Harald Binder

Assessment of survival prediction models based on microarray data

MOTIVATION In the process of developing risk prediction models, various steps of model building and model selection are involved. If this process is not adequately controlled, overfitting may result in serious overoptimism leading to potentially erroneous conclusions. METHODS For right censored time-to-event data, we estimate the prediction error for assessing the performance of a risk prediction model (Gerds and Schumacher, 2006; Graf et al., 1999). Furthermore, resampling methods are used to detect overfitting and resulting overoptimism and to adjust the estimates of prediction error (Gerds and Schumacher, 2007). RESULTS We show how and to what extent the methodology can be used in situations characterized by a large number of potential predictor variables where overfitting may be expected to be overwhelming. This is illustrated by estimating the prediction error of some recently proposed techniques for fitting a multivariate Cox regression model applied to the data of a prognostic study in patients with diffuse large-B-cell lymphoma (DLBCL). AVAILABILITY Resampling-based estimation of prediction error curves is implemented in an R package called pec available from the authors.

Treatment-effect estimates adjusted for small-study effects via a limit meta-analysis

Statistical heterogeneity and small-study effects are 2 major issues affecting the validity of meta-analysis. In this article, we introduce the concept of a limit meta-analysis, which leads to shrunken, empirical Bayes estimates of study effects after allowing for small-study effects. This in turn leads to 3 model-based adjusted pooled treatment-effect estimators and associated confidence intervals. We show how visualizing our estimators using the radial plot indicates how they can be calculated using existing software. The concept of limit meta-analysis also gives rise to a new measure of heterogeneity, termed G(2), for heterogeneity that remains after small-study effects are accounted for. In a simulation study with binary data and small-study effects, we compared our proposed estimators with those currently used together with a recent proposal by Moreno and others. Our criteria were bias, mean squared error (MSE), variance, and coverage of 95% confidence intervals. Only the estimators arising from the limit meta-analysis produced approximately unbiased treatment-effect estimates in the presence of small-study effects, while the MSE was acceptably small, provided that the number of studies in the meta-analysis was not less than 10. These limit meta-analysis estimators were also relatively robust against heterogeneity and one of them had a relatively small coverage error.

Boosting for high-dimensional time-to-event data with competing risks

MOTIVATION For analyzing high-dimensional time-to-event data with competing risks, tailored modeling techniques are required that consider the event of interest and the competing events at the same time, while also dealing with censoring. For low-dimensional settings, proportional hazards models for the subdistribution hazard have been proposed, but an adaptation for high-dimensional settings is missing. In addition, tools for judging the prediction performance of fitted models have to be provided. RESULTS We propose a boosting approach for fitting proportional subdistribution hazards models for high-dimensional data, that can e.g. incorporate a large number of microarray features, while also taking clinical covariates into account. Prediction performance is evaluated using bootstrap.632+ estimates of prediction error curves, adapted for the competing risks setting. This is illustrated with bladder cancer microarray data, where simultaneous consideration of both, the event of interest and competing events, allows for judging the additional predictive power gained from incorporating microarray measurements. AVAILABILITY The proposed boosting approach is implemented in the R package CoxBoost and prediction error estimation in the package peperr, both available from CRAN.

Design of a placebo-controlled, randomized study of the efficacy of repetitive transcranial magnetic stimulation for the treatment of chronic tinnitus

BackgroundChronic tinnitus is a frequent condition, which can have enormous impact on patients life and which is very difficult to treat. Accumulating data indicate that chronic tinnitus is related to dysfunctional neuronal activity in the central nervous system. Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive method which allows to focally modulate neuronal activity. An increasing amount of studies demonstrate reduction of tinnitus after repeated sessions of low-frequency rTMS and indicate that rTMS might represent a new promising approach for the treatment of tinnitus. However available studies have been mono-centric and are characterized by small sample sizes. Therefore, this multi-center trial will test the efficacy of rTMS treatment in a large sample of chronic tinnitus patients.Methods/DesignThis is a randomized, placebo-controlled, double-blind multi-center trial of two weeks 1 Hz rTMS-treatment in chronic tinnitus patients. Eligible patients will be randomized to either 2 weeks real or sham rTMS treatment. Main eligibility criteria: male or female individuals aged 18–70 years with chronic tinnitus (duration > 6 months), tinnitus-handicap-inventory-score ≥ 38, age-adjusted normal sensorineural hearing (i.e. not more than 5 dB below the 10% percentile of the appropriate age and gender group (DIN EN ISO 7029), conductive hearing loss ≤ 15dB. The primary endpoint is a change of tinnitus severity according to the tinnitus questionnaire of Goebel and Hiller (baseline vs. end of treatment period). A total of 138 patients are needed to detect a clinical relevant change of tinnitus severity (i.e. 5 points on the questionnaire of Goebel and Hiller; alpha = 0.05; 1-beta = 0.80). Assuming a drop-out rate of less than 5% until the primary endpoint, 150 patients have to be randomized to guarantee the target number of 138 evaluable patients. The study will be conducted by otorhinolaryngologists and psychiatrists of 7 university hospitals and 1 municipal hospital in Germany.DiscussionThis study will provide important information about the efficacy of rTMS in the treatment of chronic tinnitus.Trial registrationCurrent Controlled Trials ISRCTN89848288

Stability investigations of multivariable regression models derived from low- and high-dimensional data.

Multivariable regression models can link a potentially large number of variables to various kinds of outcomes, such as continuous, binary, or time-to-event endpoints. Selection of important variables and selection of the functional form for continuous covariates are key parts of building such models but are notoriously difficult due to several reasons. Caused by multicollinearity between predictors and a limited amount of information in the data, (in)stability can be a serious issue of models selected. For applications with a moderate number of variables, resampling-based techniques have been developed for diagnosing and improving multivariable regression models. Deriving models for high-dimensional molecular data has led to the need for adapting these techniques to settings where the number of variables is much larger than the number of observations. Three studies with a time-to-event outcome, of which one has high-dimensional data, are used to illustrate several techniques. Investigations at the covariate level and at the predictor level are seen to provide considerable insight into model stability and performance. While some areas are indicated where resampling techniques for model building still need further refinement, our case studies illustrate that these techniques can already be recommended for wider use.

Modulating cerebello-thalamocortical pathways by neuronavigated cerebellar repetitive transcranial stimulation (rTMS)

OBJECTIVES Increasing evidence suggests that dysfunctions of the cortico-cerebello-thalamocortical circuit are involved in the pathophysiology of neuropsychiatric disorders. This study explores the effects of cerebellar repetitive transcranial magnetic stimulation (rTMS) on cerebello-thalamocortical pathways. METHODS Ten healthy volunteers received MRI-guided rTMS in four separate sessions (120% motor threshold, 1000 stimuli) over either the medial or the right lateral cerebellum using frequencies of 1 and 10 Hz. Motor cortex excitability was assessed before and after the intervention by paired-pulse transcranial magnetic stimulation. RESULTS Depending on stimulation frequency, cerebellar rTMS differentially modified intracortical inhibition. Low frequency rTMS increased short intracortical inhibition (SICI), whereas high frequency rTMS had no significant effect on SICI. CONCLUSIONS These results suggest that rTMS over the cerebellum can modulate cerebello-thalamocortical pathways in a frequency-specific manner.

Sparse regression techniques in low-dimensional survival data settings

In high-dimensional data settings, sparse model fits are desired, which can be obtained through shrinkage or boosting techniques. We investigate classical shrinkage techniques such as the lasso, which is theoretically known to be biased, new techniques that address this problem, such as elastic net and SCAD, and boosting technique CoxBoost and extensions of it, which allow to incorporate additional structure. To examine, whether these methods, that are designed for or frequently used in high-dimensional survival data analysis, provide sensible results in low-dimensional data settings as well, we consider the well known GBSG breast cancer data. In detail, we study the bias, stability and sparseness of these model fitting techniques via comparison to the maximum likelihood estimate and resampling, and their prediction performance via prediction error curve estimates.

The forgotten caregivers: Siblings of schizophrenic patients

Objective. The situation of caregivers of psychiatric patients is mostly focussed on burdens of parents or spouses of patients. The burden of siblings due to the illness, however, is mostly underestimated and disregarded. Methods. Thirty-seven narrative interviews with siblings of schizophrenia inpatients were analysed by using a summarizing content analysis. The founded global statements were quantitatively analysed. Regression-analysis as well as regression trees were used to evaluate the data linked with sociodemographic and disease-related variables of the patient and siblings. Results. The results showed a high proportion of siblings engaged in caregiving activities. A total of 492 individual statements were summarized in 26 global types of statements. The three most often reported burdens by the healthy siblings are: “Handling the symptoms of illness” (100%), “Emotional burden due to the illness of the sibling” (100%) and “Uncertainty in judging what amount of stress the patient can cope with” (81.1%). Linear regression and regression tree analysis show predictors for higher burdened siblings. Conclusion. Siblings of schizophrenia patients are burdened in various aspects and in a specific matter. Their special needs will therefore have to be recognised before they can receive appropriate intervention.

Comment on ‘network-constrained regularization and variable selection for analysis of genomic data’

Contact: binderh@fdm.uni-freiburg.de Supplementary information: Supplementary data are available at Bioinformatics online.

Parallelized prediction error estimation for evaluation of high-dimensional models

UNLABELLED There is a multitude of new techniques that promise to extract predictive information in bioinformatics applications. It has been recognized that a first step for validation of the resulting model fits should rely on proper use of resampling techniques. However, this advice is frequently not followed, potential reasons being difficulty of correct implementation and computational demand. This is addressed by the R package peperr, which is designed for reliable prediction error estimation through resampling, potentially accelerated by parallel execution on a compute cluster. Its interface allows easy connection to newly developed model fitting routines. Performance evaluation of the latter is furthermore guided by diagnostic plots, which helps to detect specific problems due to high-dimensional data structures. AVAILABILITY http://cran.r-project.org, http://www.imbi.uni-freiburg.de/parallel. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.