Individualizing therapy for acute leukemia

Direct translation of cancer-specific genomic information into effective personalized therapies for acute leukemias has proven to be difficult. The aim of this study was to utilize novel tools for detailed characterization of genomic, transcriptomic, and functional aberrations in acute leukemia to gain understanding of disease pathology and guide individualized therapies. The main methods used were ex vivo drug sensitivity and resistance testing (DSRT), exome and transcriptome (RNA) sequencing, all of which were facilitated by extensive biobanking. In study I, we developed DSRT platform performed on primary leukemic cells ex vivo for identification of effective patient-specific drugs. Exome and RNA sequencing of serial samples were used for molecular characterization and to understand mechanisms of acquired resistance. Although AML samples exhibited unique DSRT profiles, responses could be clustered in five distinctive groups. Individualized treatment of refractory patients with DSRT-guided therapy resulted in meaningful clinical responses in 3/7 patients. In study II, we studied molecular drivers for relapsed T-cell acute lymphoblastic leukemia (T-ALL) and found novel STAT5B mutations. Functional studies demonstrated these mutations to enhance the transcriptional activity and to induce constitutive phosphorylation of STAT5B. The mutated blasts showed elevated BCL-XL expression and sensitivity to the pan-BCL-2 inhibitor navitoclax. Targeted sequencing revealed activating STAT5B mutations in 6/68 patients. In study III, we first evaluated ex vivo BCL-2 inhibitor sensitivity of AML cells, then we systematically assessed whether these responses correlated to specific mutations or gene expression signatures. BCL-2 inhibitor sensitivity associated with mutations in IDH1/IDH2 and WT1, as well as with aberrations in chromatin modifiers. Importantly, overexpression of a specific set of HOX genes predicted highly selective responses to BCL-2 inhibition. In study IV, we developed a national hematological biobank to allow researchers to access high-quality samples with accompanying clinical data. The samples from three collection time-points—diagnosis, potential remission, and relapse—are available. For this study, we evaluated the quality of stored samples and demonstrated that extracted DNA and RNA remain usable for demanding down-stream experiments.