PhD defence S.J.F. (Sjoerd) Hermans

Below is a brief summary of the dissertation:

Acute myeloid leukemia (AML) is a rare type of blood cancer, which requires intensive chemotherapy and a donor stem cell transplant to provide the highest chance of cure. However, the 5-year overall survival rate is still rather limited. Novel treatments are preferably investigated in randomized clinical trials (RCTs), as it is the most reliable study design. An important drawback, however, is that RCTs require a comparatively large sample size. Novel treatments are predominantly developed for higly-selected subgroups of patients, making it increasingly difficult to perform RCTs. A potential solution to perform comparative studies is to use external control patients, i.e., patients who were not previously included in clinical trials, but did receive the standard of care. The first part of this thesis addressed the methodological and statistical limitations of RWD; simultaneously, an automated system was developed to prospectively monitor RWD for the potential use as external controls. In addition to the use of a proper control group, the analytical methods is also of importance. The second part of this thesis illustrated the use of a more detailed survival analysis (“multi-state modelling”; MSM) to investigate treatment efficacy and adverse events in more detail in AML and multiple myeloma. MSM provided more details of the treatment effect in all analyses, as compared to standard survival methods. Lastly, recommendations concerning the use of RWD and MSM in prospective RCTs were postulated.