Cookie Settings

We use cookies to optimize our website. These include cookies that are necessary for the operation of the site, as well as those that are only used for anonymous statistic. You can decide for yourself which categories you want to allow. Further information can be found in our data privacy protection .


These cookies are necessary to run the core functionalities of this website and cannot be disabled.

Name Webedition CMS
Purpose This cookie is required by the CMS (Content Management System) Webedition for the system to function correctly. Typically, this cookie is deleted when the browser is closed.
Name econda
Purpose Session cookie emos_jcsid for the web analysis software econda. This runs in the “anonymized measurement” mode. There is no personal reference. As soon as the user leaves the site, tracking is ended and all data in the browser are automatically deleted.

These cookies help us understand how visitors interact with our website by collecting and analyzing information anonymously. Depending on the tool, one or more cookies are set by the provider.

Name econda
Purpose Statistics
External media

Content from external media platforms is blocked by default. If cookies from external media are accepted, access to this content no longer requires manual consent.

Name YouTube
Purpose Show YouTube content
Name Twitter
Purpose activate Twitter Feeds

Involvement in clinical trials

Members of the Division of Biostatistics are designated as responsible biometricians in a large number of phase I - III cancer clinical trials. This includes participation in the study design, calculation of power and sample size, development of statistical analysis plans, as well as the biometric analysis and final reporting and publication. In particular, we are the Biometric center for the German-Austrian Acute Myeloid Leukemia Study Group (AMLSG), the German-speaking Myeloma-Multicenter Group (GMMG) and the German Multicenter Study Group on Adult Acute Lymphoblastic Leukemia (GMALL). In addition, we support the NCT Trial Center and the Amyloidosis Center at the Heidelberg University Hospital in clinical and registry trials. We are responsible biometricians for trials and registries run by the Hopp Children's Cancer Center KiTZ. With our activities we are involved in practice changing clinical trials.
Motivated by our involvement as responsible statisticians for a large number of clinical trials, we develop methods for design and analysis of clinical trials, both in the Bayesian framework as well as the frequentist setting.

Bayesian phase I and phase II clinical trial design

Software available

There is strong interest and hope to improve efficacy of clinical trials with small sample sizes by using Bayesian trial designs, due to their flexibility and/or the possibility to easily implement borrowing of external information via informative prior distributions. One situation in which the number of patients that can be recruited is often limited is that of pediatric trials. A pediatric trial which sees the involvement of the Division is the INFORM2 phase I/II trial series run by the Hopp Children's Cancer Center KiTZ. The trial series has motivated a number of research projects in the area of Bayesian Phase I dose-finding trials, as well as in later Bayesian Phase II trials.

In Bayesian Phase II trials, assessment of frequentist error rates (test error rates, estimation error) is often still of relevance. It is important to note that, if control of frequentist type I error is required, and a uniformly most powerful test is available, no power can be gained by borrowing of information, a result that we have formally shown. On the other hand, 'local' improvements in test and estimation error rates can be achieved if external and current trial information are consistent. In this context, assessing and limiting the impact of information borrowing under different degrees of heterogeneity between current and external data sources, is of strong relevance. Development of devoted tools to perform such an assessment and, more broadly, to assist with the trial design are an area of former and current research in the Division.


Figure: Prior informativeness in terms of Effective Current Sample Size. Informativeness of a prior derived from an external study is quantified in terms of number of patients from the current trial, for different assumed true parameter configurations in the current study: under no/small heterogeneity, incorporation of external information is beneficial and results in a positive number of additional patients. Moderate heterogeneity results in worsened MSE and thus is equated to patients subtracted from the current study. Finally, in large heterogeneity situations, adaptive priors fully discard prior information, thus the effective sample size is null. See Wiesenfarth and Calderazzo (2020) for further details.

  • Calderazzo S., Kopp-Schneider A., Robust incorporation of historical information with known type I error rate inflation. Biom J. 66(1):e2200322 (2024)
  • Calderazzo S., Wiesenfarth M., Kopp-Schneider A., A decision-theoretic approach to Bayesian clinical trial design and evaluation of robustness to prior-data conflict. Biostatistics 23(1): 328-344 (2022)
  • Calderazzo S., Tarima S., Reid C., Flournoy N., Friede T., Geller N., Rosenberger J.L., Stallard N., Ursino M., Vandemeulebroecke M., Van Lancker K., Zohar S., Coping with Information Loss and the Use of Auxiliary Sources of Data: A Report from the NISS Ingram Olkin Forum Series on Unplanned Clinical Trial Disruptions. arXiv:2206.11238 (2022)
  • Kopp-Schneider A., Wiesenfarth M., Witt R., Edelmann D., Witt O., Abel U., Monitoring futility and efficacy in phase II trials with Bayesian posterior distributions - A calibration approach. Biom. J. 61: 488-502 (2019)
  • Kopp-Schneider A., Calderazzo S., Wiesenfarth M., Power gains by using external information in clinical trials are typically not possible when requiring strict type I error control. Biom. J.62(2): 361-374 (2020)
  • Kopp-Schneider A., Wiesenfarth M., Held L., Calderazzo S. Simulating and reporting frequentist operating characteristics of clinical trials that borrow external information: Towards a fair comparison in case of one-arm and hybrid control two-arm trials. Pharm Stat 23(1):4-19 (2024)
  • Wiesenfarth M., Calderazzo S., Quantification of prior impact in terms of effective current sample size. Biometrics. 76(1): 326-336 (2020)
  • Zocholl D, Wiesenfarth M., Rauch G., Kopp-Schneider A., On the feasibility of pediatric dose-finding trials in small samples with information from a preceding trial in adults. J. Biopharm. Stat. 32(5): 652-670 (2022)

Frequentist clinical trial design and analysis

Software available

In close cooperation with the needs of clinical researchers, several methodological developments to the design of clinical trials have been made, e.g.:

A novel dose finding design for phase-I studies has been proposed. In this design, a run-in intra-patient dose escalation part is followed by a rule- or model-based dose-finding design.
In many phase II cancer studies, the population under consideration is highly heterogeneous in terms of clinical, demographical, and biological covariates and response probabilities of patients may strongly vary. In this case, the operating characteristics of classical clinical trial designs heavily depend on the covariates of patients entering the study. To deal with this issue, we have derived modifications of Simon's optimal two-stage design correcting for heterogeneous populations using historical control data. This design is currently used for the TEAM study on relapsed and refractory AML.

Moreover, methods to improve the efficiency for evaluating clinical endpoints of trials have been identified:

Precision oncology trials often investigate multiple biomarkers simultaneously in independent subtrials. A challenge in the evaluation of such a trial is the small sample size in low-prevalence biomarker arms. We proposed to use the Firth correction to adjust for the small sample size bias and found that inclusion of biomarker-negative patients in the analysis can lead to further but small improvements in bias and standard deviation of the estimates.
The analysis of precision oncology trial results is often hampered by small sample sizes and/or the lack of an appropriate control arm. In consequence, methods have been proposed that take each patient as their own control by relating time-to-progression under the current to the previous therapy. Since the current use of this progression-free-survival ratio has numerous shortcomings, we propose alternative methods based on a marginal Cox model or a marginal accelerated failure time model for clustered time-to-event data with superior performance.

We are currently working on several other methodological projects that are directly motivated by our cooperation with clinical researchers, e.g.:

In sequentially randomized designs patients are initially allocated to first-line treatment groups. If patients meet certain criteria, they are potentially randomized to second-stage treatments. Estimation of survival distributions associated with the corresponding treatment policies is not straightforward. We are investigating new approaches to simulate and analyze two-stage trial designs.
The aim of a randomized trial is not necessarily to demonstrate the superiority of a new therapy over a standard therapy. The benefit for the patient may also be given if better tolerability, compliance and /or quality of life is achieved by the new therapy. Obviously, efficacy of new therapy must still be ensured. We are working on methodology for non-inferiority trials, in which the non-inferiority of the new experimental therapy compared to the standard therapy is tested under control of the type I error.

  • Edelmann D. et al., Adjusting Simon's optimal two-stage design for heterogeneous populations based on stratification or using historical controls. Biom. J. 62(2): 311-329 (2020).
  • Edelmann D. et al., The progression-free-survival ratio in molecularly-aided tumor trials - a critical examination of current practice and suggestions for alternative methods. Under revision
  • Habermehl C. et al., Addressing small sample size bias in multiple-biomarker trials: Inclusion of biomarker-negative patients and Firth correction. Biom. J. 60(2): 275-287 (2018).
  • Labrenz J. et al., Performance of phase-I dose finding designs with and without a run-in intra-patient dose escalation stage. Pharm. Stat. in press (2022)

to top
powered by webEdition CMS