HHS Public Access Author manuscript Author Manuscript
Thromb Haemost. Author manuscript; available in PMC 2016 April 01. Published in final edited form as: Thromb Haemost. 2015 April ; 113(4): 772–781. doi:10.1160/TH14-08-0670.
A genome-wide association study of heparin-induced thrombocytopenia using an electronic medical record Jason H Karnes1, Robert M Cronin1, Jerome Rollin2, Alexander Teumer3, Claire Pouplard2, Christian M Shaffer1, Carmelo Blanquicett1, Erica A Bowton1, James D Cowan1, Jonathan D Mosley1, Sara L Van Driest1, Peter E Weeke1,4, Quinn S Wells1, Tamam Bakchoul3, Joshua C Denny1, Andreas Greinacher3, Yves Gruel2, and Dan M Roden1
Author Manuscript
1Vanderbilt 2UMR
University School of Medicine, Nashville, TN
CNRS 7292 & University Hospital of Tours, Tours Cedex, France
3University
of Greifswald, Greifswald, Germany
4Copenhagen
University Hospital, Gentofte, Denmark
Abstract
Author Manuscript
Heparin-induced thrombocytopenia (HIT) is an unpredictable, potentially catastrophic adverse effect of heparin treatment resulting from an immune response to platelet factor 4 (PF4)/heparin complexes. No genome-wide evaluations have been performed to identify potential genetic influences on HIT. Here, we performed a genome-wide association study (GWAS) and candidate gene study using HIT cases and controls identified using electronic medical records (EMRs) coupled to a DNA biobank and attempted to replicate GWAS associations in an independent cohort. We subsequently investigated influences of GWAS-associated single nucleotide polymorphisms (SNPs) on PF4/heparin antibodies in non-heparin treated individuals. In a recessive model, we observed significant SNP associations (OR 18.52 [6.33–54.23], p=3.18×10−9) with HIT near the T-Cell Death-Associated Gene 8 (TDAG8). These SNPs are in linkage disequilibrium with a missense TDAG8 SNP. TDAG8 SNPs trended toward an association with HIT in replication analysis (OR 5.71 [0.47–69.22], p=0.17), and the missense SNP was associated with PF4/heparin antibody levels and positive PF4/heparin antibodies in non-heparin treated patients (OR 3.09 [1.14–8.13], p=0.02). In the candidate gene study, SNPs at HLA-DRA were nominally associated with HIT (OR 0.25 [0.15–0.44], p=2.06×10−6). Further study of TDAG8 and HLA-DRA SNPs is warranted to assess their influence on the risk of developing HIT.
Author Manuscript
Corresponding Author: Dan M Roden, M.D., MRB4 1285B, 2215B Garland Avenue, Nashville, TN, 37232-0575, Telephone: 615-322-0067, Fax: 615-343-4522, [email protected]. Authorship Contributions For the discovery GWAS cohort, JHK, RMC, EAB, JDM, SVD, PEW, QSW, JCD, and DMR designed the study, interpreted results; JHK, RMC, CB, EAB, and JCD designed case and controls algorithms, reviewed electronic medical records, and adjudicated case and control subjects; JHK, CMS, EAB, JDM, and JCD performed statistical analysis and provided bioinformatic support. For the SHIP study, AG designed the study; AT, AG, and TB interpreted results; AT performed statistical analysis. For the independent replication population, YG designed the study; JR, CP, and YG interpreted results; JR and JHK performed statistical analysis. All authors helped to write the manuscript. Disclosure of Conflicts of Interest The authors declare no competing financial interests.
Karnes et al.
Page 2
Author Manuscript
Introduction Heparin-induced thrombocytopenia (HIT) is an antibody-mediated condition of platelet activation in patients receiving unfractionated heparin (UFH) and low molecular weight heparin (LMWH)(1). HIT develops in up to 0.5–5% of patients treated with heparin anticoagulants, has a greater than 30% mortality rate, and results in catastrophic thromboembolic complications, including life- and limb-threatening thrombosis(2–5). Further complicating the use of heparin anticoagulants, prevention of HIT-related thrombosis is currently possible only after manifestations of HIT are evident and the disease process has already begun(6, 7). The inability to predict HIT thus represents a liability with heparin administration and a priori identification of patients with a high HIT risk could allow for prevention of HIT.
Author Manuscript Author Manuscript
HIT is closely associated with the development of antibodies to complexes of heparin and platelet factor 4 (PF4), a protein normally found in the alpha granules of platelets(8). Although several studies have identified genetic polymorphisms such as the Fc receptor RIIA (FCGR2A) –H131R and Fc receptor RIIIA (FCGR3A) –F158V polymorphisms(9–13) associated with HIT, none have identified genetic risk factors that can reliably predict HIT. Other work has identified genomic predictors of diverse types of immune-mediated adverse drug reactions(14–16), suggesting that strong genetic influences on HIT risk might still be identified. This is of special interest, as HIT is an example for a distinct group of immune reactions, which fulfills neither the criteria of a primary nor the criteria of a typical secondary immune reaction(17–19). The clinical spectrum of HIT includes distinct stages of progression including 1) lack of PF4/heparin antibodies; 2) presence of PF4/heparin antibodies without thrombocytopenia; 3) PF4/heparin antibodies with thrombocytopenia; and 4) PF4/heparin antibodies with thrombocytopenia and thrombosis.(3, 20) As a result, single nucleotide polymorphisms (SNPs) may exert their effects only at distinct stages in the progression of HIT and different SNPs may be predictors of PF4/heparin antibody formation, thrombocytopenia, or thrombosis. The objective of this study was to identify genetic predictors of HIT using genome-wide association study (GWAS) and candidate gene approaches. We conducted this study using HIT cases and heparin-exposed controls identified in an electronic medical record (EMR) linked to a DNA repository. Associated SNPs from the HIT GWAS were genotyped in an independent replication cohort of HIT cases and heparin-treated controls. We also tested whether GWAS-identified SNPs were associated with PF4/heparin antibody levels in nonheparin treated patients.
Author Manuscript
Materials and Methods Study Populations BioVU (Discovery Population)—The discovery population was identified in BioVU, the Vanderbilt DNA databank that links DNA extracted from discarded blood samples to deidentified EMRs(21). This study was approved by the Institutional Review Board at Vanderbilt University as described previously(21). We developed an algorithm to identify individuals with a diagnosis of HIT in the EMR. The algorithm used ICD9 diagnostic codes,
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 3
Author Manuscript
lab results, and natural language processing(22, 23) to analyze narrative text in order to identify patients treated with UFH or LMWH who tested positive for PF4/heparin antibodies (optical density [OD] level >0.7 absorbance units [AU]). UFH or LMWH-treated controls were also identified in the EMR using a control algorithm that employed similar techniques. HIT cases were confirmed by chart review by at least two physician or pharmacist reviewers. HIT cases were scored with the 4Ts scoring system, which estimates the likelihood of HIT based on timing, magnitude, and potential causes of thrombocytopenia and presence of thrombosis and with a maximum 4Ts score of eight(6, 24). Additional information regarding the BioVU resource and case and control algorithms is available in the Supplemental Materials.
Author Manuscript Author Manuscript
Replication of GWAS SNPs—The patient group for replication has been previously described(11, 13) and additional details regarding the replication study population are detailed in the Supplemental Materials. In brief, the replication cohort was comprised of 94 Caucasian individuals with definite HIT who developed delayed-onset thrombocytopenia and/or a significant decrease in platelet count after a median duration of heparin treatment of 10 days. Both PF4-specific ELISA (HAT® GTI, Brookfield, WI) and serotonin release assay were positive in all patients. The first control group (Abneg) consisted of 178 patients who had undergone cardiopulmonary bypass surgery and received high doses of UFH but did not develop an immune response to heparin. The second control group (Abpos) consisted of 154 patients who had also undergone cardiopulmonary bypass surgery with heparin treatment. All had developed significant levels of PF4-specific antibodies but none had a positive serotonin release assay and none had significant abnormal evolution in platelet count in the post-operative period. Blood samples were collected after obtaining informed consent according to the Helsinki declaration principles. Both the University of Tours ethics committee and the Ministry of Research had previously approved collection of DNA from patients for genetic studies (Agreement number DC2008-308).
Author Manuscript
SNP Association with PF4/heparin Antibody Development—The Study of Health in Pomerania (SHIP) is a population-based project in West Pomerania, Germany that has been previously described(25–27). SHIP participants were recruited from population registries by written invitation regardless of previous anticoagulant treatment. Additional details regarding the SHIP population are available in the Supplemental Materials. The SHIP population comprised 4,308 participants who had sera collected and tested for anti-PF4/ heparin IgG, IgA, and IgM antibodies by enzyme immunoassay(28), including a high heparin inhibition step as previously described(29). Reproducible optical density levels (ODs)>0.5 that were inhibited by high heparin (100 IU/mL) ≥ 40% were considered positive. All subjects gave their written informed consent in accordance with the Declaration of Helsinki. The study protocols were approved by the local ethics committee of Greifswald University. Genotyping, Imputation, and Quality Control BioVU (Discovery Population)—Samples were genotyped using the Illumina® HumanOmni1-QUAD and HumanOmni5-QUAD BeadChip genome-wide platforms. The discovery GWAS was performed on genotype data imputed from SNPs present on the
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 4
Author Manuscript
intersection of these platforms (n=730,803 SNPs). Before imputation, data were cleaned using the quality control (QC) pipeline developed by the eMERGE Genomics Working Group(30). Genomic imputation was then performed for samples using IMPUTE2(31) and data from the 1000 genomes project(32). Imputed SNPs were retained if their call probability was ≥90%. A total of 7,604,479 SNPs with mean call rate 99.0% were included in the imputed SNP dataset.
Author Manuscript
Samples were classified as being of European or African descent using ancestry informative markers (AIMs) input into STRUCTURE using Hapmap reference populations(33). Principal components (PC) analysis was also performed on AIMs using the EIGNESTRAT method implemented in PLINK(34, 35). To minimize confounding by population stratification, GWAS analyses were adjusted for PCs 1 and 2. Sensitivity analysis was performed in subjects with greater than 90% STRUCTURE-defined European ancestry. Classical HLA alleles were imputing from SNP data on genome-wide platforms using SNP2HLA with the Type 1 Diabetes Genetics Consortium (T1DGC) reference panel and a marker window size of 1000(36). Additional information regarding imputation, quality control, functional SNP analysis, and analysis of race/ethnicity is available in the Supplemental Materials. Replication of GWAS SNPs—Genotyping and quality control procedures in the replication cohort used a polymerase chain reaction-high resolution melting procedure and are described in detail in the Supplemental Materials. MAFs and the distribution of genotype groups were compared in HIT patients, Abpos and Abneg patients.
Author Manuscript
SNP Association with PF4/heparin Antibody Development—The SHIP samples were genotyped using the Affymetrix Genome-Wide Human SNP Array 6.0. Hybridization of genomic DNA was done in accordance with the manufacturer’s standard recommendations. Genotypes were determined using the Birdseed2 clustering algorithm. For QC purposes, several control samples where added. On the chip level, only subjects with a genotyping rate on QC probe sets (QC call rate) of at least 86% were included. Finally, all arrays had a sample call rate > 92%. The overall genotyping efficiency of the GWA was 98.6 %. Imputation of genotypes in SHIP was performed as described in the discovery cohort with IMPUTE-2 using the 1000Genomes v3 reference panel. Statistical Analysis
Author Manuscript
BioVU (Discovery Population)—Differences in characteristics between cases and controls were determined using chi square and t-tests. In GWAS analysis, single-locus tests of association were performed using logistic regression in an additive, dominant, and recessive genetic model adjusted for age, gender, type of heparin anticoagulant (UFH versus LMWH), and PCs 1 and 2 for all SNPs in a total of 951 individuals regardless of race/ethnic group. Significance was considered at a genome-wide-corrected significance level of 5.0×10−8. All statistical analyses were performed using PLINK(35) and SAS 9.3 (SAS, Cary, NC). GWAS SNPs were carried forward for replication based on strength of association, direct genotyping versus imputation, and putative functional consequences.
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 5
Author Manuscript
For the candidate gene analysis, the same statistical models were employed on a restricted set of genotyped SNPs in candidate genes. Candidate genes, including HLA class I and class II genes, were selected based on previous literature and HIT pathophysiology and are listed in Table 1. SNPs were identified from non-imputed data in candidate genes plus or minus 10 kilobases. SNPs in the region between HLA-A and HLA-DPB1 were not included unless located within 10 kilobases of a specified candidate gene. A total of 1,412 SNPs with minor allele frequencies (MAFs) greater than 0.01 were tested and we implemented a Bonferroni cutoff of alpha=3.54×10−5 (0.05/1,412). In an additional exploratory analysis, imputed classical HLA alleles were tested for association with HIT in an unadjusted additive model.
Author Manuscript
Replication of GWAS SNPs—To test the association of genotyped SNPs and HIT, multivariate logistic regression was used. Age and gender-adjusted odds ratios (OR) and 95% confidence intervals (CI) were generated with HIT defined as the outcome compared to both groups without HIT (Abpos and Abneg patients) and compared to Abneg patients in a recessive model. Results were considered significant with a two-sided alpha=0.05.
Author Manuscript
SNP Association with PF4/heparin Antibody Development—In the SHIP cohort, we assessed the association between GWAS-associated SNPs and PF4/heparin IgG titer levels (enzyme immunoassay OD levels) as well as formation of positive PF4/heparin antibody tests. Two IgG measurements were available for each patient and mean PF4/ heparin IgG ODs were analyzed as continuous variables after square root transformation. Association of SNPs with PF4/heparin IgG levels were determined using linear regression adjusted for age and gender with alpha=0.05. Effect sizes of SNPs are reported using coefficients (β) and standard errors (SEs). To evaluate the association between SNPs (exposure) and anti-PF4/heparin Ab status (dependent variable), conditional logistic regression models with fixed effects were used, adjusting for age and gender. ORs and CIs for positive anti-PF4/heparin status was tested for association with SNP genotype in a recessive model and determined by exact methods with median-unbiased estimates. Statistical analyses were performed in R.
Results Genome-Wide Association Study in the EMR Discovery Population
Author Manuscript
A total of 73 HIT cases were identified from BioVU and 67 of these cases were successfully genotyped after QC filters. The mean 4Ts score for HIT cases was 5.2 (standard deviation [SD] 0.8). (Table 2) The clinical characteristics of patients with HIT diagnosis are presented in Supplemental Table S2. We identified 884 controls from BioVU that were matched to HIT cases by exposure to UFH or LMWH and were successfully genotyped. The number of consecutive days of exposure to UFH or LMWH was significantly increased in HIT cases versus controls (9.7 [SD 6.0]) versus 3.4 [SD 4.0], p
Thromb Haemost. Author manuscript; available in PMC 2016 April 01. Published in final edited form as: Thromb Haemost. 2015 April ; 113(4): 772–781. doi:10.1160/TH14-08-0670.
A genome-wide association study of heparin-induced thrombocytopenia using an electronic medical record Jason H Karnes1, Robert M Cronin1, Jerome Rollin2, Alexander Teumer3, Claire Pouplard2, Christian M Shaffer1, Carmelo Blanquicett1, Erica A Bowton1, James D Cowan1, Jonathan D Mosley1, Sara L Van Driest1, Peter E Weeke1,4, Quinn S Wells1, Tamam Bakchoul3, Joshua C Denny1, Andreas Greinacher3, Yves Gruel2, and Dan M Roden1
Author Manuscript
1Vanderbilt 2UMR
University School of Medicine, Nashville, TN
CNRS 7292 & University Hospital of Tours, Tours Cedex, France
3University
of Greifswald, Greifswald, Germany
4Copenhagen
University Hospital, Gentofte, Denmark
Abstract
Author Manuscript
Heparin-induced thrombocytopenia (HIT) is an unpredictable, potentially catastrophic adverse effect of heparin treatment resulting from an immune response to platelet factor 4 (PF4)/heparin complexes. No genome-wide evaluations have been performed to identify potential genetic influences on HIT. Here, we performed a genome-wide association study (GWAS) and candidate gene study using HIT cases and controls identified using electronic medical records (EMRs) coupled to a DNA biobank and attempted to replicate GWAS associations in an independent cohort. We subsequently investigated influences of GWAS-associated single nucleotide polymorphisms (SNPs) on PF4/heparin antibodies in non-heparin treated individuals. In a recessive model, we observed significant SNP associations (OR 18.52 [6.33–54.23], p=3.18×10−9) with HIT near the T-Cell Death-Associated Gene 8 (TDAG8). These SNPs are in linkage disequilibrium with a missense TDAG8 SNP. TDAG8 SNPs trended toward an association with HIT in replication analysis (OR 5.71 [0.47–69.22], p=0.17), and the missense SNP was associated with PF4/heparin antibody levels and positive PF4/heparin antibodies in non-heparin treated patients (OR 3.09 [1.14–8.13], p=0.02). In the candidate gene study, SNPs at HLA-DRA were nominally associated with HIT (OR 0.25 [0.15–0.44], p=2.06×10−6). Further study of TDAG8 and HLA-DRA SNPs is warranted to assess their influence on the risk of developing HIT.
Author Manuscript
Corresponding Author: Dan M Roden, M.D., MRB4 1285B, 2215B Garland Avenue, Nashville, TN, 37232-0575, Telephone: 615-322-0067, Fax: 615-343-4522, [email protected]. Authorship Contributions For the discovery GWAS cohort, JHK, RMC, EAB, JDM, SVD, PEW, QSW, JCD, and DMR designed the study, interpreted results; JHK, RMC, CB, EAB, and JCD designed case and controls algorithms, reviewed electronic medical records, and adjudicated case and control subjects; JHK, CMS, EAB, JDM, and JCD performed statistical analysis and provided bioinformatic support. For the SHIP study, AG designed the study; AT, AG, and TB interpreted results; AT performed statistical analysis. For the independent replication population, YG designed the study; JR, CP, and YG interpreted results; JR and JHK performed statistical analysis. All authors helped to write the manuscript. Disclosure of Conflicts of Interest The authors declare no competing financial interests.
Karnes et al.
Page 2
Author Manuscript
Introduction Heparin-induced thrombocytopenia (HIT) is an antibody-mediated condition of platelet activation in patients receiving unfractionated heparin (UFH) and low molecular weight heparin (LMWH)(1). HIT develops in up to 0.5–5% of patients treated with heparin anticoagulants, has a greater than 30% mortality rate, and results in catastrophic thromboembolic complications, including life- and limb-threatening thrombosis(2–5). Further complicating the use of heparin anticoagulants, prevention of HIT-related thrombosis is currently possible only after manifestations of HIT are evident and the disease process has already begun(6, 7). The inability to predict HIT thus represents a liability with heparin administration and a priori identification of patients with a high HIT risk could allow for prevention of HIT.
Author Manuscript Author Manuscript
HIT is closely associated with the development of antibodies to complexes of heparin and platelet factor 4 (PF4), a protein normally found in the alpha granules of platelets(8). Although several studies have identified genetic polymorphisms such as the Fc receptor RIIA (FCGR2A) –H131R and Fc receptor RIIIA (FCGR3A) –F158V polymorphisms(9–13) associated with HIT, none have identified genetic risk factors that can reliably predict HIT. Other work has identified genomic predictors of diverse types of immune-mediated adverse drug reactions(14–16), suggesting that strong genetic influences on HIT risk might still be identified. This is of special interest, as HIT is an example for a distinct group of immune reactions, which fulfills neither the criteria of a primary nor the criteria of a typical secondary immune reaction(17–19). The clinical spectrum of HIT includes distinct stages of progression including 1) lack of PF4/heparin antibodies; 2) presence of PF4/heparin antibodies without thrombocytopenia; 3) PF4/heparin antibodies with thrombocytopenia; and 4) PF4/heparin antibodies with thrombocytopenia and thrombosis.(3, 20) As a result, single nucleotide polymorphisms (SNPs) may exert their effects only at distinct stages in the progression of HIT and different SNPs may be predictors of PF4/heparin antibody formation, thrombocytopenia, or thrombosis. The objective of this study was to identify genetic predictors of HIT using genome-wide association study (GWAS) and candidate gene approaches. We conducted this study using HIT cases and heparin-exposed controls identified in an electronic medical record (EMR) linked to a DNA repository. Associated SNPs from the HIT GWAS were genotyped in an independent replication cohort of HIT cases and heparin-treated controls. We also tested whether GWAS-identified SNPs were associated with PF4/heparin antibody levels in nonheparin treated patients.
Author Manuscript
Materials and Methods Study Populations BioVU (Discovery Population)—The discovery population was identified in BioVU, the Vanderbilt DNA databank that links DNA extracted from discarded blood samples to deidentified EMRs(21). This study was approved by the Institutional Review Board at Vanderbilt University as described previously(21). We developed an algorithm to identify individuals with a diagnosis of HIT in the EMR. The algorithm used ICD9 diagnostic codes,
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 3
Author Manuscript
lab results, and natural language processing(22, 23) to analyze narrative text in order to identify patients treated with UFH or LMWH who tested positive for PF4/heparin antibodies (optical density [OD] level >0.7 absorbance units [AU]). UFH or LMWH-treated controls were also identified in the EMR using a control algorithm that employed similar techniques. HIT cases were confirmed by chart review by at least two physician or pharmacist reviewers. HIT cases were scored with the 4Ts scoring system, which estimates the likelihood of HIT based on timing, magnitude, and potential causes of thrombocytopenia and presence of thrombosis and with a maximum 4Ts score of eight(6, 24). Additional information regarding the BioVU resource and case and control algorithms is available in the Supplemental Materials.
Author Manuscript Author Manuscript
Replication of GWAS SNPs—The patient group for replication has been previously described(11, 13) and additional details regarding the replication study population are detailed in the Supplemental Materials. In brief, the replication cohort was comprised of 94 Caucasian individuals with definite HIT who developed delayed-onset thrombocytopenia and/or a significant decrease in platelet count after a median duration of heparin treatment of 10 days. Both PF4-specific ELISA (HAT® GTI, Brookfield, WI) and serotonin release assay were positive in all patients. The first control group (Abneg) consisted of 178 patients who had undergone cardiopulmonary bypass surgery and received high doses of UFH but did not develop an immune response to heparin. The second control group (Abpos) consisted of 154 patients who had also undergone cardiopulmonary bypass surgery with heparin treatment. All had developed significant levels of PF4-specific antibodies but none had a positive serotonin release assay and none had significant abnormal evolution in platelet count in the post-operative period. Blood samples were collected after obtaining informed consent according to the Helsinki declaration principles. Both the University of Tours ethics committee and the Ministry of Research had previously approved collection of DNA from patients for genetic studies (Agreement number DC2008-308).
Author Manuscript
SNP Association with PF4/heparin Antibody Development—The Study of Health in Pomerania (SHIP) is a population-based project in West Pomerania, Germany that has been previously described(25–27). SHIP participants were recruited from population registries by written invitation regardless of previous anticoagulant treatment. Additional details regarding the SHIP population are available in the Supplemental Materials. The SHIP population comprised 4,308 participants who had sera collected and tested for anti-PF4/ heparin IgG, IgA, and IgM antibodies by enzyme immunoassay(28), including a high heparin inhibition step as previously described(29). Reproducible optical density levels (ODs)>0.5 that were inhibited by high heparin (100 IU/mL) ≥ 40% were considered positive. All subjects gave their written informed consent in accordance with the Declaration of Helsinki. The study protocols were approved by the local ethics committee of Greifswald University. Genotyping, Imputation, and Quality Control BioVU (Discovery Population)—Samples were genotyped using the Illumina® HumanOmni1-QUAD and HumanOmni5-QUAD BeadChip genome-wide platforms. The discovery GWAS was performed on genotype data imputed from SNPs present on the
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 4
Author Manuscript
intersection of these platforms (n=730,803 SNPs). Before imputation, data were cleaned using the quality control (QC) pipeline developed by the eMERGE Genomics Working Group(30). Genomic imputation was then performed for samples using IMPUTE2(31) and data from the 1000 genomes project(32). Imputed SNPs were retained if their call probability was ≥90%. A total of 7,604,479 SNPs with mean call rate 99.0% were included in the imputed SNP dataset.
Author Manuscript
Samples were classified as being of European or African descent using ancestry informative markers (AIMs) input into STRUCTURE using Hapmap reference populations(33). Principal components (PC) analysis was also performed on AIMs using the EIGNESTRAT method implemented in PLINK(34, 35). To minimize confounding by population stratification, GWAS analyses were adjusted for PCs 1 and 2. Sensitivity analysis was performed in subjects with greater than 90% STRUCTURE-defined European ancestry. Classical HLA alleles were imputing from SNP data on genome-wide platforms using SNP2HLA with the Type 1 Diabetes Genetics Consortium (T1DGC) reference panel and a marker window size of 1000(36). Additional information regarding imputation, quality control, functional SNP analysis, and analysis of race/ethnicity is available in the Supplemental Materials. Replication of GWAS SNPs—Genotyping and quality control procedures in the replication cohort used a polymerase chain reaction-high resolution melting procedure and are described in detail in the Supplemental Materials. MAFs and the distribution of genotype groups were compared in HIT patients, Abpos and Abneg patients.
Author Manuscript
SNP Association with PF4/heparin Antibody Development—The SHIP samples were genotyped using the Affymetrix Genome-Wide Human SNP Array 6.0. Hybridization of genomic DNA was done in accordance with the manufacturer’s standard recommendations. Genotypes were determined using the Birdseed2 clustering algorithm. For QC purposes, several control samples where added. On the chip level, only subjects with a genotyping rate on QC probe sets (QC call rate) of at least 86% were included. Finally, all arrays had a sample call rate > 92%. The overall genotyping efficiency of the GWA was 98.6 %. Imputation of genotypes in SHIP was performed as described in the discovery cohort with IMPUTE-2 using the 1000Genomes v3 reference panel. Statistical Analysis
Author Manuscript
BioVU (Discovery Population)—Differences in characteristics between cases and controls were determined using chi square and t-tests. In GWAS analysis, single-locus tests of association were performed using logistic regression in an additive, dominant, and recessive genetic model adjusted for age, gender, type of heparin anticoagulant (UFH versus LMWH), and PCs 1 and 2 for all SNPs in a total of 951 individuals regardless of race/ethnic group. Significance was considered at a genome-wide-corrected significance level of 5.0×10−8. All statistical analyses were performed using PLINK(35) and SAS 9.3 (SAS, Cary, NC). GWAS SNPs were carried forward for replication based on strength of association, direct genotyping versus imputation, and putative functional consequences.
Thromb Haemost. Author manuscript; available in PMC 2016 April 01.
Karnes et al.
Page 5
Author Manuscript
For the candidate gene analysis, the same statistical models were employed on a restricted set of genotyped SNPs in candidate genes. Candidate genes, including HLA class I and class II genes, were selected based on previous literature and HIT pathophysiology and are listed in Table 1. SNPs were identified from non-imputed data in candidate genes plus or minus 10 kilobases. SNPs in the region between HLA-A and HLA-DPB1 were not included unless located within 10 kilobases of a specified candidate gene. A total of 1,412 SNPs with minor allele frequencies (MAFs) greater than 0.01 were tested and we implemented a Bonferroni cutoff of alpha=3.54×10−5 (0.05/1,412). In an additional exploratory analysis, imputed classical HLA alleles were tested for association with HIT in an unadjusted additive model.
Author Manuscript
Replication of GWAS SNPs—To test the association of genotyped SNPs and HIT, multivariate logistic regression was used. Age and gender-adjusted odds ratios (OR) and 95% confidence intervals (CI) were generated with HIT defined as the outcome compared to both groups without HIT (Abpos and Abneg patients) and compared to Abneg patients in a recessive model. Results were considered significant with a two-sided alpha=0.05.
Author Manuscript
SNP Association with PF4/heparin Antibody Development—In the SHIP cohort, we assessed the association between GWAS-associated SNPs and PF4/heparin IgG titer levels (enzyme immunoassay OD levels) as well as formation of positive PF4/heparin antibody tests. Two IgG measurements were available for each patient and mean PF4/ heparin IgG ODs were analyzed as continuous variables after square root transformation. Association of SNPs with PF4/heparin IgG levels were determined using linear regression adjusted for age and gender with alpha=0.05. Effect sizes of SNPs are reported using coefficients (β) and standard errors (SEs). To evaluate the association between SNPs (exposure) and anti-PF4/heparin Ab status (dependent variable), conditional logistic regression models with fixed effects were used, adjusting for age and gender. ORs and CIs for positive anti-PF4/heparin status was tested for association with SNP genotype in a recessive model and determined by exact methods with median-unbiased estimates. Statistical analyses were performed in R.
Results Genome-Wide Association Study in the EMR Discovery Population
Author Manuscript
A total of 73 HIT cases were identified from BioVU and 67 of these cases were successfully genotyped after QC filters. The mean 4Ts score for HIT cases was 5.2 (standard deviation [SD] 0.8). (Table 2) The clinical characteristics of patients with HIT diagnosis are presented in Supplemental Table S2. We identified 884 controls from BioVU that were matched to HIT cases by exposure to UFH or LMWH and were successfully genotyped. The number of consecutive days of exposure to UFH or LMWH was significantly increased in HIT cases versus controls (9.7 [SD 6.0]) versus 3.4 [SD 4.0], p
