Citation Context

...arkers are assessed from a sample, creating inflation in type I errors. The Bonferroni adjustment for p-values is a common method to use for multiple comparisons, but when the number of comparisons is large, this 16 method can be too conservative. Letting k equal the number of comparisons and α equal the selected type I error rate, the Bonferroni method adjusts the error rate by α/k. For large values of k, the adjustment becomes radically small, reducing the chance that any hypothesis be rejected. Controlling for the false discovery rate (FDR) using a method proposed by Benjamini and Hochberg [28] is more practical for proteomic or genomic experiments comparing several potential biomarkers among patients [29, 30]. In this method, unadjusted p-values are first ordered from smallest to largest, and the rank is recorded. The adjustment to the error rate is calculated as α*m/k for each p-value, where α is the error rate, m is the rank, and k is the total number of comparisons made. The adjusted p-value is referred to as the Q-value in the Benjamini-Hochberg approach. This correction is more suitable for experiments with large k as it is less likely to overlook statistically significant res...

Citation Context

...to develop a non-invasive alternative to coronary angiography for detection of CAD. This study is unique, in that it applies to the specific population of symptomatic patients that are referred for cardiac catheterization. The proteomics analysis was conducted in two stages. In stage one, 239 samples (138 with CAD and 101 with normal coronary arteries) were assayed for 24 proteins. A scoring algorithm was generated off these 239 samples to measure the predictive ability of the proteins. This scoring algorithm was developed with a Monte Carlo optimization technique using a Metropolis algorithm [38] to derive the numerical coefficients, rather than the maximum likelihood approach used in logistic regression. 5-fold cross validation was used to estimate the bias in the ROC curve and to generate relevant sensitivity and specificity measurements. In the 21 following analysis, this process is duplicated using logistic regression rather than methods used to generate the previous scoring algorithm. In stage 2, assays were run on 120 additional samples (71 with CAD and 49 with normal coronary arteries) for validation of the algorithm, but for economic reasons, the researchers excluded assaying ...

Citation Context

... costly and invasive in nature. 2 Novel discoveries in biomarker research have a significant impact in the area of public health, providing alternative diagnostic methods for currently used invasive procedures, thus reducing the existing medical complications and economic burden of such procedures. 1.1 CORONARY ARTERY DISEASE The application of biomarker research related to coronary artery disease (CAD) is the primary focus of this thesis. In the United States, CAD is the leading cause of mortality, accounting for about one of every six deaths. In 2009, 386,324 deaths due to CAD were recorded [3]. The disease occurs when the coronary arteries harden and narrow, due to atherosclerosis, preventing oxygen from reaching the heart. The most common symptom of coronary artery disease is angina, or chest pain, but a patient may also experience fatigue, light-headedness, or shortness of breath. However, sometimes an individual will experience myocardial infarction (a heart attack) or immediate death without having any of the previous symptoms. The use of screening procedures allows for early detection of the disease so successful interventions can be performed to reduce chances of infarct or d...

Citation Context

...sed. If a linear assumption is validated, continuous variables will provide more powerful statistical results than their factored counterparts. Therefore, categorization of continuous variables is valid in an exploratory process, but final analysis should be conducted on the continuous form of the data. If the data is truly expected to be non-linear with respect to the log-odds of the outcome, some more advanced modeling techniques can be used to address the situations. 2.1.3.1 Fractional Polynomials The idea of fractional polynomials in regression is discussed in detail by Royston and Altman [23]. Fractional polynomials are used to transform continuous data to investigate improvements in model fit, compared to the straight line model , or in other words, the model without 14 a transformation applied to the covariate x [24]. Transformations are usually applied to continuous covariates in the event there is a nonlinear relationship with the dependent variable, in order to obtain better estimates for model coefficients. In regular polynomial regression, the independent variable, x, is taken to an nth power (i.e. x 2 , x 3 , x 4 , etc.) to describe a nonlinear relationship it may have wit...

Citation Context

...ut final analysis should be conducted on the continuous form of the data. If the data is truly expected to be non-linear with respect to the log-odds of the outcome, some more advanced modeling techniques can be used to address the situations. 2.1.3.1 Fractional Polynomials The idea of fractional polynomials in regression is discussed in detail by Royston and Altman [23]. Fractional polynomials are used to transform continuous data to investigate improvements in model fit, compared to the straight line model , or in other words, the model without 14 a transformation applied to the covariate x [24]. Transformations are usually applied to continuous covariates in the event there is a nonlinear relationship with the dependent variable, in order to obtain better estimates for model coefficients. In regular polynomial regression, the independent variable, x, is taken to an nth power (i.e. x 2 , x 3 , x 4 , etc.) to describe a nonlinear relationship it may have with the dependent variable (in the case of logistic regression, this would be the log-odds). The fractional polynomial method extends the current theory of polynomial regression by including negative and fractional powers for the cov...

Citation Context

...heterization procedure. Valid diagnostic tests should maintain very high levels of sensitivity. In order to characterize measures of sensitivity and specificity, receiver operating characteristic curves are usually generated. Sensitivity and Specificity Disease Status + - Test Result + True Positive (A) False Positive (B) Total, positive - False Negative (C) True Negative (D) Total, negative Total, disease Total, no disease 18 2.2.3 Receiver Operating Characteristic Curves The Receiver Operating Characteristic (ROC) curve is a way to visualize and gauge the performance of a set of classifiers [31]. ROC analysis is the principal method for evaluating sensitivity and specificity of a classifier and proves to be a useful tool in the evaluation of biomarkers [26]. In general, a measurement of the area under the ROC curve (AUC) is reported to compare the intrinsic accuracy of different tests [32]. The ROC curve is generated by plotting a set of thresholds according to their corresponding true-positive and false positive rates, or sensitivity and 1-specificity. When generating ROC curves for logistic models with several predictors, retrospective calculations of the ROC curve tend to give inf...

Citation Context

...ssification of subjects according to a dichotomous outcome. Many times, in health sciences, the goal is to differentiate those with and without a specific disease. Logistic regression has the ability to model the probability of disease, or any categorical outcome, and how the addition or subtraction of predictor variables affects that probability [18]. As opposed to linear regression methods, the logistic regression model estimates a linear function based on the log-odds of disease. This is because the relationship between the probability of the outcome and its predictors is usually nonlinear [19]. A unit increase in the predictor will have less of an impact when the probability of disease is close to 0 or 1, forming a logistic function, demonstrated by figure 2. Where equals the probability of disease and β0 and β represent numerical coefficients, the analytical form of this logit function is . The regression model can then be written out as ( ) Notice that the left side of the equation is the natural log of odds equation specified in section 2.1.1. Exponentiation both sides of the regression model then gives odds ratio estimates for the β’s. With some simple algebra, the regr...

Citation Context

...heart’s ventricles. Investigation into the biological pathways for atherosclerosis involving inflammation, plaque instability, thrombosis, and remodeling of the extracellular matrix, has identified several biomarkers associated with acute coronary syndromes [11]. An up-regulation of proteins responding to these biological processes can provide useful diagnostic information for cardiac complications. For example, troponin is widely one of the most popular biomarkers for heart disease, where elevated levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severity of disease better than previous attempts that have focused on single markers [13, 14, 15]. Previous research has supported moderate improvement in risk models of coronary disease by implementing multiple biomarkers among other populations [16]. Adaptation of this multi-marker approach may point to a specific set of markers that would imp...

Citation Context

...iction error. In this process, the data is split into k number of blocks. A predictive model is generated based on the K-1 partitions and used to score the Kth block (figure 5). This process is repeated until predicted probabilities have been generated for all the observations. 5-fold and 10-fold cross-validation are the most common forms of K-fold cross validation, using 80% of the data to score the other 20%, or 90% of the data to score the other 10%, respectively [36]. 19 External validation is one of the best ways to determine how well a logistic model can perform in clinical applications [37]. The data made available to the researchers or the statistician is used as a training set, where the predictive model is generated. The test set comes from additional experiments where the data has not been seen prior to developing the model. This can sometimes be emulated in an internal validation process where the data set is split into a training set and validation set prior to analysis. The validation data is then scored using the predictive model from the training set. In order to use this technique to estimate the prediction error of a model, the data set needs to be sizable enough to s...

Citation Context

...ide useful diagnostic information for cardiac complications. For example, troponin is widely one of the most popular biomarkers for heart disease, where elevated levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severity of disease better than previous attempts that have focused on single markers [13, 14, 15]. Previous research has supported moderate improvement in risk models of coronary disease by implementing multiple biomarkers among other populations [16]. Adaptation of this multi-marker approach may point to a specific set of markers that would improve risk assessment among symptomatic patients, as defined by this paper. 6 Identifying biomarkers in serum of symptomatic patients could lead to the development of a clinical assay to use as a diagnostic method for those patients with a high pretest probability and likelihood for CAD. Instead of referring patients for catheterization based on a c...

Citation Context

...igure 1 demonstrates the patient flow from experiencing symptoms to diagnosis. 1.1.2 Biomarkers and Coronary Artery Disease Cardiovascular disease is often accompanied with sources of inflammation and plaque instability, followed by thrombosis within the arterial regions of the heart. Resulting ischemia may be followed by remodeling of the heart’s ventricles. Investigation into the biological pathways for atherosclerosis involving inflammation, plaque instability, thrombosis, and remodeling of the extracellular matrix, has identified several biomarkers associated with acute coronary syndromes [11]. An up-regulation of proteins responding to these biological processes can provide useful diagnostic information for cardiac complications. For example, troponin is widely one of the most popular biomarkers for heart disease, where elevated levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severi...

Citation Context

...uable an experience without the support of these individuals. 1 1.0 INTRODUCTION Biological markers, more commonly referred to as “biomarkers,” refer to observable measurements derived from a patient that can be used to describe certain biological developments, including disease status, risk, or prognosis for that patient. According to an NIH working group, the definition of a biomarker is standardized to be “a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention [1]. These biomarkers can be classified into separate categories based on their clinical properties. For the purpose of this paper, the term biomarkers will be used to denote biological components that indicate disease status of an individual; they are disease biomarkers consisting of diagnostic properties. More specifically, this paper is interested in circulating biomarkers ascertained from advanced proteomics methods. Previously, biomarkers were commonly found to be simple physiological measurements, such as one’s blood pressure or heart rate, but have now evolved into complex imaging techniqu...

Citation Context

...ide useful diagnostic information for cardiac complications. For example, troponin is widely one of the most popular biomarkers for heart disease, where elevated levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severity of disease better than previous attempts that have focused on single markers [13, 14, 15]. Previous research has supported moderate improvement in risk models of coronary disease by implementing multiple biomarkers among other populations [16]. Adaptation of this multi-marker approach may point to a specific set of markers that would improve risk assessment among symptomatic patients, as defined by this paper. 6 Identifying biomarkers in serum of symptomatic patients could lead to the development of a clinical assay to use as a diagnostic method for those patients with a high pretest probability and likelihood for CAD. Instead of referring patients for catheterization based on a c...

Citation Context

...ore accurate odds-ratios for covariates within the model. 2.2 EVALUATION OF BIOMARKERS One of the main uses of biomarkers is to make a diagnosis more reliable, more rapidly, and inexpensive compared to existing methods [25]. However, proper evaluation of a biomarker needs to be assessed before it can be determined useful. Clinicians looking to implement biomarkers into their clinical practice are most concerned with diagnostic accuracy. Diagnostic accuracy refers to the ability of a biomarker to classify subjects into clinically relevant groups and is the general purpose of biomarker analyses [26]. That is, can a biomarker accurately distinguish between those patients that truly do have a disease and those that in fact do not have the disease? Several statistical tools can be used to measure diagnostic accuracy to determine if it should be used in clinical practice. 2.2.1 Multiple Comparisons Valid biomarkers should have a greater presence in the affected individuals than the unaffected individuals. [27]. Statistical comparative tests such as the Student’s t-test or Wilcoxon rank-sum test are commonly used to detect statistically significant differences in biomarker concentrations amon...

Citation Context

...thm had AUC measurements ranging from .82-.84 and specificities 45 of 43% - 58% at 95% sensitivity, while the predictive model generated here produced an AUC of .86 and 38% specificity achieved at 95% sensitivity. Concentrations of OPN produced the strongest relationship of any protein, which is a significant result of this analysis in itself, as OPN has been linked to heart disease through recent studies [39]. In regards to its relationship with coronary disease, OPN is a glycoprotein/cytokine of the extracellular matrix that has shown implications of roles in cardiac remodeling and fibrosis [40]. Other studies suggest that OPN is associated with calcification in coronary arteries [41, 42]. However, up-regulation of this protein has been linked to many other pathologies as well, including myeloma, multiple sclerosis, bone destruction, and cancer, preventing a direct association with heart disease to be made [43]. This makes it difficult to draw conclusions on the clinical usefulness of OPN as a biomarker for CAD without further understanding of the protein’s precise function. IL1β is a cytokine of the interleukin family known to be involved in inflammatory response. The inflammatory p...

Citation Context

...ide useful diagnostic information for cardiac complications. For example, troponin is widely one of the most popular biomarkers for heart disease, where elevated levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severity of disease better than previous attempts that have focused on single markers [13, 14, 15]. Previous research has supported moderate improvement in risk models of coronary disease by implementing multiple biomarkers among other populations [16]. Adaptation of this multi-marker approach may point to a specific set of markers that would improve risk assessment among symptomatic patients, as defined by this paper. 6 Identifying biomarkers in serum of symptomatic patients could lead to the development of a clinical assay to use as a diagnostic method for those patients with a high pretest probability and likelihood for CAD. Instead of referring patients for catheterization based on a c...

Citation Context

...1β is a cytokine of the interleukin family known to be involved in inflammatory response. The inflammatory process has been discussed as a significant mediator in the development of atherosclerosis, and the gene encoding the IL1β protein has been linked to coronary artery disease in Brazilian populations [44]. Apo-B100 is a lipid binding protein that is responsible for carrying low density lipoproteins (LDL, aka “bad cholesterol”) to tissues. ApoB100 has been declared as a more reliable indicator of risk of heart disease than LDL and a standardized assay for the protein can be used clinically [45]. Fibrinogen is known to play a role in blood clot formation and concentration levels have been notably increased in patients with cardiovascular disease. Thrombosis, the formation of blood clots, has been recognized as the basis for many cardiac cases involving myocardial infarction, ischemic death, and unstable angina pectoris [46]. 46 Proteomic biomarkers add predictive value to clinical characteristics The addition of a proteomic multi-marker panel to common clinical risk factors of heart disease greatly improved discrimination among symptomatic patients. For the sample of symptomatic pati...

Citation Context

...t a model with a set of predictors that produces the lowest deviance, indicating a closer fit to the model [19]. Figure 2. Graphical form of a logit function A unit increase in x has little impact on the probability of disease when the probability is close to 0 or 1. 0 .2 .4 .6 .8 1 P ro b a b ili ty o f D is e a s e 0 1 2 3 4 5 x Graph of the logit function 11 When combining several markers for prediction, which is becoming more and more popular with proteomic and genomic technologies, logistic regression serves as a useful tool for finding the best set of markers to use as a diagnostic tool [20]. However, using multiple signature of biomarkers for diagnostic tests leads to more difficulties in selecting the most predictive set of markers from a large list of candidates [21]. A simple approach to the variable selection process to obtain the most parsimonious model is forward stepwise selection, where variables are entered into the final model based on statistically significant relationships with the outcome. This differs from standard forward selection because variables that have entered the model in the stepwise method will also have potential to exit the model, based on the statisti...

Citation Context

...multi-marker panels suggested by LaFramboise et al, the 4-marker panel derived from the logistic regression process was similar. The best 4-marker panels derived by the study’s scoring algorithm had AUC measurements ranging from .82-.84 and specificities 45 of 43% - 58% at 95% sensitivity, while the predictive model generated here produced an AUC of .86 and 38% specificity achieved at 95% sensitivity. Concentrations of OPN produced the strongest relationship of any protein, which is a significant result of this analysis in itself, as OPN has been linked to heart disease through recent studies [39]. In regards to its relationship with coronary disease, OPN is a glycoprotein/cytokine of the extracellular matrix that has shown implications of roles in cardiac remodeling and fibrosis [40]. Other studies suggest that OPN is associated with calcification in coronary arteries [41, 42]. However, up-regulation of this protein has been linked to many other pathologies as well, including myeloma, multiple sclerosis, bone destruction, and cancer, preventing a direct association with heart disease to be made [43]. This makes it difficult to draw conclusions on the clinical usefulness of OPN as a bi...

Citation Context

...amber of the heart, where the dye is then released into the bloodstream [4]. Coronary angiography has been a highly efficient screening procedure for the detection of coronary stenosis and is regarded as the current gold standard for determining clinically significant CAD among symptomatic patients, but complications arising 3 from the procedure have been criticized [5]. Several common complications include arrhythmias (mostly attributed to anxiety about the procedure), bleeding and hematoma around the femoral artery region, allergic reactions to the injected dye, and anesthetic complications [6, 7]. Furthermore, a patient undergoing the catheterization procedure is exposed to localized x-ray radiation for an extended period of time, increasing the risk of cancer and other genetic effects [8]. The alarming rate of CAD has led to an increase in the number of cardiac catheterizations performed in hospitals, thus increasing the incidence of these complications. Almost half of the patients referred for catheterizations are found to have insignificant coronary lesions, and are unnecessarily exposed to procedural complications [9]. One alternative to the invasive procedure would include the id...

Citation Context

... reactions to the injected dye, and anesthetic complications [6, 7]. Furthermore, a patient undergoing the catheterization procedure is exposed to localized x-ray radiation for an extended period of time, increasing the risk of cancer and other genetic effects [8]. The alarming rate of CAD has led to an increase in the number of cardiac catheterizations performed in hospitals, thus increasing the incidence of these complications. Almost half of the patients referred for catheterizations are found to have insignificant coronary lesions, and are unnecessarily exposed to procedural complications [9]. One alternative to the invasive procedure would include the identification of biomarkers existing in a patient’s bloodstream. Biomarker discovery regarding CAD would reveal a safer, more pragmatic diagnostic procedure than coronary angiography with cardiac catheterization. 1.1.1 Symptomatic Patients Patients referred for catheterization come in to the emergency room (ER) or heart clinic showing symptoms of CAD. This cohort of patients excludes those that have experienced a cardiac event, such as myocardial infarction, who skip the ER and are immediately sent for percutaneous intervention. Fo...

Citation Context

...arate categories based on their clinical properties. For the purpose of this paper, the term biomarkers will be used to denote biological components that indicate disease status of an individual; they are disease biomarkers consisting of diagnostic properties. More specifically, this paper is interested in circulating biomarkers ascertained from advanced proteomics methods. Previously, biomarkers were commonly found to be simple physiological measurements, such as one’s blood pressure or heart rate, but have now evolved into complex imaging techniques and multi-marker genomic/proteomic panels [2]. This revolution allows researchers to interrogate blood and serum samples for potential markers that may not correspond with a patient’s sense of well-being, but are evidently affecting the disease status of an individual. This method of diagnosis is especially attractive in areas where the incidence of disease is high and current diagnostic methods are both costly and invasive in nature. 2 Novel discoveries in biomarker research have a significant impact in the area of public health, providing alternative diagnostic methods for currently used invasive procedures, thus reducing the existing ...

Citation Context

...ical imaging is used to detect a dye injected into the arteries by way of cardiac catheterization. This involves the insertion of a catheter, a thin and flexible tube, through a brachial or femoral artery and up to the aorta and chamber of the heart, where the dye is then released into the bloodstream [4]. Coronary angiography has been a highly efficient screening procedure for the detection of coronary stenosis and is regarded as the current gold standard for determining clinically significant CAD among symptomatic patients, but complications arising 3 from the procedure have been criticized [5]. Several common complications include arrhythmias (mostly attributed to anxiety about the procedure), bleeding and hematoma around the femoral artery region, allergic reactions to the injected dye, and anesthetic complications [6, 7]. Furthermore, a patient undergoing the catheterization procedure is exposed to localized x-ray radiation for an extended period of time, increasing the risk of cancer and other genetic effects [8]. The alarming rate of CAD has led to an increase in the number of cardiac catheterizations performed in hospitals, thus increasing the incidence of these complications....

Citation Context

...amber of the heart, where the dye is then released into the bloodstream [4]. Coronary angiography has been a highly efficient screening procedure for the detection of coronary stenosis and is regarded as the current gold standard for determining clinically significant CAD among symptomatic patients, but complications arising 3 from the procedure have been criticized [5]. Several common complications include arrhythmias (mostly attributed to anxiety about the procedure), bleeding and hematoma around the femoral artery region, allergic reactions to the injected dye, and anesthetic complications [6, 7]. Furthermore, a patient undergoing the catheterization procedure is exposed to localized x-ray radiation for an extended period of time, increasing the risk of cancer and other genetic effects [8]. The alarming rate of CAD has led to an increase in the number of cardiac catheterizations performed in hospitals, thus increasing the incidence of these complications. Almost half of the patients referred for catheterizations are found to have insignificant coronary lesions, and are unnecessarily exposed to procedural complications [9]. One alternative to the invasive procedure would include the id...

Citation Context

... as the current gold standard for determining clinically significant CAD among symptomatic patients, but complications arising 3 from the procedure have been criticized [5]. Several common complications include arrhythmias (mostly attributed to anxiety about the procedure), bleeding and hematoma around the femoral artery region, allergic reactions to the injected dye, and anesthetic complications [6, 7]. Furthermore, a patient undergoing the catheterization procedure is exposed to localized x-ray radiation for an extended period of time, increasing the risk of cancer and other genetic effects [8]. The alarming rate of CAD has led to an increase in the number of cardiac catheterizations performed in hospitals, thus increasing the incidence of these complications. Almost half of the patients referred for catheterizations are found to have insignificant coronary lesions, and are unnecessarily exposed to procedural complications [9]. One alternative to the invasive procedure would include the identification of biomarkers existing in a patient’s bloodstream. Biomarker discovery regarding CAD would reveal a safer, more pragmatic diagnostic procedure than coronary angiography with cardiac ca...

Citation Context

... regarding CAD would reveal a safer, more pragmatic diagnostic procedure than coronary angiography with cardiac catheterization. 1.1.1 Symptomatic Patients Patients referred for catheterization come in to the emergency room (ER) or heart clinic showing symptoms of CAD. This cohort of patients excludes those that have experienced a cardiac event, such as myocardial infarction, who skip the ER and are immediately sent for percutaneous intervention. For the patients received in the ER or heart clinic, an assessment of the individual is performed to determine the pretest probability of having CAD [10]. This may include looking at a patient’s medical history and existing clinical characteristics (obesity, smoking, age, etc.). If the pretest probability for CAD is low to intermediate, a non-invasive stress test and/or 4 Figure 1. Patient experience during the diagnostic process of coronary artery disease electrocardiogram (EKG) may also be taken into consideration to determine the likelihood of disease. Following the conclusion of an insignificant pretest probability, the patient may be 5 treated for his or her symptoms and sent home, being reassured there is insignificant evidence of diseas...

Citation Context

...levels of this protein points to the extent of injury to the heart during myocardial infarction [12]. However, there has been less success in regards to the clinical application of biomarkers to determine the degree of coronary artery disease among symptomatic patients. One suggestion is that a combination of protein changes in serum can address the severity of disease better than previous attempts that have focused on single markers [13, 14, 15]. Previous research has supported moderate improvement in risk models of coronary disease by implementing multiple biomarkers among other populations [16]. Adaptation of this multi-marker approach may point to a specific set of markers that would improve risk assessment among symptomatic patients, as defined by this paper. 6 Identifying biomarkers in serum of symptomatic patients could lead to the development of a clinical assay to use as a diagnostic method for those patients with a high pretest probability and likelihood for CAD. Instead of referring patients for catheterization based on a clinical assessment, stress test, and/or EKG, a less costly blood assay can be performed to filter out symptomatic patients that would otherwise be diagnos...

Citation Context

... little impact on the probability of disease when the probability is close to 0 or 1. 0 .2 .4 .6 .8 1 P ro b a b ili ty o f D is e a s e 0 1 2 3 4 5 x Graph of the logit function 11 When combining several markers for prediction, which is becoming more and more popular with proteomic and genomic technologies, logistic regression serves as a useful tool for finding the best set of markers to use as a diagnostic tool [20]. However, using multiple signature of biomarkers for diagnostic tests leads to more difficulties in selecting the most predictive set of markers from a large list of candidates [21]. A simple approach to the variable selection process to obtain the most parsimonious model is forward stepwise selection, where variables are entered into the final model based on statistically significant relationships with the outcome. This differs from standard forward selection because variables that have entered the model in the stepwise method will also have potential to exit the model, based on the statistical significance of their relationship with the outcome once new predictors are added. As previously discussed, the main assumption of logistic regression is that predictors within t...

Citation Context

...heavily skewed right distribution. A natural logarithmic transformation to the variable more closely approximates a normal curve (right). 0 .0 1 .0 2 .0 3 D e n s it y 0 50 100 150 200 OPN Distribution of OPN 0 .1 .2 .3 .4 D e n s it y 0 1 2 3 4 5 logOPN Distribution of ln(OPN) 13 If the log-linear assumption of logistic regression is violated, the predictive model will produce inaccurate estimates for the odds-ratios. Dichotomization or categorization of continuous predictors is commonly used in exploratory stages to fit logistic regression models when the linear relationship is questionable [22]. Categorization of variables into two or more categories is often done in medical research as a way to simplify the interpretation of odds ratios, creating regression models with step functions. Factoring by tertiles, quartiles, or quintiles is commonly seen in proteomic analysis when clinically relevant thresholds are not available [20]. This crude approach to categorization can be used to identify a log-linear relationship between the outcome and its predictor. Moving forward with the factored continuous variables may present complications for clinical interpretation. First, the cutpoints u...

Citation Context

... to modeling continuous covariates in an appropriate functional form, as opposed to categorization of these covariates which may present several disadvantages and statistically significant loss of information. Provided a nonlinear relationship exists between the dependent and independent 15 variables, fitting a logistic regression model with fractional polynomials will produce more accurate odds-ratios for covariates within the model. 2.2 EVALUATION OF BIOMARKERS One of the main uses of biomarkers is to make a diagnosis more reliable, more rapidly, and inexpensive compared to existing methods [25]. However, proper evaluation of a biomarker needs to be assessed before it can be determined useful. Clinicians looking to implement biomarkers into their clinical practice are most concerned with diagnostic accuracy. Diagnostic accuracy refers to the ability of a biomarker to classify subjects into clinically relevant groups and is the general purpose of biomarker analyses [26]. That is, can a biomarker accurately distinguish between those patients that truly do have a disease and those that in fact do not have the disease? Several statistical tools can be used to measure diagnostic accuracy ...

Citation Context

...with diagnostic accuracy. Diagnostic accuracy refers to the ability of a biomarker to classify subjects into clinically relevant groups and is the general purpose of biomarker analyses [26]. That is, can a biomarker accurately distinguish between those patients that truly do have a disease and those that in fact do not have the disease? Several statistical tools can be used to measure diagnostic accuracy to determine if it should be used in clinical practice. 2.2.1 Multiple Comparisons Valid biomarkers should have a greater presence in the affected individuals than the unaffected individuals. [27]. Statistical comparative tests such as the Student’s t-test or Wilcoxon rank-sum test are commonly used to detect statistically significant differences in biomarker concentrations among disease categories. In most research studies, multiple biomarkers are assessed from a sample, creating inflation in type I errors. The Bonferroni adjustment for p-values is a common method to use for multiple comparisons, but when the number of comparisons is large, this 16 method can be too conservative. Letting k equal the number of comparisons and α equal the selected type I error rate, the Bonferroni metho...

Citation Context

...a common method to use for multiple comparisons, but when the number of comparisons is large, this 16 method can be too conservative. Letting k equal the number of comparisons and α equal the selected type I error rate, the Bonferroni method adjusts the error rate by α/k. For large values of k, the adjustment becomes radically small, reducing the chance that any hypothesis be rejected. Controlling for the false discovery rate (FDR) using a method proposed by Benjamini and Hochberg [28] is more practical for proteomic or genomic experiments comparing several potential biomarkers among patients [29, 30]. In this method, unadjusted p-values are first ordered from smallest to largest, and the rank is recorded. The adjustment to the error rate is calculated as α*m/k for each p-value, where α is the error rate, m is the rank, and k is the total number of comparisons made. The adjusted p-value is referred to as the Q-value in the Benjamini-Hochberg approach. This correction is more suitable for experiments with large k as it is less likely to overlook statistically significant results that may be masked by more conservative approaches. 2.2.2 Sensitivity and Specificity Sensitivity and specificity...

Citation Context

...otting a set of thresholds according to their corresponding true-positive and false positive rates, or sensitivity and 1-specificity. When generating ROC curves for logistic models with several predictors, retrospective calculations of the ROC curve tend to give inflated assessments of score performance [33]. A predictive model will almost always fit better to the data it was constructed around than when applied to future data. This is because the ROC curve is generated by a process of resubstitution, where the model is constructed using the available data, and then validated on the same data [34]. Some more advanced techniques for handling the upward bias of the ROC curve have been discussed [35], but among the simplest and most common methods are k-fold cross validation and external validation. K-fold cross-validation is an internal validation method for estimating the prediction error. In this process, the data is split into k number of blocks. A predictive model is generated based on the K-1 partitions and used to score the Kth block (figure 5). This process is repeated until predicted probabilities have been generated for all the observations. 5-fold and 10-fold cross-validation a...

Citation Context

...itivity, while the predictive model generated here produced an AUC of .86 and 38% specificity achieved at 95% sensitivity. Concentrations of OPN produced the strongest relationship of any protein, which is a significant result of this analysis in itself, as OPN has been linked to heart disease through recent studies [39]. In regards to its relationship with coronary disease, OPN is a glycoprotein/cytokine of the extracellular matrix that has shown implications of roles in cardiac remodeling and fibrosis [40]. Other studies suggest that OPN is associated with calcification in coronary arteries [41, 42]. However, up-regulation of this protein has been linked to many other pathologies as well, including myeloma, multiple sclerosis, bone destruction, and cancer, preventing a direct association with heart disease to be made [43]. This makes it difficult to draw conclusions on the clinical usefulness of OPN as a biomarker for CAD without further understanding of the protein’s precise function. IL1β is a cytokine of the interleukin family known to be involved in inflammatory response. The inflammatory process has been discussed as a significant mediator in the development of atherosclerosis, and ...

Citation Context

...his analysis in itself, as OPN has been linked to heart disease through recent studies [39]. In regards to its relationship with coronary disease, OPN is a glycoprotein/cytokine of the extracellular matrix that has shown implications of roles in cardiac remodeling and fibrosis [40]. Other studies suggest that OPN is associated with calcification in coronary arteries [41, 42]. However, up-regulation of this protein has been linked to many other pathologies as well, including myeloma, multiple sclerosis, bone destruction, and cancer, preventing a direct association with heart disease to be made [43]. This makes it difficult to draw conclusions on the clinical usefulness of OPN as a biomarker for CAD without further understanding of the protein’s precise function. IL1β is a cytokine of the interleukin family known to be involved in inflammatory response. The inflammatory process has been discussed as a significant mediator in the development of atherosclerosis, and the gene encoding the IL1β protein has been linked to coronary artery disease in Brazilian populations [44]. Apo-B100 is a lipid binding protein that is responsible for carrying low density lipoproteins (LDL, aka “bad cholester...

Citation Context

...yeloma, multiple sclerosis, bone destruction, and cancer, preventing a direct association with heart disease to be made [43]. This makes it difficult to draw conclusions on the clinical usefulness of OPN as a biomarker for CAD without further understanding of the protein’s precise function. IL1β is a cytokine of the interleukin family known to be involved in inflammatory response. The inflammatory process has been discussed as a significant mediator in the development of atherosclerosis, and the gene encoding the IL1β protein has been linked to coronary artery disease in Brazilian populations [44]. Apo-B100 is a lipid binding protein that is responsible for carrying low density lipoproteins (LDL, aka “bad cholesterol”) to tissues. ApoB100 has been declared as a more reliable indicator of risk of heart disease than LDL and a standardized assay for the protein can be used clinically [45]. Fibrinogen is known to play a role in blood clot formation and concentration levels have been notably increased in patients with cardiovascular disease. Thrombosis, the formation of blood clots, has been recognized as the basis for many cardiac cases involving myocardial infarction, ischemic death, and ...

Citation Context

...ding protein that is responsible for carrying low density lipoproteins (LDL, aka “bad cholesterol”) to tissues. ApoB100 has been declared as a more reliable indicator of risk of heart disease than LDL and a standardized assay for the protein can be used clinically [45]. Fibrinogen is known to play a role in blood clot formation and concentration levels have been notably increased in patients with cardiovascular disease. Thrombosis, the formation of blood clots, has been recognized as the basis for many cardiac cases involving myocardial infarction, ischemic death, and unstable angina pectoris [46]. 46 Proteomic biomarkers add predictive value to clinical characteristics The addition of a proteomic multi-marker panel to common clinical risk factors of heart disease greatly improved discrimination among symptomatic patients. For the sample of symptomatic patients studied, the data resembled the current status quo, where the use of clinical characteristics provides zero ability to detect a patient who does not need to undergo catheterization. None of the clinical characteristics had adequate discriminatory power before the inclusion of proteins. The added discriminatory power of OPN, alon...