Stefan Wagenpfeil

Smoothing Hazard Functions and Time-Varying Effects in Discrete Duration and Competing Risks Models

Abstract State-space or dynamic approaches to discrete or grouped duration data with competing risks or multiple terminating events allow simultaneous modeling and smooth estimation of hazard functions and time-varying effects in a flexible way. Full Bayesian or posterior mean estimation, using numerical integration techniques or Monte Carlo methods, can become computationally rather demanding or even infeasible for higher dimensions and larger datasets. Therefore, based on previous work on filtering and smoothing for multicategorical time series and longitudinal data, our approach uses posterior mode estimation. Thus we have to maximize posterior densities or, equivalently, a penalized likelihood, which enforces smoothness of hazard functions and time-varying effects by a roughness penalty. Dropping the Bayesian smoothness prior and adopting a nonparametric viewpoint, one might also start directly from maximizing this penalized likelihood. We show how Fisher scoring smoothing iterations can be carried ou...

Psoriasin (S100A7) up-regulation in oral squamous cell carcinoma and its relation to clinicopathologic features

Proteomic analysis recently suggested aberrant psoriasin (S100A7) expression in oral squamous cell carcinoma (OSCC). In this study, OSCC specimens and matching normal oral tissues from 45 patients who had undergone ablative surgery were examined. Increased psoriasin expression at mRNA level was observed in OSCC samples by quantitative real-time RT-PCR (p=0.015). Immunofluorescence analysis with psoriasin antibody confirmed these observations. Moreover, significantly increased mRNA ratios between malignant and normal samples were correlated with early UICC stage (p=0.006), T1/T2 tumour classification (p=0.043), absence of cervical lymph node metastasis (p=0.027) and age under 65 (p=0.009). Additionally, well-differentiated tumour tissues demonstrated a significantly higher psoriasin expression than moderate and poor differentiated carcinomas (p=0.018). Based on this data, we conclude that psoriasin is a positive marker for oral cancerogenesis and early tumour progression.

Penalized likelihood estimation and iterative Kalman smoothing for non-Gaussian dynamic regression models

Abstract Dynamic regression or state space models provide a flexible framework for analyzing non-Gaussian time series and longitudinal data, covering for example models for discrete longitudinal observations. As for non-Gaussian random coefficient models, a direct Bayesian approach leads to numerical integration problems, often intractable for more complicated data sets. Recent Markov chain Monte Carlo methods avoid this by repeated sampling from approximative posterior distributions, but there are still open questions about sampling schemes and convergence. In this article we consider simpler methods of inference based on posterior modes or, equivalently, maximum penalized likelihood estimation. From the latter point of view, the approach can also be interpreted as a nonparametric method for smoothing time-varying coefficients. Efficient smoothing algorithms are obtained by iteration of common linear Kalman filtering and smoothing, in the same way as estimation in generalized linear models with fixed effects can be performed by iteratively weighted least squares estimation. The algorithm can be combined with an EM-type method or cross-validation to estimate unknown hyper- or smoothing parameters. The approach is illustrated by applying to a binary time series and a multicategorical longitudinal data set.

Cost Analysis Comparing an Anthracycline/Docetaxel Regimen to CMF in Patients with Early Stage Breast Cancer

Background: Taxane-based adjuvant chemotherapy is the current standard for node-positive breast cancer patients. Recent data identified relevant patient subgroups with questionable benefit. To estimate the incremental burden on health care resources and costs, we compared a modern sequential regimen (4×epirubicin/cyclophosphamide; 4×docetaxel: EC→DOC) to CMF. Patients and Methods: Data were obtained alongside the phase III WSG-AGO Intergroup trial (2000–2005). A cohort of 110 patients receiving 1,047 chemotherapy cycle days at 38 study sites was analyzed from a hospital perspective. Results: Mean age was 52.4 years. Mean costs for the EC→DOC group (n = 54) totaled €8,459 per patient (95% confidence interval (CI): €7,785–9,132) with cytostatic drug costs being the largest burden (€5,673; 67%). CMF was significantly (–41.2%) less expensive (€4,973; 95% CI: €4,706–5,240), and toxicity-associated rehospitalization was reduced by half (CMF: n = 4, EC→DOC:n =8). Conclusions: Our results demonstrate a substantial budget increase attributable to introduction of taxanes to adjuvant chemotherapy of breast cancer. Data will allow estimating cost-effectiveness of individualized chemotherapy strategies.

Statistical analysis of combined dose effects for experiments with two agents

OBJECTIVE Classical isobologram analysis offers a way for analysing combined drug effects in dose-response experiments statistically. The aim is to determine as to whether two agents or drugs can be considered synergistic or antagonistic in their effect. METHODS AND MATERIALS We describe a MATLAB-based software tool for automated isobologram analysis and computation of combination indices. Statistical issues like estimation together with respective confidence intervals are of key interest. Additional predictive values are computed to facilitate a more easy interpretation of obtained results. RESULTS Analysis of an experimental and a real in vitro data set demonstrates the approach and the way of interpreting results. Results are summarized in two ways: tables and graphical displays containing classical isobolograms. CONCLUSION Our package supplements the clinical software-equipment and is a tool for automatic evaluation of combined dose-response experiments in experimental oncology in the urologic clinic.

[The influence of surgical and restorative dental treatment prior to cardiac valve surgery on the long-term demand of dental treatment: a prospective clinical study].

The study objective was to evaluate the long-term influence of non-radical surgical and restorative dental treatment modalities prior to elective cardiac valve replacement on the subsequent dental treatment demand. A total of 305 patients preceding cardiac valve surgery were screened and the appropriate dental treatment was initiated. After 36 months 80 patients were re-evaluated clinically, of which 60 required dental restorations of 155 teeth, mostly due to periodontal pathology. Independent of the sub-group there was a statistically substantial increase of the treatment demand compared to the time of initial examination. In addition, at the time of final re-evaluation the definite treatment need significantly increased far beyond anticipation, potentially due to inadequate dental procedures during the follow-up interval. Irrespective of any dental treatment or antibiotic application, endocarditis did not occur in any patient. Non-radical dental restoration prior to cardiac valve replacement can only be successful, if a standardized dental follow-up with common monitoring forms is provided. The risk of prosthetic valve infective endocarditis based on a dental focus is probably overestimated.SummaryThe study objective was to evaluate the long-term influence of non-radical surgical and restorative dental treatment modalities prior to elective cardiac valve replacement on the subsequent dental treatment demand. A total of 305 patients preceding cardiac valve surgery were screened and the appropriate dental treatment was initiated. After 36 months 80 patients were re-evaluated clinically, of which 60 required dental restorations of 155 teeth, mostly due to periodontal pathology. Independent of the sub-group there was a statistically substantial increase of the treatment demand compared to the time of initial examination. In addition, at the time of final re-evaluation the definite treatment need significantly increased far beyond anticipation, potentially due to inadequate dental procedures during the follow-up interval. Irrespective of any dental treatment or antibiotic application, endocarditis did not occur in any patient. Non-radical dental restoration prior to cardiac valve replacement can only be successful, if a standardized dental follow-up with common monitoring forms is provided. The risk of prosthetic valve infective endocarditis based on a dental focus is probably overestimated.ZusammenfassungStudienziel war die Evaluation des Einflusses einer nicht-radikalen Zahnsanierung vor einer geplanten Herzklappenersatzoperation auf den langfristigen postoperativen zahnärztlichen Behandlungsbedarf. 305 Patienten wurden vor einer Klappenersatzoperation hinsichtlich des zahnärztlichen Fokussanierungsbedarfs gescreent und notwendige Behandlungen eingeleitet. Nach 36 Monaten konnten 80 dieser Patienten klinisch reevaluiert werden. Bei 60 Patienten waren 155 Zähne behandlungsbedürftig, meistens aufgrund parodontaler Schädigungen. Gruppenunabhängig lag ein statistisch signifikanter Anstieg des Behandlungsbedarfs gegenüber dem initialen Erhebungsbefund vor. Darüber hinaus zeigte sich zum Nachbeobachtungszeitpunkt eine signifikant über der Erwartung liegende tatsächliche Therapienotwendigkeit, was möglicherweise auf nicht adäquate Therapiemaßnahmen während des Follow-up-Intervalls zurückzuführen ist. Eine Endokarditis war ungeachtet einer präoperativen Sanierung und einer meistens erfolgten prophylaktischen Antibiotikagabe im Gesamtkollektiv nicht aufgetreten. Nicht-radikale zahnärztliche Behandlungen vor einer Herzklappenersatzoperation können nur dann erfolgreich sein, wenn auch eine anschließende engmaschige Nachsorge unter standardisierten Nachbeobachtungsrichtlinien mit entsprechenden Erhebungsbögen gewährleistet ist. Die Gefahr einer von einem unbehandelten dentalen Fokus ausgehenden Ersatzklappenendokarditis wird wahrscheinlich überschätzt.

Isobologram Analysis in MATLAB for Combined Effects of Two Agents in Dose-Response Experiments

Isobologram analysis is a way of exploring and visualizing combined drug effects. The aim of this work is to determine as to whether two agents can be considered synergistic or antagonistic. Resulting from a clinical consulting case in urology we developed a MATLAB-based software tool for automated isobologram analysis. In this way we supplement the clinical software-equipment in our laboratory and encourage the evaluation of combined dose-response experiments. Analysis of an example data set demonstrates the approach and the way of interpreting obtained results.

A MATLAB-Based Software Tool for Changepoint Detection and Nonlinear Regression in Dose-Response Relationships

Resulting from a clinical consulting case in urology we developed a software tool for determining nonlinear dose-response relationships. Unlike most existing statistical software packages, we directly compute and display analytical pointwise 95% confidence intervals for the prediction result. Furthermore, user-defined changepoints with 95% confidence interval can be calculated in order to estimate the dosage for a 50% response rate, for instance. This is necessary to compare the effect of different retinoids, tumor cell lines, etc. In this way we supplement the clinical software-equipment in our laboratory and encourage the evaluation of dose-response data. The numerical and computational problems arising with nonlinear regression, 4-parameter logistic as well as log-logit modelling and the respective confidence intervals are addressed in particular. Analysis of real data and an example data set demonstrate the approach. A demo version of the software tool can be downloaded from the first authors homepage.

Eine klinisch prospektive Untersuchung zum Einfluss der Fokussanierung vor Herzklappenoperation auf den langfristigen zahnärztlichen Behandlungsbedarf

The study objective was to evaluate the long-term influence of non-radical surgical and restorative dental treatment modalities prior to elective cardiac valve replacement on the subsequent dental treatment demand. A total of 305 patients preceding cardiac valve surgery were screened and the appropriate dental treatment was initiated. After 36 months 80 patients were re-evaluated clinically, of which 60 required dental restorations of 155 teeth, mostly due to periodontal pathology. Independent of the sub-group there was a statistically substantial increase of the treatment demand compared to the time of initial examination. In addition, at the time of final re-evaluation the definite treatment need significantly increased far beyond anticipation, potentially due to inadequate dental procedures during the follow-up interval. Irrespective of any dental treatment or antibiotic application, endocarditis did not occur in any patient. Non-radical dental restoration prior to cardiac valve replacement can only be successful, if a standardized dental follow-up with common monitoring forms is provided. The risk of prosthetic valve infective endocarditis based on a dental focus is probably overestimated.SummaryThe study objective was to evaluate the long-term influence of non-radical surgical and restorative dental treatment modalities prior to elective cardiac valve replacement on the subsequent dental treatment demand. A total of 305 patients preceding cardiac valve surgery were screened and the appropriate dental treatment was initiated. After 36 months 80 patients were re-evaluated clinically, of which 60 required dental restorations of 155 teeth, mostly due to periodontal pathology. Independent of the sub-group there was a statistically substantial increase of the treatment demand compared to the time of initial examination. In addition, at the time of final re-evaluation the definite treatment need significantly increased far beyond anticipation, potentially due to inadequate dental procedures during the follow-up interval. Irrespective of any dental treatment or antibiotic application, endocarditis did not occur in any patient. Non-radical dental restoration prior to cardiac valve replacement can only be successful, if a standardized dental follow-up with common monitoring forms is provided. The risk of prosthetic valve infective endocarditis based on a dental focus is probably overestimated.ZusammenfassungStudienziel war die Evaluation des Einflusses einer nicht-radikalen Zahnsanierung vor einer geplanten Herzklappenersatzoperation auf den langfristigen postoperativen zahnärztlichen Behandlungsbedarf. 305 Patienten wurden vor einer Klappenersatzoperation hinsichtlich des zahnärztlichen Fokussanierungsbedarfs gescreent und notwendige Behandlungen eingeleitet. Nach 36 Monaten konnten 80 dieser Patienten klinisch reevaluiert werden. Bei 60 Patienten waren 155 Zähne behandlungsbedürftig, meistens aufgrund parodontaler Schädigungen. Gruppenunabhängig lag ein statistisch signifikanter Anstieg des Behandlungsbedarfs gegenüber dem initialen Erhebungsbefund vor. Darüber hinaus zeigte sich zum Nachbeobachtungszeitpunkt eine signifikant über der Erwartung liegende tatsächliche Therapienotwendigkeit, was möglicherweise auf nicht adäquate Therapiemaßnahmen während des Follow-up-Intervalls zurückzuführen ist. Eine Endokarditis war ungeachtet einer präoperativen Sanierung und einer meistens erfolgten prophylaktischen Antibiotikagabe im Gesamtkollektiv nicht aufgetreten. Nicht-radikale zahnärztliche Behandlungen vor einer Herzklappenersatzoperation können nur dann erfolgreich sein, wenn auch eine anschließende engmaschige Nachsorge unter standardisierten Nachbeobachtungsrichtlinien mit entsprechenden Erhebungsbögen gewährleistet ist. Die Gefahr einer von einem unbehandelten dentalen Fokus ausgehenden Ersatzklappenendokarditis wird wahrscheinlich überschätzt.