Mixture modeling for joint analysis of survival, discrete, and continuous data

Mixture modeling is commonly used to model categorical latent variables that represent subpopulations in which population membership is unknown but can be inferred from the data. In relatively recent years, the potential of finite mixture models has been applied in time-to-event data. However, the commonly used survival mixture model assumes that the effects of the covariates involved in failure times differ across latent classes, but the covariate distribution is homogeneous. The aim of this dissertation is to develop a method to examine time-to-event data in the presence of unobserved heterogeneity under a framework of mixture modeling. A joint model is developed to incorporate the latent survival trajectory along with the observed information for the joint analysis of a time-to-event variable, its discrete and continuous covariates, and a latent class variable. It is assumed that the effects of covariates on survival times and the distribution of covariates vary across different latent classes. The unobservable survival trajectories are identified through estimating the probability that a subject belongs to a particular class based on observed information. We applied this method to a Hodgkin lymphoma study with long-term follow-up and observed four distinct latent classes in terms of long-term survival and distributions of prognostic factors. Our results from simulation studies and from the Hodgkin lymphoma study demonstrated the superiority of our joint model compared with the conventional survival model. This flexible inference method provides more accurate estimation and accommodates unobservable heterogeneity among individuals while taking involved interactions between covariates into consideration.^

Analyzing left-truncated right-censored data with uncertain onset time with parametric models

Prevalent sampling is an efficient and focused approach to the study of the natural history of disease. Right-censored time-to-event data observed from prospective prevalent cohort studies are often subject to left-truncated sampling. Left-truncated samples are not randomly selected from the population of interest and have a selection bias. Extensive studies have focused on estimating the unbiased distribution given left-truncated samples. However, in many applications, the exact date of disease onset was not observed. For example, in an HIV infection study, the exact HIV infection time is not observable. However, it is known that the HIV infection date occurred between two observable dates. Meeting these challenges motivated our study. We propose parametric models to estimate the unbiased distribution of left-truncated, right-censored time-to-event data with uncertain onset times. We first consider data from a length-biased sampling, a specific case in left-truncated samplings. Then we extend the proposed method to general left-truncated sampling. With a parametric model, we construct the full likelihood, given a biased sample with unobservable onset of disease. The parameters are estimated through the maximization of the constructed likelihood by adjusting the selection bias and unobservable exact onset. Simulations are conducted to evaluate the finite sample performance of the proposed methods. We apply the proposed method to an HIV infection study, estimating the unbiased survival function and covariance coefficients. ^

Exon involvement in the regulation of MuSVts110 RNA splicing

MuSVts110 is a conditionally defective mutant of Moloney murine sarcoma virus which undergoes a novel tmperature-dependent splice event at growth temperatures of 33$\sp\circ$C or lower. Relative to wild-type MuSV-124, MuSVts110 contains a 1487 base deletion spanning from the 3$\sp\prime$ end of the p30 gag coding region to just downstream of the first v-mos initiation codon. As a result, the gag and mos genes are fused out of frame and no v-mos protein is expressed. However, upon a shift to 33$\sp\circ$C or lower, a splice event occurs which removes 431 bases, realigns the gag and mos genes, and allows read-through translation of a P85gag-mos transforming protein. Interestingly, while the cryptic splice sites utilized in MuSVts110 are present and unaltered in MuSV-124, they are never used. Due to the 1487 base deletion, the MuSV-124 intron was reduced from 1919 to 431 bases suggesting that intron size might be involved in the activation of these cryptic splice sites in MuSVts110. Since the splicing phenotype of the MuSVts110 equivalent (TS32 DNA) which contains the identical 1487 base deletion introduced into otherwise wild-type MuSV-124 DNA, was indistinguishable from authentic MuSVts110, it was concluded that this deletion alone is responsible for activation of the cryptic splice sites used in MuSVts110. These results also confirmed that thermodependent splicing is an intrinsic property of the viral RNA and not due to some cellular defect. Furthermore, analysis of gag gene deletion and frameshift MuSVts110 mutants demonstrated that viral gag gene proteins do not play a role in regulation of MuSVts110 splicing. Instead, cis-acting viral sequences appear to mediate regulation of the splice event.^ Our initial observation that truncation of the MuSVts110 transcript, leaving only residual amounts of the flanking exon sequences, completely abolished splicing activity argued that exon sequences might participate in the regulation of the splice event.^ Analysis of exon sequence involvement has also identified cis-acting sequences important in the thermodependence of the splice event. Data suggest that regulation of the MuSVts110 splice event involves multiple interactions between specific intron and exon sequences and spliceosome components which together limit splicing activity to temperatures of 33$\sp\circ$C or lower while simultaneously restricting splicing to a maximum of 50% efficiency. (Abstract shortened with permission of author.) ^

Kawasaki syndrome: Cluster analysis of incident cases presenting to a pediatric tertiary care hospital in Texas from January 1, 2005 to December 31, 2009

Background: Despite almost 40 years of research into the etiology of Kawasaki Syndrome (KS), there is little research published on spatial and temporal clustering of KS cases. Previous analysis has found significant spatial and temporal clustering of cases, therefore cluster analyses were performed to substantiate these findings and provide insight into incident KS cases discharged from a pediatric tertiary care hospital. Identifying clusters from a single institution would allow for prospective analysis of risk factors and potential exposures for further insight into KS etiology. ^ Methods: A retrospective study was carried out to examine the epidemiology and distribution of patients presenting to Texas Children’s Hospital in Houston, Texas, with a diagnosis of Acute Febrile Mucocutaneous Lymph Node Syndrome (MCLS) upon discharge from January 1, 2005 to December 31, 2009. Spatial, temporal, and space-time cluster analyses were performed using the Bernoulli model with case and control event data. ^ Results: 397 of 102,761 total patients admitted to Texas Children’s Hospital had a principal or secondary diagnosis of Acute Febrile MCLS upon over the 5 year period. Demographic data for KS cases remained consistent with known disease epidemiology. Spatial, temporal, and space-time analyses of clustering using the Bernoulli model demonstrated no statistically significant clusters. ^ Discussion: Despite previous findings of spatial-temporal clustering of KS cases, there were no significant clusters of KS cases discharged from a single institution. This implicates the need for an expanded approach to conducting spatial-temporal cluster analysis and KS surveillance given the limitations of evaluating data from a single institution.^

Effect of patient accrual patterns on power and size of log-rank test for time-to-event outcome in clinical trials

The determination of size as well as power of a test is a vital part of a Clinical Trial Design. This research focuses on the simulation of clinical trial data with time-to-event as the primary outcome. It investigates the impact of different recruitment patterns, and time dependent hazard structures on size and power of the log-rank test. A non-homogeneous Poisson process is used to simulate entry times according to the different accrual patterns. A Weibull distribution is employed to simulate survival times according to the different hazard structures. The current study utilizes simulation methods to evaluate the effect of different recruitment patterns on size and power estimates of the log-rank test. The size of the log-rank test is estimated by simulating survival times with identical hazard rates between the treatment and the control arm of the study resulting in a hazard ratio of one. Powers of the log-rank test at specific values of hazard ratio (≠1) are estimated by simulating survival times with different, but proportional hazard rates for the two arms of the study. Different shapes (constant, decreasing, or increasing) of the hazard function of the Weibull distribution are also considered to assess the effect of hazard structure on the size and power of the log-rank test. ^