Background
Biliary tract cancers (BTC) represent an important rare cancer type in Veterans. The heterogeneity of BTC has revealed distinct molecular subtypes, however a majority of patients remain without precision-based targeted therapeutics. Epigenomic remodeling has been considered as a shared mechanism of therapeutic resistance. Cyclin dependant kinase 7 (CDK7) is an emerging therapeutic target that functions by phosphorylation of RNA polymerase II and cell cycle progression. Here, we investigate CDK7 inhibition using small molecule inhibition (SY-5609) across a panel of BTC organoid models.
Methods
PCOs were expanded from patient-derived tissues and shared models provided from the NCI. Organoid response was tracked from growth using Z-stacked high content imaging (Cytation5) to track individual organoid growth and established viability markers of Caspase-3/7 (C3/7) and ToPro3, for induced apoptosis and necrosis for phenotypic screening. Treatment groups included media control, positive control (cycloheximide) 200uM continuous, gemcitabine (gem) 10uM 24h, cisplatin (cis) 5uM 48h, combination gem+cis, and SY-5609 10nM 144h. Glass’s delta was used to standardize effect size relative to media control.
Results
Patient-derived cancer organoids were generated across four unique models including pathogenic (A-B) IDH1 p.R132G, (C) FGFR2-HPGDS fusion and (D) non-targetable molecular profile (CCNE1 amplified, BRCA1 splice variant). In the non-targeted model, CDK7 inhibition achieved growth arrest +2.0% (SY-5607) v. +43.0% (media control) with effect size >1.1. This response was similar to standard of care gem+cis with growth of +1.5% and augmented using the combination of gem+SY-5609 -3.1% with effect size of >1.3. When treated with CDK7 inhibition, persistent growth was seen across models of IDH1 mutant and FGFR2-HPGDS2 fusion cancers. High content imaging revealed subclonal populations with failed induction of apoptosis and necrosis at 144h, suggestive of the critical need to address intrinsic resistant populations to both SOC chemotherapy and novel targeted strategies.
Conclusions
Across a diversity of BTC cancer models, CDK7 inhibition was found to achieve growth arrest in a CCNE1 amplified cancer model. High content imaging of organoids can identify subclonal resistant populations as a critical unmet need in future therapeutic development. Ongoing work is adapting these techniques to multiple small molecule inhibitors that target transcription including EZH1/2 and CDK9.