B, African Americans; W, Whites; H, Hispanics; A, Asians; O, others; F, female; M, male.

D. Wang et al. / Forensic Science International 237 (2014) 90–99

the southeastern region of Norway where ethnicity was not specified in the paper, but is expected to be homogenously White. In addition, the RyR2 gene was not tested in that study. The role of KCNQ1 and KCNH2 genes were also examined in 163 consecutive bodies found in water in Finland where the cohort is presumably Caucasians [30]. Furthermore, three Long QT genes (KCNQ1, KCNH2 and SCN5A) and several major hypertrophic cardiomyopathy genes were examined in 37 presumable Caucasian SCD victims in a study conducted by Brion’s group [31]. Our infant cohort represented 60.3% African Americans, 17.7% Hispanics, 7.1% Whites, and 4.3% Asians, and the non-infant cohort represented 35.3% African Americans, 26.3% Hispanics, 25.6% Whites, and 12.8% Asians. All 274 cases in our SUD cohort were investigated within a single medical examiner’s office where investigation protocol is relatively standardized and stringent cohort inclusion criteria are established for this study, which is important for ascertaining the role of cardiac channelopathy in sudden unexplained deaths and correlating genotype and phenotype results when possible. Our data suggested that African American infants had the highest risks of SUD compared to infants in other ethnicity groups. In addition, age-at-death among non-infants by ethnicity data suggested that African Americans died of SUD at significantly younger age compared to other ethnic groups (23.7 years vs. 30 .3 years, mean age-at-death, p = 0.006). This result agreed with the concept that African Americans are more susceptible to various cardiovascular diseases associated with early death [32–35]. The reasons for such findings remained to be further explored in order to determine whether the social and economic differences or accessibility to proper health care, or genetic make-up made the differences. 4.2. Principal findings in this study In agreement with previously published studies [5–8,36], we found that a cardiac channelopathy may help to explain the cause of death in approximately 13.5% of infant and 19.8% of non-infant SUD victims. We found most channelopathy-associated variants appeared in the SCN5A gene, and the roles of other cardiac channelopathy genes became more prominent in non-infants than in infants. KCNH2 and RyR2 variants were more frequently observed in females and males, respectively. In addition, concerning the activities at the time of death, we noticed more variants in the SCN5A gene from decedents who were sleeping and observed more variants in the RyR2 gene in the decedents who were active [37,38]. However, putative channelopathy-associated variants displayed notable differences between our data and the published data: (1) we observed more KCNQ1 mutations in African Americans, which has not been previously reported (Table 5). (2) Others reported more significant roles for KCNQ1 [8] and RyR2 [8] genes in adults, whereas we observed a high percentage of SCN5A mutations in adults (Table 4). In addition, it is interesting that two common variants, SCN5A–S524Y and SCN5A–S1103Y, were higher in Whites in our study cohort compared to the ethnically matched population in ESP (no statistical analysis performed due to small sample size) (Table 6). It has been reported that certain environmental or genetic modifiers may interact with the common cardiac channelopathy gene variants. For example, SCN5A–S524Y reportedly displayed a different pharmacological response from the wild-type gene to various antiarrhythmic agents, including quinidine and flecainide [39]. SCN5A–S1103Y was reported to increase the rate of cardiac sodium channel activation and to alter the inactivation of gating at low pH [29]. It is unknown if any genetic modifiers existed in the White population from our cohorts.

97

It was reported that people carrying double putative channelopathy-associated variants often manifest earlier on-set or more severe presentation of the diseases [40]. In our study, three decedents carried double variants, two infants (2 months and 5 weeks old, respectively) and one adult (41 years old male). The double variant carriers represented 10.5% (2/19) among infants who were positive for a putative disease-associated variant, and 3.8% (1/26) non-infants who were positive for a putative diseaseassociated variant. Although the number is too small to evaluate clinical significance, we believe as we expand the number of disease genes in our testing, we may be able to ascertain the significance of infant decedents carrying double or multiple putative disease-causing variants versus adult decedents or known channelopathy patients. 4.3. How postmortem molecular testing results can help the families? Postmortem molecular testing for cardiac channelopathies is an important diagnostic tool which establishes an etiologically specific cause of death in many cases for which the cause of death would otherwise remain unexplained or unknown. Such testing provides relevant information to at-risk family members and, therefore, enables them to seek clinical evaluation and genetic counseling before a catastrophic manifestation of the disease strikes [12,31]. Here, we illustrate this concept through two of our recently published studies [41,42]. Both studies involved putative channelopathy-causing variants identified in the 66 cases that were tested in our lab but excluded in this study cohort because of the presence of other compounding contributing factors to the cause of death. Both studies were conducted through collaborating with Einstein/Montefiore Cardiogenetics Program, a clinical genetic program with adult and pediatric cardiology expertise. In one case, the KCNQ1–S277L variant was found in a decedent who presented with sudden cardiac death in the presence of cocaine use. Through electrophysiology and biochemical analysis, it was demonstrated that KCNQ1–S277L variant in the decedent resulted in defective channels that compromised the repolarization reserve, thereby enhancing the arrhythmic susceptibility to pharmacological blockage of the I (Kr) current by cocaine [41]. High-risk relatives were tested for the presence of this variant and were subjected to close cardiac monitoring and drug education. Using similar functional analysis methods, in a second study, we demonstrated that another variant, KCNH2–G816V, caused a trafficking defect that acted in a partially dominant negative manner; possible hypokalemia in the decedent, which induced wild-type KCNH2 protein degradation combined with haploinsufficiency further compromising the repolarization reserve and causing the lethal arrhythmia [42]. This information becomes important in counseling and treatment of the at-risk family members to avoid hypokelemic events. Furthermore, both studies illustrated the importance of postmortem cardiac channelopathy testing in complex cases where channelopathy is not the only risk or underlying cause of death. 4.4. Challenges of interpreting cardiac channelopathy variants Interpreting the clinical effect of a variant identified in a cardiac channelopathy gene in postmortem molecular diagnostics can be quite challenging. There are several layers of complexities: (1) unlike in a clinical setting, there are no clinical symptoms or clinical evaluation of the subject that can be used to correlate with the molecular findings in the postmortem diagnosis; (2) often neither in vitro functional studies nor extensive family studies are available to assist with the variant interpretation; (3) most types of cardiac channelopathies are inherited in an autosomal dominant

98

D. Wang et al. / Forensic Science International 237 (2014) 90–99

manner, exhibiting both reduced penetrance (when a diseasecausing variant carrier does not manifest a disease at all) and variable expressivity (when different carriers of the same diseasecausing variant manifest different symptoms of a wide spectrum of disease, ranging from mild, to severe, to life-threatening); (4) because of the rapid changes of the current literature, fast expanding databases, and the dynamic function of prediction programs, interpretation of a variant may evolve over time. Currently, however, there is a lack of consensus and guidelines in the field of cardiac channelopathies for cardiac channelopathy variant classification. There are also limitations associated with the published literature, the ESP and 1000 genome databases, and in silico function prediction programs. Early literatures (published before the release of the ESP and 1000 genome project dataset) often used a small numbers of local population controls to evaluate a putative disease associated variant. Typically a variant was reported in a single patient in a family in which there were not enough affected individuals available to demonstrate co-segregation of the variant and the disease. Also, in vivo or in vitro functional studies were not always available. So the certainty of the pathogenicity of a variant is limited. On the other hand, even though the ESP and 1000 genome databases contain large numbers of people within ethnically defined populations, approximately 6500 and 1092 people, respectively, the clinical evaluations, especially cardiac evaluations of the individuals in the databases are unknown to the end users, which make these databases less than ideal to be used as ‘‘healthy controls’’. In addition, there is no ethnically matched population for Hispanics, Asians or mixed ethnicities in the ESP database, and the population in 1000 genome is divided by the geographic regions, and not by genetic ethnicities. Nevertheless, the MAF in the ESP and 1000 genome databases provide useful information to re-evaluate a variant if it is rare or common [43–45]. Furthermore, in silico functional prediction programs, such as Polyphen 2 [17] and SIFT [18], can be helpful for variant evaluation. Both programs estimate the impact of an amino acid change in the structure and function of a protein using physical and comparative considerations. However, these two programs do not always give concordant prediction results due to different analysis algorithm, and both programs were shown to have limited accuracy of prediction (<60%) [21]. 4.5. Limitations of this study and future direction Due to the technological capacity of the traditional Sangersequencing method, we were only able to test selected exons of RyR2 gene, and 5 out of the 13 currently known long QT syndrome genes [13]. We believe it is important that additional arrhythmia genes are included in the future testing panel. In addition, it is also important to test and evaluate the candidate genes involved in pathways leading to cardiomyopathies, disturbances of the central nervous system, immune response, metabolism, and beyond. As molecular diagnostics laboratories begin to embrace next generation technologies which enable faster, cheaper and more efficient testing, our understanding of the inherited diseases underlying sudden unexplained death will expand [31]. Genetic testing information should be provided to the family members with proper counseling along with the choices of further clinical evaluation. This will undoubtedly further advance the quality of forensic pathologists’ investigations and increase the benefits to at-risk families. Source of funding This project was partially supported by Awards no. 2009-CD-BX0049 and 2011-DN-BX-K535, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice.

Conflict of interest statement There is no conflict of interest to be disclosed for any author on this paper. Acknowledgement We thank Dr. Jason K. Graham for critical review of this manuscript.

References [1] M.J. Ackerman, D.J. Tester, D.J. Driscoll, Molecular autopsy of sudden unexplained death in the young, Am. J. Forensic Med. Pathol. 22 (2001) 105–111. [2] T.J. Mathews, M.F. MacDorman, Infant mortality statistics from the 2008 period linked birth/infant death data set, Natl. Vital Stat. Rep. 60 (2012) 27. [3] M. Arnestad, L. Crotti, T.O. Rognum, R. Insolia, M. Pedrazzini, C. Ferrandi, et al., Prevalence of long-QT syndrome gene variants in sudden infant death syndrome, Circulation 115 (2007) 361–367. [4] S.S. Chugh, O. Senashova, A. Watts, P.T. Tran, Z. Zhou, Q. Gong, et al., Postmortem molecular screening in unexplained sudden death, J. Am. Coll. Cardiol. 43 (2004) 1625–1629. [5] P.A. Gladding, C.A. Evans, J. Crawford, S.K. Chung, A. Vaughan, D. Webster, et al., Posthumous diagnosis of long QT syndrome from neonatal screening cards, Heart Rhythm 7 (2010) 481–486. [6] J.R. Skinner, J. Crawford, W. Smith, A. Aitken, D. Heaven, C.A. Evans, et al., Prospective, population-based long QT molecular autopsy study of postmortem negative sudden death in 1–40 year olds, Heart Rhythm 8 (2011) 412–419. [7] D.J. Tester, M.J. Ackerman, Postmortem long QT syndrome genetic testing for sudden unexplained death in the young, J. Am. Coll. Cardiol. 49 (2007) 240–246. [8] D.J. Tester, A. Medeiros-Domingo, M.L. Will, C.M. Haglund, M.J. Ackerman, Cardiac channel molecular autopsy: insights from 173 consecutive cases of autopsynegative sudden unexplained death referred for postmortem genetic testing, Mayo Clin. Proc. 87 (2012) 524–539. [9] Centers for Disease Control and Prevention, Guidelines for death scene investigation of sudden unexplained infant deaths: recommendations of the interagency panel on sudden infant death syndrome, MMWR Morb. Mort. Wkly. 45 (1996) 1–22. [10] C. Basso, M. Burke, P. Fornes, P.J. Gallagher, R.H. De Gouveia, M. Sheppard, et al., Guidelines for autopsy investigation of sudden cardiac death, Pathologica 102 (2010) 391–404. [11] L.B. Shields, J.C. Hunsaker 3rd, T.S. Corey, D. Stewart, Is SIDS on the rise??, J. Ky. Med. Assoc. 105 (2007) 343–353. [12] K. Michaud, P. Mangin, B.S. Elger, Genetic analysis of sudden cardiac death victims: a survey of current forensic autopsy practices, Int. J. Legal Med. 125 (2011) 359–366. [13] D.J. Tester, M.J. Ackerman, Cardiomyopathic and channelopathic causes of sudden unexplained death in infants and children, Annu. Rev. Med. 60 (2009) 69–84. [14] J.T. den Dunnen, S.E. Antonarakis, Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion, Hum. Mutat. 15 (2000) 7–12. [15] Exome Variant Server NGESPE, Seattle, WA, http://evs.gs.washington.edu/EVS/ (accessed 13.06.13). [16] G.R. Abecasis, A. Auton, L.D. Brooks, M.A. DePristo, R.M. Durbin, R.E. Handsaker, et al., An integrated map of genetic variation from 1092 human genomes, Nature 491 (2012) 56–65. [17] I.A. Adzhubei, S. Schmidt, L. Peshkin, V.E. Ramensky, A. Gerasimova, P. Bork, et al., A method and server for predicting damaging missense mutations, Nat. Methods 7 (2010) 248–249. [18] P. Kumar, S. Henikoff, P.C. Ng, Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm, Nat. Protoc. 4 (2009) 1073– 1081. [19] Population of Infants CaTbRE, NYC 2010 Census, http://www.nyc.gov/html/dcp/ html/census/demo_tables_2010.shtml (accessed 30.10.13). [20] C.S. Richards, S. Bale, D.B. Bellissimo, S. Das, W.W. Grody, M.R. Hegde, et al., ACMG recommendations for standards for interpretation and reporting of sequence variations: revisions 2007, Genet. Med. 10 (2008) 294–300. [21] S.E. Flanagan, A.M. Patch, S. Ellard, Using SIFT and PolyPhen to predict loss-offunction and gain-of-function mutations, Genet. Test. Mol. Biomark. 14 (2010) 533–537. [22] K.A. Frazer, S.S. Murray, N.J. Schork, E.J. Topol, Human genetic variation and its contribution to complex traits, Nat. Rev. Genet. 10 (2009) 241–251. [23] D.J. Tester, M.L. Will, C.M. Haglund, M.J. Ackerman, Compendium of cardiac channel mutations in 541 consecutive unrelated patients referred for long QT syndrome genetic testing, Heart Rhythm 2 (2005) 507–517. [24] P. Yang, H. Kanki, B. Drolet, T. Yang, J. Wei, P.C. Viswanathan, et al., Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes, Circulation 105 (2002) 1943–1948. [25] Z.A. McCrossan, T.K. Roepke, A. Lewis, G. Panaghie, G.W. Abbott, Regulation of the Kv2.1 potassium channel by MinK and MiRP1, J. Membr. Biol. 228 (2009) 1–14. [26] F. Sesti, G.W. Abbott, J. Wei, K.T. Murray, S. Saksena, P.J. Schwartz, et al., A common polymorphism associated with antibiotic-induced cardiac arrhythmia, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 10613–10618.

D. Wang et al. / Forensic Science International 237 (2014) 90–99 [27] J. Wu, W. Shimizu, W.G. Ding, S. Ohno, F. Toyoda, H. Itoh, et al., KCNE2 modulation of Kv4.3 current and its potential role in fatal rhythm disorders, Heart Rhythm 7 (2010) 199–205. [28] D.J. Tester, M.J. Ackerman, Sudden infant death syndrome: how significant are the cardiac channelopathies, Cardiovasc. Res. 67 (2005) 388–396. [29] L.D. Plant, P.N. Bowers, Q. Liu, T. Morgan, T. Zhang, M.W. State, et al., A common cardiac sodium channel variant associated with sudden infant death in African Americans, SCN5A S1103Y, J. Clin. Invest. 116 (2006) 430–435. [30] P. Lunetta, A. Levo, P.J. Laitinen, H. Fodstad, K. Kontula, A. Sajantila, Molecular screening of selected long QT syndrome (LQTS) mutations in 165 consecutive bodies found in water, Int. J. Legal Med. 117 (2003) 115–117. [31] C. Allegue, R. Gil, A. Blanco-Verea, M. Santori, M. Rodriguez-Calvo, L. Concheiro, et al., Prevalence of HCM and long QT syndrome mutations in young sudden cardiac death-related cases, Int. J. Legal Med. 125 (2011) 565–572. [32] E. Barnett, J. Halverson, Disparities in premature coronary heart disease mortality by region and urbanicity among black and white adults ages 35–64, 1985–1995, Public Health Rep. 115 (2000) 52–64. [33] A.D. Callow, Cardiovascular disease 2005–the global picture, Vasc. Pharmacol. 45 (2006) 302–307. [34] K.C. Ferdinand, African American heart failure trial: role of endothelial dysfunction and heart failure in African Americans, Am. J. Cardiol. 99 (2007) 3D–6D. [35] C. Fincher, J.E. Williams, V. MacLean, J.J. Allison, C.I. Kiefe, J. Canto, Racial disparities in coronary heart disease: a sociological view of the medical literature on physician bias, Ethn. Dis. 14 (2004) 360–371. [36] D.J. Tester, D.B. Spoon, H.H. Valdivia, J.C. Makielski, M.J. Ackerman, Targeted mutational analysis of the RyR2-encoded cardiac ryanodine receptor in sudden unexplained death: a molecular autopsy of 49 medical examiner/coroner’s cases, Mayo Clin. Proc. 79 (2004) 1380–1384.

99

[37] S.G. Priori, C. Napolitano, M. Memmi, B. Colombi, F. Drago, M. Gasparini, et al., Clinical and molecular characterization of patients with catecholaminergic polymorphic ventricular tachycardia, Circulation 106 (2002) 69–74. [38] P.J. Schwartz, S.G. Priori, C. Spazzolini, A.J. Moss, G.M. Vincent, C. Napolitano, et al., Genotype-phenotype correlation in the long-QT syndrome: gene-specific triggers for life-threatening arrhythmias, Circulation 103 (2001) 89–95. [39] M. Shuraih, T. Ai, M. Vatta, Y. Sohma, E.M. Merkle, E. Taylor, et al., A common SCN5A variant alters the responsiveness of human sodium channels to class I antiarrhythmic agents, J. Cardiovasc. Electrophysiol. 18 (2007) 434–440. [40] H. Itoh, W. Shimizu, K. Hayashi, K. Yamagata, T. Sakaguchi, S. Ohno, et al., Long QT syndrome with compound mutations is associated with a more severe phenotype: a Japanese multicenter study, Heart Rhythm 7 (2010) 1411–1418. [41] J. Chen, M. Weber, S.Y. Um, C.A. Walsh, Y. Tang, T.V. McDonald, A dual mechanism for I(Ks) current reduction by the pathogenic mutation KCNQ1-S277L, Pacing Clin. Electrophysiol. 34 (2011) 1652–1664. [42] Y. Krishnan, R. Zheng, C. Walsh, Y. Tang, T.V. McDonald, Partially dominant mutant channel defect corresponding with intermediate LQT2 phenotype, Pacing Clin. Electrophysiol. 35 (2012) 3–16. [43] C. Andreasen, L. Refsgaard, J.B. Nielsen, A. Sajadieh, B.G. Winkel, J. Tfelt-Hansen, et al., Mutations in genes encoding cardiac ion channels previously associated with sudden infant death syndrome (SIDS) are present with high frequency in new exome data, Can. J. Cardiol. 29 (9) (2013) 1104–1109. [44] L. Refsgaard, A.G. Holst, G. Sadjadieh, S. Haunso, J.B. Nielsen, M.S. Olesen, High prevalence of genetic variants previously associated with LQT syndrome in new exome data, Eur. J. Hum. Genet. 20 (2012) 905–908. [45] B. Risgaard, R. Jabbari, L. Refsgaard, A. Holst, S. Haunso, A. Sadjadieh, et al., High prevalence of genetic variants previously associated with Brugada syndrome in new exome data, Clin. Genet. 84 (5) (2013) 489–495.