Bayesian predictive modeling for genomic based personalized treatment selection.
ABSTRACT: Efforts to personalize medicine in oncology have been limited by reductive characterizations of the intrinsically complex underlying biological phenomena. Future advances in personalized medicine will rely on molecular signatures that derive from synthesis of multifarious interdependent molecular quantities requiring robust quantitative methods. However, highly parameterized statistical models when applied in these settings often require a prohibitively large database and are sensitive to proper characterizations of the treatment-by-covariate interactions, which in practice are difficult to specify and may be limited by generalized linear models. In this article, we present a Bayesian predictive framework that enables the integration of a high-dimensional set of genomic features with clinical responses and treatment histories of historical patients, providing a probabilistic basis for using the clinical and molecular information to personalize therapy for future patients. Our work represents one of the first attempts to define personalized treatment assignment rules based on large-scale genomic data. We use actual gene expression data acquired from The Cancer Genome Atlas in the settings of leukemia and glioma to explore the statistical properties of our proposed Bayesian approach for personalizing treatment selection. The method is shown to yield considerable improvements in predictive accuracy when compared to penalized regression approaches.
Project description:Despite recent advancements in "omic" technologies, personalized medicine has not realized its fullest potential due to isolated and incomplete application of gene expression tools. In many instances, pharmacogenomics is being interchangeably used for personalized medicine, when actually it is one of the many facets of personalized medicine. Herein, we highlight key issues that are hampering the advancement of personalized medicine and highlight emerging predictive tools that can serve as a decision support mechanism for physicians to personalize treatments.
Project description:The central theme of personalized medicine is the premise that an individual's unique physiologic characteristics play a significant role in both disease vulnerability and in response to specific therapies. The major goals of personalized medicine are therefore to predict an individual's susceptibility to developing an illness, achieve accurate diagnosis, and optimize the most efficient and favorable response to treatment. The goal of achieving personalized medicine in psychiatry is a laudable one, because its attainment should be associated with a marked reduction in morbidity and mortality. In this review, we summarize an illustrative selection of studies that are laying the foundation towards personalizing medicine in major depressive disorder, bipolar disorder, and schizophrenia. In addition, we present emerging applications that are likely to advance personalized medicine in psychiatry, with an emphasis on novel biomarkers and neuroimaging.
Project description:Can genetic screening be used to personalize education for students? Genome-wide association studies (GWAS) screen an individual's DNA for specific variations in their genome, and how said variations relate to specific traits. The variations can then be assigned a corresponding weight and summed to produce polygenic scores (PGS) for given traits. Though first developed for disease risk, PGS is now used to predict educational achievement. Using a novel simulation method, this paper examines if PGS could advance screening in schools, a goal of personalized education. Results show limited potential benefits for using PGS to personalize education for individual students. However, further analysis shows PGS can be effectively used alongside progress monitoring measures to screen for learning disability risk. Altogether, PGS is not useful in personalizing education for every child but has potential utility when used simultaneously with additional screening tools to help determine which children may struggle academically.
Project description:<h4>Introduction</h4>A common approach to personalizing psychological interventions is the allocation of treatment modules to individual patients based on cut-off scores on questionnaires, which are mostly based on group studies. However, this way, intraindividual variation and temporal dynamics are not taken into account. Automated individual time series analyses are a possible solution, since these can identify the factors influencing the targeted symptom in a specific individual, and associated modules can be allocated accordingly. The aim of this study was to illustrate how automated individual time series analyses can be applied to personalize cognitive behavioral therapy for cancer-related fatigue in cancer survivors and how this procedure differs from allocating modules based on questionnaires.<h4>Methods</h4>This study was a case report series (<i>n</i> = 3). Patients completed ecological momentary assessments at the start of therapy, and after three treatment modules (approximately 14 weeks). Assessments were analyzed with AutoVAR, an R package that automates the process of finding optimal vector autoregressive models. The results informed the treatment plan.<h4>Results</h4>Three cases were described. From the ecological momentary assessments and automated time series analyses three individual treatment plans were constructed, in which the most important predictor for cancer-related fatigue was treated first. For two patients, this led to the treatment ending after the follow-up ecological momentary assessments. One patient continued treatment until six months, the standard treatment time in regular treatment. All three treatment plans differed from the treatment plans informed by questionnaire scores.<h4>Discussion</h4>This study is one of the first to apply time series analyses in systematically personalizing psychological treatment. An important strength of this approach is that it can be used for every modular cognitive behavioral intervention where each treatment module addresses specific maintaining factors. Whether or not personalized CBT is more efficacious than standard, non-personalized CBT remains to be determined in controlled studies comparing it to usual care.
Project description:Radiotherapy remains an important treatment modality in nearly two thirds of all cancers, including the primary curative or palliative treatment of breast cancer. Unfortunately, largely due to tumor heterogeneity, tumor radiotherapy response rates can vary significantly, even between patients diagnosed with the same tumor type. Although in recent years significant technological advances have been made in the way radiation can be precisely delivered to tumors, it is proving more difficult to personalize radiotherapy regimens based on cancer biology. Biomarkers that provide prognostic or predictive information regarding a tumor's intrinsic radiosensitivity or its response to treatment could prove valuable in helping to personalize radiation dosing, enabling clinicians to make decisions between different treatment options whilst avoiding radiation-induced toxicity in patients unlikely to gain therapeutic benefit. Studies have investigated numerous ways in which both patient and tumor radiosensitivities can be assessed. Tumor molecular profiling has been used to develop radiosensitivity gene signatures, while the assessment of specific intracellular or secreted proteins, including circulating tumor cells, exosomes and DNA, has been performed to identify prognostic or predictive biomarkers of radiation response. Finally, the investigation of biomarkers related to radiation-induced toxicity could provide another means by which radiotherapy could become personalized. In this review, we discuss studies that have used these methods to identify or develop prognostic/predictive signatures of radiosensitivity, and how such assays could be used in the future as a means of providing personalized radiotherapy.
Project description:Highlighting tumoral mutations is a key step in oncology for personalizing care. Considering the genetic heterogeneity in a tumor, software used for detecting mutations should clearly distinguish real tumor events of interest that could be predictive markers for personalized medicine from false positives. OutLyzer is a new variant-caller designed for the specific and sensitive detection of mutations for research and diagnostic purposes. It is based on statistic and local evaluation of sequencing background noise to highlight potential true positive variants. 130 previously genotyped patients were sequenced after enrichment by capturing the exons of 22 genes. Sequencing data were analyzed by HaplotypeCaller, LofreqStar, Varscan2 and OutLyzer. OutLyzer had the best sensitivity and specificity with a fixed limit of detection for all tools of 1% for SNVs and 2% for Indels. OutLyzer is a useful tool for detecting mutations of interest in tumors including low allele-frequency mutations, and could be adopted in standard practice for delivering targeted therapies in cancer treatment.
Project description:Colorectal cancer (CRC) is the third most common cancer diagnosed worldwide and is heterogeneous both morphologically and molecularly. In an era of personalized medicine, the greatest challenge is to predict individual response to therapy and distinguish patients likely to be cured with surgical resection of tumors and systemic therapy from those resistant or non-responsive to treatment. Patients would avoid futile treatments, including clinical trial regimes and ultimately this would prevent under- and over-treatment and reduce unnecessary adverse side effects. In this review, the potential of specific biomarkers will be explored to address two key questions-1) Can the prognosis of patients that will fare well or poorly be determined beyond currently recognized prognostic indicators? and 2) Can an individual patient's response to therapy be predicted and those who will most likely benefit from treatment/s be identified? Identifying and validating key prognostic and predictive biomarkers and an understanding of the underlying mechanisms of drug resistance and toxicity in CRC are important steps in order to personalize treatment. This review addresses recent data on biological prognostic and predictive biomarkers in CRC. In addition, patient cohorts most likely to benefit from currently available systemic treatments and/or targeted therapies are discussed in this review.
Project description:In recent years, there has been an emphasis on personalizing breast cancer treatment in order to avoid the debilitating side effects caused by broad-spectrum chemotherapeutic drug treatment. Development of personalized medicine requires the identification of proteins that are expressed by individual tumors. Herein, we reveal the identity of plasma membrane proteins that are overexpressed in estrogen receptor ?-positive, HER2-positive, and triple negative breast cancer cells. The proteins we identified are involved in maintaining protein structure, intracellular homeostasis, and cellular architecture; enhancing cell proliferation and invasion; and influencing cell migration. These proteins may be useful for breast cancer detection and/or treatment.
Project description:Early intervention in psychosis is crucial to improving patient response to treatment and the functional deficits that critically affect their long-term quality of life. Stratification tools are needed to personalize functional deficit prevention strategies at an early stage. In the present study, we applied topological tools to analyze symptoms of early psychosis patients, and detected a clear stratification of the cohort into three groups. One of the groups had a significantly better psychosocial outcome than the others after a 3-year clinical follow-up. This group was characterized by a metabolic profile indicative of an activated antioxidant response, while that of the groups with poorer outcome was indicative of oxidative stress. We replicated in a second cohort the finding that the three distinct clinical profiles at baseline were associated with distinct outcomes at follow-up, thus validating the predictive value of this new stratification. This approach could assist in personalizing treatment strategies.
Project description:Idiopathic pulmonary fibrosis (IPF), the most common fibrotic lung disease, is a chronic disease of unknown etiology with a very high mortality. Personalized medicine focuses on the use of the individual's molecular and 'omic' (i.e., genomic, epigenomic and proteomic) information to direct more efficient and cost-effective strategies for prevention, diagnosis, outcome prediction and treatment of diseases. In this review, we describe the use and promise of applying 'omic' technologies to the familial and sporadic forms of IPF as a means to personalize diagnosis and outcome prediction in IPF. The validation and implementation of such approaches will be crucial to personalize IPF patient care, prioritize lung transplant and stratify patients for drug studies, as well as, in the future, predict response to therapies as they emerge.