Use of Ex Vivo Patient-Derived Tumor Organotypic Spheroids to Identify Combination Therapies for HER2 Mutant Non-Small Cell Lung Cancer

Purpose: Evaluating drug responses using primary patient-derived cells ex vivo represents a potentially rapid and efficient approach to screening for new treatment approaches. Here, we sought to identify neratinib combinations in HER2 mutant non-small cell lung cancer (NSCLC) patient xenograft-derived organotypic spheroids (XDOTS) using a short-term ex vivo system. Experimental Design: We generated two HER2-mutant NSCLC PDX models [DFCI359 (HER2 exon19 755_757LREdelinsRP) and DFCI315 (HER2 exon20 V777_G778insGSP)] and used the PDX tumors to generate XDOTS. Tumor spheroids were grown in a microfluidic device and treated ex vivo with neratinibbased drug combinations. Live/dead quantification was performed by dual-labeling deconvolution fluorescence microscopy. The most efficacious ex vivo combination was subsequently validated in vivo using the DFCI359 and DFCI315 PDXs and a HER2(YVMA) genetically engineered mouse model. Results: Both neratinib and afatinib, but not gefitinib, induced cell death in DFCI359 XDOTS. The combinations of neratinib/trastuzumab and neratinib/temsirolimus enhanced the therapeutic benefit of neratinib alone in DFCI315 and DFCI359. The combination of neratinib and trastuzumab in vivo was more effective compared with single-agent neratinib or trastuzumab and was associated with more robust inhibition of HER2 and downstream signaling. Conclusions: The XDOTS platform can be used to evaluate therapies and therapeutic combinations ex vivo using PDX tumors. This approach may accelerate the identification and clinical development of therapies for targets with no or few existing models and/or therapies.