State-transition Modeling of Blood Transcriptome Predicts Disease Evolution and Treatment Response in Chronic Myeloid Leukemia (CML)

Chronic myeloid leukemia (CML) is initiated and initially maintained solely by the fusion gene BCR-ABL, encoding a mutant protein targeted in the clinic with tyrosine kinase inhibitors (TKIs) While TKI treatment is effective in inducing long-term remission, frequently is not curative. For these reasons, CML is an ideal disease to test our hypothesis that that transcriptome-based state-transition models accurately predict cancer evolution and treatment response. We hypothesized that transcriptome-based state-transition models accurately predict cancer evolution and treatment response. To test our hypothesis, we collected time-sequential peripheral blood samples from cohorts of tetracycline-off (Tet-Off) BCR-ABL-inducible transgenic mice. Using time-series bulk RNA-seq on the whole transcriptome to capture a system-wide view of all disease states, we applied state-transition theory to mathematically model CML development. We included four experimental cohorts of mice that were sampled weekly for 18 weeks or until mice became moribund with disease: Tet-on control mice where BCR-ABL expression was suppressed (n=3); Tet-off CML mice had BCR-ABL expression that induced disease that mimics human chronic phase (CP) CML (n=6); a Tet-off, Tet-on (TOTO) cohort where BCR-ABL expression allowed disease development and was then suppressed by turning Tet-on to simulate a hypothetical best-case treatment scenario (n=4); and TKI cohort to simulate a clinical setting where BCR-ABL expression was induced (Tet-off) and remained on during and after a four week nilotinib treatment window (n=7).