Analysis of novel therapeutic options for adenoid cystic carcinoma of the salivary glands

Abstract

Adenoid cystic adenocarcinoma (ACC) of the salivary glands is characterised by slow growth, frequent metastasis and low survival rates. ACC is resistant to conventional chemotherapy and currently, no systemic agent has proven effective. MYB activation is a hallmark of ACC, with the most common genetic aberration being the MYB-NFIB translocation, occurring in 50-60% of cases. Research has been hindered by the lack of validated pre-clinical ACC models, hence there is an urgent need to develop relevant models to test novel therapeutic options. Therefore, I aimed to establish patient-derived xenograft (PDX)-derived pre-clinical in vitro models, and to identify potential drug targets and therapies based on the properties of eight sequenced metastatic ACC patient tumour samples. The success rate of ACC PDX establishment was 54.5%; three of which possessed the MYB-NFIB translocation, one of the MYBL1-NFIB translocation, and the other two expressed high MYB via alternative mechanisms. Using ACC PDX-derived primary cells I successfully established and developed culture conditions for 2D growth and 3C ACC organoid models that maintain high MYB expression and histologically recapitulate the PDX tumours derived from. I optimise 2D and 3D drug sensitivity assays to assess potential therapies for ACC, targeting genes and pathways identified using the patient tumour sequencing data. A promising drug combination of a CDK2, 5, 9 inhibitor and bromodomain inhibitor was highly synergistic in the PDX-derived models. A short-term in vivo drug assay in one PDX line identified engrailed-1 (EN1), MYC and SOX9 as potential candidate biomarkers for therapeutic activity of the drug combination, and Ki-67 staining identified that the combination treatment significantly reduced the proportion of actively proliferating cells. On-going long-term in vivo assays will determine whether this promising combination inhibits ACC growth. Here I discuss the optimisation and establishment of relevant PDX-derived ACC 2D and 3D organoid models representing MYB-NFIB and MYB driven ACC subsets. I also demonstrate the application of these models by identifying a promising novel drug combination for ACC that could potentially inform clinical trials.