The evaluation of genetic tests

Scientific advances in genetics and molecular biology have been very successful in advancing our knowledge of biological mechanisms in health and disease, and in catalysing a variety of technological innovations. The number of genetic tests available has consequently increased exponentially over the last few years. Their development has not been accompanied by processes and systems to evaluate these tests in a proper and formal manner to establish their clinical validity and utility. A framework for the evaluation of genetic tests has been developed. This paper reviews the current practice of genetic test evaluation, highlighting the limitations and future challenges in this area of public health.     

A comparison of univariate and multivariate tests for genetic linkage

A variety of robust and model-dependent genetic linkage methods were applied to log transformed lipid levels from a large pedigree in which the LDL receptor defect has been shown to segregate by molecular biologic techniques. Application of the Haseman-Elston and a variance-components based test for linkage identified LDL and cholesterol as cosegregating with the marker C3, which is genetically linked to the LDL receptor defect. Consideration of lipid fractions as a multivariate response identified (0.723 × cholesterol) - (0.551 × triglycerides) as most strongly supporting evidence for linkage with C3. Subsequent segregation and linkage analyses provided support for an autosomal dominant major gene influencing either LDL or the function of cholesterol and triglycerides. Genetic linkage to LDL was only mildly supported, with a maximum lod score of 0.51 at a recombination fraction of θ = 0.33. Genetic linkage of the linear function to C3 was more strongly supported, with a maximum lod score of 1.69 at θ = 0.09. Bivariate analysis of clinical affection (with either type IIa or type IIb hyperlipidemia) and quantitative measures (LDL or the linear function) generally led to decreased lod scores, indicating, in this pedigree, loss of information when using clinical affection. © 1993 Wiley-Liss, Inc.     

Marketing genetic tests: empowerment or snake oil?

Genetic tests are currently being offered to the general public with little oversight and regulation as to which tests are allowed to be sold clinically and little control over the marketing and promotion of sales and use. This article provides discussion and data to indicate that the general public holds high opinions of genetic testing and that current media outlets for public education on genetic testing are not adequate to increase accurate knowledge of genetics. The authors argue that more regulation is needed to control and correct this problem in the United States.     

Robust trend tests for genetic association using matched case-control design

Trend tests for genetic association using a matched case-control design are studied, which allows for a variable number of controls per case. However, the tests depend on the scores based on the underlying genetic model, thus it may result in substantial loss of power when the model is misspecified. Since the mode of inheritance may be unknown for complex diseases, robust trend tests in matched case-control studies are developed. Simulation is conducted to compare the trend tests and the robust trend tests under various genetic models. The results are applied to detect candidate-gene association using an example from a case-control aetiologic study of sarcoidosis.     

Why disclosure of genetic tests for health insurance should be voluntary

Whether the disclosure of genetic (and non-genetic) information should be mandatory or voluntary is of concern. At present there are two opposing camps--the insurance industry argues for mandatory disclosure to avoid problems of adverse selection, and genetic interest groups argue for voluntary disclosure, fearing discrimination and public reluctance to have tests, thus missing out on any benefits from subsequent interventions. Policy has to balance these sides of the debate, as reducing adverse selection and encouraging maximum take-up of tests are both important for public health. We outline why voluntary disclosure is the optimal policy as it creates an incentive for people both to undergo testing and to disclose their result thus reducing the potential for adverse section.     

Comparison of statistics in association tests of genetic markers for survival outcomes

Computationally efficient statistical tests are needed in association testing of large scale genetic markers for survival outcomes. In this study, we explore several test statistics based on the Cox proportional hazards model for survival data. First, we consider the classical partial likelihood‐based Wald and score tests. A revised way to compute the score statistics is explored to improve the computational efficiency. Next, we propose a Cox–Snell residual‐based score test, which allows us to handle the controlling variables more conveniently. We also illustrated the incorporation of these three tests into a permutation procedure to adjust for the multiple testing. In addition, we examine a simulation‐based approach proposed by Lin (2005) to adjust for multiple testing. We presented the comparison of these four statistics in terms of type I error, power, family‐wise error rate, and computational efficiency under various scenarios via extensive simulation. Copyright © 2013 John Wiley & Sons, Ltd.     

Life insurance and genetic tests: Risks for insurers and society

The uncertainties surrounding the rapid advances being made in genetic testing and its subsequent use in the life insurance underwriting process has caused much consternation and apprehension throughout society and the insurance industry in recent years. The diagnostic and predictive power of genetic testing has introduced a new variable for insurers to select and classify applicants for life insurance. However, such use by insurers present many complex social, ethical and regulatory questions. This paper explores life insurers' use of genetic test results and the consequences for insurers should they be denied access to this actuarially relevant material. In addition, the risks posed for society from insurers' use of this information is analysed from the results of a survey and case study conducted in the Republic of Ireland in 2007. Given that the future use of genetic test results by life insurance companies depends heavily on public acceptance, it is important to understand reactions of the public, in relation to their risk perception and acceptance of this particular commercial use of genetic testing. To conclude, the paper briefly explores general international reaction to the risks posed by the use of genetic test results in underwriting life insurance applicants.     

Analyze multivariate phenotypes in genetic association studies by combining univariate association tests

Multivariate phenotypes are frequently encountered in genome‐wide association studies (GWAS). Such phenotypes contain more information than univariate phenotypes, but how to best exploit the information to increase the chance of detecting genetic variant of pleiotropic effect is not always clear. Moreover, when multivariate phenotypes contain a mixture of quantitative and qualitative measures, limited methods are applicable. In this paper, we first evaluated the approach originally proposed by O'Brien and by Wei and Johnson that combines the univariate test statistics and then we proposed two extensions to that approach. The original and proposed approaches are applicable to a multivariate phenotype containing any type of components including continuous, categorical and survival phenotypes, and applicable to samples consisting of families or unrelated samples. Simulation results suggested that all methods had valid type I error rates. Our extensions had a better power than O'Brien's method with heterogeneous means among univariate test statistics, but were less powerful than O'Brien's with homogeneous means among individual test statistics. All approaches have shown considerable increase in power compared to testing each component of a multivariate phenotype individually in some cases. We apply all the methods to GWAS of serum uric acid levels and gout with 550,000 single nucleotide polymorphisms in the Framingham Heart Study. Genet. Epidemiol. 34:444–454, 2010. © 2010 Wiley‐Liss, Inc.     

Likelihood ratio tests for maternal and fetal genetic effects on obstetric complications

It is well recognized that both maternal and fetal genes could contribute to susceptibility for obstetric complications. Logistic regression models are usually adopted to model the separate or joint action of maternal and fetal loci with case‐control data. The standard likelihood ratio tests (LRTs) can be used to test the significance of appropriate odds ratio parameters. This method, although simple to implement, fails to exploit a unique feature of genetic epidemiology studies of obstetric complications. Specifically, it does not take into consideration the correlation between the maternal and offspring genomes. We propose novel LRT that take advantage of this information by incorporating the fact that half of a child's genome is inherited from the mother. Our methods have substantially improved power for detecting marginal, main, and interactive maternal and fetal genotype effects, as evidenced by results from extensive simulation studies. We demonstrate our new methods by applying them to the analysis of data from a pilot study of preeclampsia. Genet. Epidemiol. 33:526–538, 2009. © 2009 Wiley‐Liss, Inc.     

Population genetic tests suggest that the epidemiologies of JCV and BKV are strikingly different

The JCV and BKV viruses have been used as markers for the study of human evolution by assuming that these viruses coevolved with their host. However, it is currently unclear whether the details of the population expansion of these viruses and humans agree. To study this in more detail, large numbers of complete genomes were used for population genetic tests to detect evidence for population expansion. Relative to the neutral expectation of no selective forces and no demographic changes, the JCV data set contained a striking excess of synonymous and non-synonymous mutations that occur only once in the data set. The same was found for non-synonymous mutations of BKV, but not at all for synonymous mutations of BKV. The different frequency spectra of mutations in JCV and BKV do not result from the inclusion of patients with clinical symptoms associated with BKV and JCV, such as nephropathy or progressive multifocal leucoencefalopathy, nor from the different numbers of genomes available for JCV and BKV. Instead, the distribution of unique mutations and population genetic models that use older mutation classes indicate a striking difference of the historical demographies of JCV and BKV with only the former virus exhibiting the evidence of demographic expansion. Our analyses expand on recent population genetic analyses that document a global population expansion of JCV by taking into account the impact of deleterious mutations and by comparing both human viruses. The striking difference between the demographics of BKV and JCV suggests that important aspects of their epidemiology remain to be discovered.     

Improved genetic association tests for an ordinal outcome representing the disease progression process

We are interested in detecting genetic variants that influence transition between discrete stages of a disease progression process, such as the natural history of progression to cervical cancer with the following four stages: (1) normal-human papillomavirus (HPV) exposed, (2) persistent infection with oncogenic HPV, (3) cervical intraepithelial neoplasia grades 2 or 3 (CIN2/3), and (4) cervical cancer. Standard statistical tests derived from the proportional odds model or polytomous regression model can be used to study this type of ordinal outcome. But these methods are either too sensitive to the proportion odds assumption or fail to take advantage of the restriction on the parameter space for the genetic variants. Two alternative tests, the maximum score test (MAX) and the adaptive P-value combination test (Adapt-P), are proposed with the aim of striking a balance between efficiency and robustness. A simulation study demonstrates that MAX and Adapt-P have the most robust performance among all considered tests under various realistic scenarios. As a demonstration, we applied the considered tests to a genetic association study of cervical cancer.     

Using needs-based frameworks for evaluating new technologies: An application to genetic tests

Given the multitude of newly available genetic tests in the face of limited healthcare budgets, the European Society of Human Genetics assessed how genetic services can be prioritized fairly. Using (health) benefit maximizing frameworks for this purpose has been criticized on the grounds that rather than maximization, fairness requires meeting claims (e.g. based on medical need) equitably. This study develops a prioritization score for genetic tests to facilitate equitable allocation based on need-based claims.It includes attributes representing health need associated with hereditary conditions (severity and progression), a genetic service's suitability to alleviate need (evidence of benefit and likelihood of positive result) and costs to meet the needs. A case study for measuring the attributes is provided and a suggestion is made how need-based claims can be quantified in a priority function. Attribute weights can be informed by data from discrete-choice experiments.Further work is needed to measure the attributes across the multitude of genetic tests and to determine appropriate weights. The priority score is most likely to be considered acceptable if developed within a decision process which meets criteria of procedural fairness and if the priority score is interpreted as “strength of recommendation” rather than a fixed cut-off value.     

On estimation of the variance in Cochran-Armitage trend tests for genetic association using case-control studies

The Cochran-Armitage trend test has been used in case-control studies for testing genetic association. As the variance of the test statistic is a function of unknown parameters, e.g. disease prevalence and allele frequency, it must be estimated. The usual estimator combining data for cases and controls assumes they follow the same distribution under the null hypothesis. Under the alternative hypothesis, however, the cases and controls follow different distributions. Thus, the power of the trend tests may be affected by the variance estimator used. In particular, the usual method combining both cases and controls is not an asymptotically unbiased estimator of the null variance when the alternative is true. Two different estimates of the null variance are available which are consistent under both the null and alternative hypotheses. In this paper, we examine sample size and small sample power performance of trend tests, which are optimal for three common genetic models as well as a robust trend test based on the three estimates of the variance and provide guidelines for choosing an appropriate test.     

The protective role of ascorbic acid on imazalil-induced genetic damage assessed by the cytogenetic tests

Ascorbic acid (AA), known as vitamin C, has important antioxidant and metabolic functions, making its incorporation into the human diet essential. On the other hand, imazalil (IMA), a commonly used fungicide in both agricultural and clinical domains is suspected to produce very serious toxic effects in vertebrates. In this study, the antigenotoxic effects of AA were studied against the genotoxic damage induced by IMA on cultured human lymphocytes using chromosomal aberration (CA) and sister chromatid exchange (SCE) as genetic end points. Human peripheral lymphocytes were treated in vitro with varying concentrations of AA (25, 50, 100, 200, and 400 μg/ml), tested in combination with IMA (336 mg/L). AA alone was not genotoxic and when combined with IMA treatment, reduced the frequencies of CAs and SCEs. A clear dose-dependent decrease in the genotoxic damage of IMA was observed, suggesting a genoprotective role of AA. In conclusion, the preventive role of AA in alleviating IMA-induced DNA damage was indicated for the first time in the present study.     

General score tests for associations of genetic markers with disease using cases and their parents

Association studies of genetic markers with disease play a critical role in the dissection of genetically complex traits because they are relatively easy to conduct and are useful for fine-scale mapping of genetic traits. The advantage of family-based controls has recently received much attention because spurious associations caused by population structure can be controlled for, and marker genotype information on diseased cases and their parents can be used to test the compound hypothesis of both linkage and linkage disequilibrium. However, debate still exists regarding the statistical methods of analysis. Herein are presented statistical methods to test for linkage (in the presence of linkage disequilibrium) between multiallelic genetic markers and disease when diseased subjects (cases) and their parents are sampled. Theoretical considerations for the development of general statistical tests are presented as well as asymptotic formulas to compute their power when planning a study. Furthermore, simulation results for nine specific statistics are used to contrast the power of these methods under different genetic mechanisms leading to disease (dominant vs. recessive, one vs. two high-risk alleles). These results demonstrate substantial gains in power for specific statistical tests designed to detect specified genetic mechanisms. However, without a priori knowledge of the likely genetic mechanism, it is desirable to rely on a fairly robust statistical method, robust so that power is not drastically lost when either dominant or recessive mechanisms are acting, and when either one or more than one marker alleles are associated with disease. Based on both theoretical and simulation results, a general score statistic, which generalizes the transmission/disequilibrium test, tends to offer sufficient power for a variety of genetic mechanisms, so that it is worth considering for broad use in studies which use genetic marker information from both diseased cases and their parents. © 1996 Wiley-Liss, Inc.     

Knowledge,attitudes and behavior of physicians regarding predictive genetic tests for breast and colorectal cancer

Background

Genetic testing for cancer susceptibility is an emerging technology in medicine. This study assessed the knowledge, attitudes and professional behavior of Italian physicians regarding the use of predictive genetic tests for breast and colorectal cancer, including the BRCA1/2 and APC tests.

Methods

A cross-sectional survey of a random sample of Italian physicians was performed in 2010 through a self-administered questionnaire.

Results

A response rate of 69.6% (1079 questionnaires) was achieved. A significant lack of knowledge was detected, particularly for APC testing. Less than half of the physicians agreed on the importance of efficacy and cost-effectiveness evidence in the selection of predictive genetic tests to be offered to the patients. Multiple logistic regression analyses showed that education had a positive influence on knowledge, attitudes and, to a lesser extent, professional use. The factor most strongly related to the physicians' use of genetic testing was patients requests for breast (odds ratio = 12.65; 95% confidence interval 7.77–20.59) or colorectal cancer tests (odds ratio = 7.02; 95% confidence interval 3.61–13.64). A high level of interest for specific training was reported by almost all physicians surveyed.

Conclusions

Targeted educational programs are needed to improve the expertise of physicians, and, ultimately, to enhance the appropriate use of genetic tests in clinical practice.