Evidence-based support for phenotypic drug discovery in acute myeloid leukemia

The discovery and development of effective drugs for cancer patients has seen limited success in the clinic from phase I trials onward. The high attrition rate of current drug development approaches requires careful evaluation to provide a better understanding of the factors that correlate with or predict positive clinical outcomes. Here, we examine pre-clinical drug development approaches and conduct a meta-analysis of 2918 clinical studies involving 466 unique drugs tested in clinical trials for acute myeloid leukemia (AML). Our goal was to determine whether there are key shared pre-clinical characteristics that ultimately relate to successful or unsuccessful drugs in patients. We provide an evidence-based recommendation for the use of phenotypic drug discovery rather than other methods during pre-clinical development. Although our analysis was limited to AML, similar analyses are likely to be informative for other tumor-specific drug discovery campaigns, informing and improving the foundational discovery screens and platforms for other cancers.

Automated summary

The research highlights the limited success in the clinic of effective drugs for cancer patients, and the need for careful evaluation of high attrition rates in current drug development approaches. A meta-analysis of 2918 clinical studies involving 466 drugs for acute myeloid leukemia (AML) was conducted to determine key shared pre-clinical characteristics that relate to successful or unsuccessful drugs in patients. The recommendation is to use phenotypic drug discovery during pre-clinical development, informing and improving foundational discovery screens and platforms for other cancers.