Bayesian generalized linear models for meta-analysis of diagnostic tests

With the recognition of the importance of evidence-based medicine, there is an emerging need for methods to systematically synthesize available data. Specifically, methods to provide accurate estimates of test characteristics for diagnostic tests are needed to help physicians make better clinical decisions. To provide more flexible approaches for meta-analysis of diagnostic tests, we developed three Bayesian generalized linear models. Two of these models, a bivariate normal and a binomial model, analyzed pairs of sensitivity and specificity values while incorporating the correlation between these two outcome variables. Noninformative independent uniform priors were used for the variance of sensitivity, specificity and correlation. We also applied an inverse Wishart prior to check the sensitivity of the results. The third model was a multinomial model where the test results were modeled as multinomial random variables. All three models can include specific imaging techniques as covariates in order to compare performance. Vague normal priors were assigned to the coefficients of the covariates. The computations were carried out using the 'Bayesian inference using Gibbs sampling' implementation of Markov chain Monte Carlo techniques. We investigated the properties of the three proposed models through extensive simulation studies. We also applied these models to a previously published meta-analysis dataset on cervical cancer as well as to an unpublished melanoma dataset. In general, our findings show that the point estimates of sensitivity and specificity were consistent among Bayesian and frequentist bivariate normal and binomial models. However, in the simulation studies, the estimates of the correlation coefficient from Bayesian bivariate models are not as good as those obtained from frequentist estimation regardless of which prior distribution was used for the covariance matrix. The Bayesian multinomial model consistently underestimated the sensitivity and specificity regardless of the sample size and correlation coefficient. In conclusion, the Bayesian bivariate binomial model provides the most flexible framework for future applications because of its following strengths: (1) it facilitates direct comparison between different tests; (2) it captures the variability in both sensitivity and specificity simultaneously as well as the intercorrelation between the two; and (3) it can be directly applied to sparse data without ad hoc correction. ^

A multivariate frailty model for disease recurrences and survival

A multivariate frailty hazard model is developed for joint-modeling of three correlated time-to-event outcomes: (1) local recurrence, (2) distant recurrence, and (3) overall survival. The term frailty is introduced to model population heterogeneity. The dependence is modeled by conditioning on a shared frailty that is included in the three hazard functions. Independent variables can be included in the model as covariates. The Markov chain Monte Carlo methods are used to estimate the posterior distributions of model parameters. The algorithm used in present application is the hybrid Metropolis-Hastings algorithm, which simultaneously updates all parameters with evaluations of gradient of log posterior density. The performance of this approach is examined based on simulation studies using Exponential and Weibull distributions. We apply the proposed methods to a study of patients with soft tissue sarcoma, which motivated this research. Our results indicate that patients with chemotherapy had better overall survival with hazard ratio of 0.242 (95% CI: 0.094 - 0.564) and lower risk of distant recurrence with hazard ratio of 0.636 (95% CI: 0.487 - 0.860), but not significantly better in local recurrence with hazard ratio of 0.799 (95% CI: 0.575 - 1.054). The advantages and limitations of the proposed models, and future research directions are discussed. ^

A Bayesian analysis for spatial processes with application to mapping

In geographical epidemiology, maps of disease rates and disease risk provide a spatial perspective for researching disease etiology. For rare diseases or when the population base is small, the rate and risk estimates may be unstable. Empirical Bayesian (EB) methods have been used to spatially smooth the estimates by permitting an area estimate to "borrow strength" from its neighbors. Such EB methods include the use of a Gamma model, of a James-Stein estimator, and of a conditional autoregressive (CAR) process. A fully Bayesian analysis of the CAR process is proposed. One advantage of this fully Bayesian analysis is that it can be implemented simply by using repeated sampling from the posterior densities. Use of a Markov chain Monte Carlo technique such as Gibbs sampler was not necessary. Direct resampling from the posterior densities provides exact small sample inferences instead of the approximate asymptotic analyses of maximum likelihood methods (Clayton & Kaldor, 1987). Further, the proposed CAR model provides for covariates to be included in the model. A simulation demonstrates the effect of sample size on the fully Bayesian analysis of the CAR process. The methods are applied to lip cancer data from Scotland, and the results are compared. ^

Bayesian and survival model for tumor relapse status and disease-specific survival, with applications to breast cancer

Breast cancer is the most common non-skin cancer and the second leading cause of cancer-related death in women in the United States. Studies on ipsilateral breast tumor relapse (IBTR) status and disease-specific survival will help guide clinic treatment and predict patient prognosis.^ After breast conservation therapy, patients with breast cancer may experience breast tumor relapse. This relapse is classified into two distinct types: true local recurrence (TR) and new ipsilateral primary tumor (NP). However, the methods used to classify the relapse types are imperfect and are prone to misclassification. In addition, some observed survival data (e.g., time to relapse and time from relapse to death)are strongly correlated with relapse types. The first part of this dissertation presents a Bayesian approach to (1) modeling the potentially misclassified relapse status and the correlated survival information, (2) estimating the sensitivity and specificity of the diagnostic methods, and (3) quantify the covariate effects on event probabilities. A shared frailty was used to account for the within-subject correlation between survival times. The inference was conducted using a Bayesian framework via Markov Chain Monte Carlo simulation implemented in softwareWinBUGS. Simulation was used to validate the Bayesian method and assess its frequentist properties. The new model has two important innovations: (1) it utilizes the additional survival times correlated with the relapse status to improve the parameter estimation, and (2) it provides tools to address the correlation between the two diagnostic methods conditional to the true relapse types.^ Prediction of patients at highest risk for IBTR after local excision of ductal carcinoma in situ (DCIS) remains a clinical concern. The goals of the second part of this dissertation were to evaluate a published nomogram from Memorial Sloan-Kettering Cancer Center, to determine the risk of IBTR in patients with DCIS treated with local excision, and to determine whether there is a subset of patients at low risk of IBTR. Patients who had undergone local excision from 1990 through 2007 at MD Anderson Cancer Center with a final diagnosis of DCIS (n=794) were included in this part. Clinicopathologic factors and the performance of the Memorial Sloan-Kettering Cancer Center nomogram for prediction of IBTR were assessed for 734 patients with complete data. Nomogram for prediction of 5- and 10-year IBTR probabilities were found to demonstrate imperfect calibration and discrimination, with an area under the receiver operating characteristic curve of .63 and a concordance index of .63. In conclusion, predictive models for IBTR in DCIS patients treated with local excision are imperfect. Our current ability to accurately predict recurrence based on clinical parameters is limited.^ The American Joint Committee on Cancer (AJCC) staging of breast cancer is widely used to determine prognosis, yet survival within each AJCC stage shows wide variation and remains unpredictable. For the third part of this dissertation, biologic markers were hypothesized to be responsible for some of this variation, and the addition of biologic markers to current AJCC staging were examined for possibly provide improved prognostication. The initial cohort included patients treated with surgery as first intervention at MDACC from 1997 to 2006. Cox proportional hazards models were used to create prognostic scoring systems. AJCC pathologic staging parameters and biologic tumor markers were investigated to devise the scoring systems. Surveillance Epidemiology and End Results (SEER) data was used as the external cohort to validate the scoring systems. Binary indicators for pathologic stage (PS), estrogen receptor status (E), and tumor grade (G) were summed to create PS+EG scoring systems devised to predict 5-year patient outcomes. These scoring systems facilitated separation of the study population into more refined subgroups than the current AJCC staging system. The ability of the PS+EG score to stratify outcomes was confirmed in both internal and external validation cohorts. The current study proposes and validates a new staging system by incorporating tumor grade and ER status into current AJCC staging. We recommend that biologic markers be incorporating into revised versions of the AJCC staging system for patients receiving surgery as the first intervention.^ Chapter 1 focuses on developing a Bayesian method to solve misclassified relapse status and application to breast cancer data. Chapter 2 focuses on evaluation of a breast cancer nomogram for predicting risk of IBTR in patients with DCIS after local excision gives the statement of the problem in the clinical research. Chapter 3 focuses on validation of a novel staging system for disease-specific survival in patients with breast cancer treated with surgery as the first intervention. ^

Longitudinal categorical data analysis: A continuous -time Markov chain approach

In this dissertation, we propose a continuous-time Markov chain model to examine the longitudinal data that have three categories in the outcome variable. The advantage of this model is that it permits a different number of measurements for each subject and the duration between two consecutive time points of measurements can be irregular. Using the maximum likelihood principle, we can estimate the transition probability between two time points. By using the information provided by the independent variables, this model can also estimate the transition probability for each subject. The Monte Carlo simulation method will be used to investigate the goodness of model fitting compared with that obtained from other models. A public health example will be used to demonstrate the application of this method. ^

BENCHMARKING AND IMPLEMENTATION OF A NEW INDEPENDENT MONTE CARLO DOSE CALCULATION QUALITY ASSURANCE AUDIT TOOL FOR CLINICAL TRIALS

Introduction Commercial treatment planning systems employ a variety of dose calculation algorithms to plan and predict the dose distributions a patient receives during external beam radiation therapy. Traditionally, the Radiological Physics Center has relied on measurements to assure that institutions participating in the National Cancer Institute sponsored clinical trials administer radiation in doses that are clinically comparable to those of other participating institutions. To complement the effort of the RPC, an independent dose calculation tool needs to be developed that will enable a generic method to determine patient dose distributions in three dimensions and to perform retrospective analysis of radiation delivered to patients who enrolled in past clinical trials. Methods A multi-source model representing output for Varian 6 MV and 10 MV photon beams was developed and evaluated. The Monte Carlo algorithm, know as the Dose Planning Method (DPM), was used to perform the dose calculations. The dose calculations were compared to measurements made in a water phantom and in anthropomorphic phantoms. Intensity modulated radiation therapy and stereotactic body radiation therapy techniques were used with the anthropomorphic phantoms. Finally, past patient treatment plans were selected and recalculated using DPM and contrasted against a commercial dose calculation algorithm. Results The multi-source model was validated for the Varian 6 MV and 10 MV photon beams. The benchmark evaluations demonstrated the ability of the model to accurately calculate dose for the Varian 6 MV and the Varian 10 MV source models. The patient calculations proved that the model was reproducible in determining dose under similar conditions described by the benchmark tests. Conclusions The dose calculation tool that relied on a multi-source model approach and used the DPM code to calculate dose was developed, validated, and benchmarked for the Varian 6 MV and 10 MV photon beams. Several patient dose distributions were contrasted against a commercial algorithm to provide a proof of principal to use as an application in monitoring clinical trial activity.

A 3-D CT assisted Monte Carlo evaluation of intracavitary brachytherapy implants

Intracavitary brachytherapy (ICB) combined with external beam irradiation for treatment of cervical cancer is highly successful in achieving local control. The M.D. Anderson Cancer Center employs Fletcher Suit Delclos (FSD) applicators. FSD applicators contain shields to limit dose to critical structures. Dosimetric evaluation of ICB implants is limited to assessing dose at reference points. These points serve as surrogates for treatment intensity and critical structure dose. Several studies have mentioned that the ICRU38 reference points inadequately characterize the dose distribution. Also, the ovoid shields are rarely considered in dosimetry. ^ The goal of this dissertation was to ascertain the influence of the ovoid shields on patient dose distributions. Monte Carlo dosimetry (MCD) was applied to patient computed tomography(CT) scans. These data were analyzed to determine the effect of the shields on dose to standard reference points and the bladder and rectum. The hypothesis of this work is that the ICRU38 bladder and rectal points computed conventionally are not clinically acceptable surrogates for the maximum dose points as determined by MCD. ^ MCD was applied to the tandem and ovoids. The FSD ovoids and tandem were modeled in a single input file that allowed dose to be calculated for any patient. Dose difference surface histograms(DDSH) were computed for the bladder and rectum. Reference point doses were compared between shielded and unshielded ovoids, and a commercial treatment planning system. ^ The results of this work showed the tandem tip screw caused a 33% reduction in dose. The ovoid shields reduced the dose by a maximum of 48.9%. DDSHs revealed on average 5% of the bladder surface area was spared 53 cGy and 5% of the rectal surface area was spared 195 cGy. The ovoid shields on average reduced the dose by 18% for the bladder point and 25% for the rectal point. The Student's t-test revealed the ICRU38 bladder and rectal points do not predict the maximum dose for these organs. ^ It is concluded that modeling the tandem and ovoid internal structures is necessary for accurate dose calculations, the bladder shielding segments may not be necessary, and that the ICRU38 bladder point is irrelevant. ^

Verification of intensity modulated stereotactic radiotherapy using Monte Carlo calculations and EPID dosimetry

The purpose of this work was to develop a comprehensive IMSRT QA procedure that examined, using EPID dosimetry and Monte Carlo (MC) calculations, each step in the treatment planning and delivery process. These steps included verification of the field shaping, treatment planning system (RTPS) dose calculations, and patient dose delivery. Verification of each step in the treatment process is assumed to result in correct dose delivery to the patient. ^ The accelerator MC model was verified against commissioning data for field sizes from 0.8 × 0.8 cm 2 to 10 × 10 cm 2. Depth doses were within 2% local percent difference (LPD) in low gradient regions and 1 mm distance to agreement (DTA) in high gradient regions. Lateral profiles were within 2% LPD in low gradient regions and 1 mm DTA in high gradient regions. Calculated output factors were within 1% of measurement for field sizes ≥1 × 1 cm2. ^ The measured and calculated pretreatment EPID dose patterns were compared using criteria of 5% LPD, 1 mm DTA, or 2% of central axis pixel value with ≥95% of compared points required to pass for successful verification. Pretreatment field verification resulted in 97% percent of the points passing. ^ The RTPS and Monte Carlo phantom dose calculations were compared using 5% LPD, 2 mm DTA, or 2% of the maximum dose with ≥95% of compared points required passing for successful verification. RTPS calculation verification resulted in 97% percent of the points passing. ^ The measured and calculated EPID exit dose patterns were compared using criteria of 5% LPD, 1 mm DTA, or 2% of central axis pixel value with ≥95% of compared points required to pass for successful verification. Exit dose verification resulted in 97% percent of the points passing. ^ Each of the processes above verified an individual step in the treatment planning and delivery process. The combination of these verification steps ensures accurate treatment delivery to the patient. This work shows that Monte Carlo calculations and EPID dosimetry can be used to quantitatively verify IMSRT treatments resulting in improved patient care and, potentially, improved clinical outcome. ^

Monte Carlo and analytical dose calculations for ocular proton therapy

Uveal melanoma is a rare but life-threatening form of ocular cancer. Contemporary treatment techniques include proton therapy, which enables conservation of the eye and its useful vision. Dose to the proximal structures is widely believed to play a role in treatment side effects, therefore, reliable dose estimates are required for properly evaluating the therapeutic value and complication risk of treatment plans. Unfortunately, current simplistic dose calculation algorithms can result in errors of up to 30% in the proximal region. In addition, they lack predictive methods for absolute dose per monitor unit (D/MU) values. ^ To facilitate more accurate dose predictions, a Monte Carlo model of an ocular proton nozzle was created and benchmarked against measured dose profiles to within ±3% or ±0.5 mm and D/MU values to within ±3%. The benchmarked Monte Carlo model was used to develop and validate a new broad beam dose algorithm that included the influence of edgescattered protons on the cross-field intensity profile, the effect of energy straggling in the distal portion of poly-energetic beams, and the proton fluence loss as a function of residual range. Generally, the analytical algorithm predicted relative dose distributions that were within ±3% or ±0.5 mm and absolute D/MU values that were within ±3% of Monte Carlo calculations. Slightly larger dose differences were observed at depths less than 7 mm, an effect attributed to the dose contributions of edge-scattered protons. Additional comparisons of Monte Carlo and broad beam dose predictions were made in a detailed eye model developed in this work, with generally similar findings. ^ Monte Carlo was shown to be an excellent predictor of the measured dose profiles and D/MU values and a valuable tool for developing and validating a broad beam dose algorithm for ocular proton therapy. The more detailed physics modeling by the Monte Carlo and broad beam dose algorithms represent an improvement in the accuracy of relative dose predictions over current techniques, and they provide absolute dose predictions. It is anticipated these improvements can be used to develop treatment strategies that reduce the incidence or severity of treatment complications by sparing normal tissue. ^