Jungnam Joo

Statistical design of personalized medicine interventions: The Clarification of Optimal Anticoagulation through Genetics (COAG) trial

BackgroundThere is currently much interest in pharmacogenetics: determining variation in genes that regulate drug effects, with a particular emphasis on improving drug safety and efficacy. The ability to determine such variation motivates the application of personalized drug therapies that utilize a patients genetic makeup to determine a safe and effective drug at the correct dose. To ascertain whether a genotype-guided drug therapy improves patient care, a personalized medicine intervention may be evaluated within the framework of a randomized controlled trial. The statistical design of this type of personalized medicine intervention requires special considerations: the distribution of relevant allelic variants in the study population; and whether the pharmacogenetic intervention is equally effective across subpopulations defined by allelic variants.MethodsThe statistical design of the Clarification of Optimal Anticoagulation through Genetics (COAG) trial serves as an illustrative example of a personalized medicine intervention that uses each subjects genotype information. The COAG trial is a multicenter, double blind, randomized clinical trial that will compare two approaches to initiation of warfarin therapy: genotype-guided dosing, the initiation of warfarin therapy based on algorithms using clinical information and genotypes for polymorphisms in CYP2C9 and VKORC1; and clinical-guided dosing, the initiation of warfarin therapy based on algorithms using only clinical information.ResultsWe determine an absolute minimum detectable difference of 5.49% based on an assumed 60% population prevalence of zero or multiple genetic variants in either CYP2C9 or VKORC1 and an assumed 15% relative effectiveness of genotype-guided warfarin initiation for those with zero or multiple genetic variants. Thus we calculate a sample size of 1238 to achieve a power level of 80% for the primary outcome. We show that reasonable departures from these assumptions may decrease statistical power to 65%.ConclusionsIn a personalized medicine intervention, the minimum detectable difference used in sample size calculations is not a known quantity, but rather an unknown quantity that depends on the genetic makeup of the subjects enrolled. Given the possible sensitivity of sample size and power calculations to these key assumptions, we recommend that they be monitored during the conduct of a personalized medicine intervention.Trial Registrationclinicaltrials.gov: NCT00839657

Conditional linkage and genome-wide association studies identify UGT1A1 as a major gene for anti-atherogenic serum bilirubin levels--the Framingham Heart Study.

OBJECTIVE AND METHODS Low bilirubin levels are significantly associated with cardiovascular diseases (CVD). In previous genome-wide linkage studies we identified a major locus on chromosome 2q harboring the candidate gene UDP-glucuronosyltransferase (UGT1A1). The activity of this enzyme is significantly influenced by a TA-repeat polymorphism in the promoter of the gene. In a prospective study individuals with genotype (TA)7/(TA)7 had significantly higher bilirubin levels and approximately one-third the risk of CVD as carriers of the wildtype (TA)6 allele. In the present study we performed a conditional linkage study to investigate whether this polymorphism explains the observed linkage peak and extended our analysis by a genome-wide association study on bilirubin levels in 1345 individuals. RESULTS After adjustment for the bilirubin variance explained by this polymorphism, the LOD score on chromosome 2q dropped from 3.8 to 0.4, demonstrating that this polymorphism explains the previous linkage result. For the genome-wide association study, the closest marker to UGT1A1 was in the top ranking SNPs. The association became even stronger when we considered the TA-repeat polymorphism in the analysis (p=2.68 x 10(-53)). Five other SNPs in other regions reached genome-wide significance without obvious connection to bilirubin metabolism. CONCLUSIONS Our studies suggest that UGT1A1 may be the major gene with strong effects on bilirubin levels and the TA-repeat polymorphism might be the key polymorphism within the gene controlling bilirubin levels. Since this polymorphism has a high frequency and a substantial impact on the development of CVD, the gene might be an important drug target.

Rationale and study design of the CardioGene Study: genomics of in-stent restenosis

BACKGROUND AND AIMS in-stent restenosis is a major limitation of stent therapy for atherosclerosis coronary artery disease. The CardioGene Study is an ongoing study of restenosis in bare mental stents (BMS) for the treatment of coronary artery disease. The overall goal is to understand the genetic determinants of the responses to vascular injury that result in the development of restenosis in some patients but not in others. Gene expression profiling at transcriptional and translational levels provides global assessment of gene activity after vascular injury and mechanistic insight. Furthermore, the delineation of genetic biomarkers would be of value in the clinical setting of risk-stratify patients prior to stent therapy. Prospective risk stratification would allow for the rational selection of specialized treatments against the development of in-stent restenosis (ISR), such as drug-eluting stents. SETTING Patients are enrolled at two sites in the US with high-volume cardiac catheterization facilities: the William Beaumont Hospital in Royal Oak, MI, USA, and the Mayo Clinic in Rochester, MN, USA. STUDY DESIGN Two complementary study designs are used to understand the molecular mechanisms of restenosis and the genetic biomarkers predictive of restenosis. First, 350 patients are enrolled prospectively at the time of stent implantation. Blood is sampled prior to stent placement and afterwards at 2 weeks and 6 months. The clinical outcome of restenosis is determined 6 and 12 months after stent placement. The primary outcome is clinical restenosis at 6 months. The major secondary outcome is clinical restenosis at 12 months. Second, a corollary case-control analysis will be carried out with the enrollment of an additional 250 cases with a history of recurrent restenosis after treatment with BMS. Controls for this analysis are derived from the prospective cohort. PATIENTS AND METHODS Consecutive patients presenting to the cardiac catheterization laboratory are screened, informed about the study and enrolled after signing the consent form. Enrollment has been completed for the prospective cohort, and enrollment of the additional group is ongoing. A standardized questionnaire is used to collect clinical data primarily through direct patient interview to assess medical history, medication use, functional status, family history, environmental factors, and social history. Further data are abstracted from the medical charts and catheterization reports. A total of 276 clinical variables are collected per individual at baseline, and 49 variables are collected at each of the 6- and 12-month follow-up visits. A Clinical Events Committee adjudicates clinical outcomes. Blood samples are processed at each clinical enrollment site using standardized operating procedures. From each blood sample, several aliquots are prepared and stored of peripheral blood mononuclear cells, granulocytes, platelets, serum, and plasma. Additionally, a portion of each patients leukocytes is cryopreserved for future cell-line creation. Samples are frozen and shipped to the National Heart, Lung and Blood Institute (NHLBI). Additional materials generated in the analysis of the samples at the NHLBI are frozen and stored, including isolated genomic DNA, total RNA, reverse transcribed cDNA libraries and labeled RNA hybridization mixtures used in microarray analysis. Per individual in the prospective cohort, high-quality transcript profiles of peripheral blood mononuclear cells at each time of blood sampling are obtained using Affymetrix U133A microarrays (Affymetrix, Santa Clara, CA, USA). Per chip, this yields 495,930 features per individual per time of sampling. This represents expression levels for 22,283 genes per patients oer time of blood sampling, including 14,500 well-characterized human genes. Proteomics of plasma is performed with multidimensional liquid chromatography and tandem mass spectrometry. Protein expression is examined similarly to mRNA expression as a measure of gene expression. Genotyping is performed in two manners. First, those genes showing differential expression at the levels of mRNA and protein are investigated using a candidate gene approach. Specific variants in known gene regulatory regions, such as promoters, are sought initially, as those variants may explain differences in expression level. Second, a genome-wide scan is used to identify genetic loci that are associated with ISR. Those regions identified are further examined for genes that show differential expression in the mRNA microarray profiling or proteomics investigations. These genes are finely investigated for candidate SNPs and other gene variants. Complementary genomic and proteomic approaches are expected to be robust. Integration of data sets is accomplished using a variety of informatics tools, organization of gene expression into functional pathways, and investigation of physical maps of up- and downregulated sets of genes. CONCLUSIONS The CardioGene Study is designed to understand ISR. Global gene and protein expression profiling define molecular phenotypes of patients. Well-defined clinical phenotypes will be paired with genomic data to define analyses aimed to achieve several goals. These include determining blood gene and protein expression in patients with ISR, investigating the genetic basis of ISR, developing predictive gene and protein biomarkers, and the identification of new targets for treatment.

Comparative analysis of treatment outcomes in patients with neuromyelitis optica spectrum disorder using multifaceted endpoints

Background: There is still an unmet need for comparative analyses of available treatment options for neuromyelitis optica spectrum disorder (NMOSD). Objective: We aimed to compare the efficacies of the immunosuppressants most commonly prescribed for patients with NMOSD using multifaceted endpoints. Methods: We conducted a retrospective analysis of treatment outcomes in 138 NMOSD patients treated with azathioprine, mycophenolate mofetil (MMF), or rituximab. The primary outcome measures were the annualized relapse rate (ARR), annualized severe relapse rate, time to first relapse, and time to first severe relapse. Results: A comparison of any relapse among the groups revealed that the azathioprine had a significantly higher risk of relapse relative to the rituximab (hazard ratio: 1.82; 95% CI: 1.1–3.1; p=0.03). A comparison of severe relapse among the groups revealed that the hazard ratios of severe relapse for the azathioprine and MMF relative to the rituximab were 11.66 (95% CI: 2.6–52.3; p=0.001) and 5.96 (95% CI: 1.0–35.1; p=0.048), respectively. The times to first relapse and first severe relapse were also significantly different among the treatment groups Conclusions: The present study showed that reductions in the risks of relapse and severe relapse differed among patients who were initially treated with azathioprine, MMF, and rituximab.

Prospective alpha allocation in the clarification of optimal anticoagulation through genetics (COAG) trial

Background The Clarification of Optimal Anticoagulation through Genetics (COAG) trial is a large, multicenter, double-blinded, randomized trial to determine whether use of a genotype-guided dosing algorithm (using clinical and genetic information) to initiate warfarin treatment will improve anticoagulation status when compared to a dosing algorithm using only clinical information. Purpose This article describes prospective alpha allocation and balanced alpha allocation for the design of the COAG trial. Methods The trial involves two possibly heterogeneous populations, which can be distinguished by the difference in warfarin dose as predicted by the two algorithms. A statistical approach is detailed, which allows an overall comparison as well as a comparison of the primary endpoint in the subgroup for which sufficiently different doses are predicted by the two algorithms. Methods of allocating alpha for these analyses are given — a prospective alpha allocation and allocating alpha so that the two analyses have equal power, which we call a ‘balanced alpha allocation.’ Results We show how to include an analysis of the primary endpoint in a subgroup as a co-primary analysis. Power can be improved by incorporating the correlation between the overall and subgroup analyses in a prospective alpha allocation approach. Balanced alpha allocation for the full cohort and subgroup tests to achieve the same desired power for both of the primary analyses is discussed in detail. Limitations In the COAG trial, it is impractical to stratify the randomization on subgroup membership because genetic information may not be available at the time of randomization. If imbalances in the treatment arms in the subgroup are found, they will need to be addressed. Conclusions The design of the COAG trial assures that the subgroup in which the largest treatment difference is expected is elevated to a co-primary analysis. Incorporating the correlation between the full cohort and the subgroup analyses provides an improvement in power for the subgroup comparison, and further improvement may be achieved via a balanced alpha allocation approach when the parameters involved in the sample size calculation are reasonably well estimated. Clinical Trials 2010; 7: 597—604. http://ctj.sagepub.com

Robust Tests in Genome-Wide Scans under Incomplete Linkage Disequilibrium

Under complete linkage disequilibrium (LD), robust tests often have greater power than Pearsons chi-square test and trend tests for the analysis of case-control genetic association studies. Robust statistics have been used in candidate-gene and genome-wide association studies (GWAS) when the genetic model is unknown. We consider here a more general incomplete LD model, and examine the impact of penetrances at the marker locus when the genetic models are defined at the disease locus. Robust statistics are then reviewed and their efficiency and robustness are compared through simulations in GWAS of 300,000 markers under the incomplete LD model. Applications of several robust tests to the Wellcome Trust Case-Control Consortium [Nature 447 (2007) 661--678] are presented.

Robust ranks of true associations in genome-wide case-control association studies

In whole-genome association studies, at the first stage, all markers are tested for association and their test statistics or p-values are ranked. At the second stage, some most significant markers are further analyzed by more powerful statistical methods. This helps reduce the number of hypotheses to be corrected for in multiple testing. Ranks of true associations in genome-wide scans using a single test statistic have been studied. In a case-control design for association, the trend test has been proposed. However, three different trend tests, optimal for the recessive, additive, and dominant models, respectively, are available for each marker. Because the true genetic model is unknown, we rank markers based on multiple test statistics or test statistics robust to model mis-specification. We studied this problem with application to Problem 3 of Genetic Analysis Workshop 15. An independent simulation study was also conducted to further evaluate the proposed procedure.

Joint Analysis of Binary and Quantitative Traits With Data Sharing and Outcome‐Dependent Sampling

We study the analysis of a joint association between a genetic marker with both binary (case‐control) and quantitative (continuous) traits, where the quantitative trait values are only available for the cases due to data sharing and outcome‐dependent sampling. Data sharing becomes common in genetic association studies, and the outcome‐dependent sampling is the consequence of data sharing, under which a phenotype of interest is not measured for some subgroup. The trend test (or Pearsons test) and F‐test are often, respectively, used to analyze the binary and quantitative traits. Because of the outcome‐dependent sampling, the usual F‐test can be applied using the subgroup with the observed quantitative traits. We propose a modified F‐test by also incorporating the genotype frequencies of the subgroup whose traits are not observed. Further, a combination of this modified F‐test and Pearsons test is proposed by Fishers combination of their P‐values as a joint analysis. Because of the correlation of the two analyses, we propose to use a Gamma (scaled chi‐squared) distribution to fit the asymptotic null distribution for the joint analysis. The proposed modified F‐test and the joint analysis can also be applied to test single trait association (either binary or quantitative trait). Through simulations, we identify the situations under which the proposed tests are more powerful than the existing ones. Application to a real dataset of rheumatoid arthritis is presented.

Intravoxel incoherent motion diffusion-weighted MR imaging of breast cancer: association with histopathological features and subtypes

OBJECTIVE To evaluate the associations between intravoxel incoherent motion (IVIM)-derived parameters and histopathological features and subtypes of breast cancer. METHODS Pre-operative MRI from 275 patients with unilateral breast cancer was analyzed. The apparent diffusion coefficient (ADC) and IVIM parameters [tissue diffusion coefficient (Dt), perfusion fraction (fp) and pseudodiffusion coefficient] were obtained from cancer and normal tissue using diffusion-weighted imaging with b-values of 0, 30, 70, 100, 150, 200, 300, 400, 500 and 800 s mm(-2). We then compared the IVIM parameters of tumours with different histopathological features and subtypes. RESULTS The ADC and Dt were lower and fp was higher in cancers than in normal tissues (p < 0.001). The Dt was lower in high Ki-67 cancer than in low Ki-67 cancer (p = 0.019), whereas ADC showed no significant difference (p = 0.309). Luminal B [human epidermal growth factor receptor 2 (HER2)-negative] cancer showed lower ADC (p = 0.003) and Dt (p = 0.001) than other types. CONCLUSION We found low tissue diffusivity in high Ki-67 cancer and luminal B (HER2-negative) cancer using IVIM imaging. ADVANCES IN KNOWLEDGE Low tissue diffusivity is more clearly shown in high Ki-67 tumours and luminal B (HER2-negative) tumours with the IVIM model.

Robust trend tests for genetic association in case-control studies using family data

We studied a trend test for genetic association between disease and the number of risk alleles using case-control data. When the data are sampled from families, this trend test can be adjusted to take into account the correlations among family members in complex pedigrees. However, the test depends on the scores based on the underlying genetic model and thus it may have substantial loss of power when the model is misspecified. Since the mode of inheritance will be unknown for complex diseases, we have developed two robust trend tests for case-control studies using family data. These robust tests have relatively good power for a class of possible genetic models. The trend tests and robust trend tests were applied to a dataset of Genetic Analysis Workshop 14 from the Collaborative Study on the Genetics of Alcoholism.