Abstract B39: Generation of head and neck cancer patient-derived xenografts with in vivo acquired cetuximab resistance

Abstract

Abstract Background: Although cetuximab has demonstrated only modest response rates in head and neck squamous cell carcinoma (HNSCC) patients in clinical studies, most HNSCC cell line xenograft models have demonstrated sensitivity to cetuximab. We sought to characterize molecular and phenotypic changes associated with the acquisition of cetuximab resistance by evaluating patient-derived xenograft (PDX) models with and without cetuximab selection in vivo. Methods: A single PDX was established from a patient tongue HNSCC and a separate PDX from an independent patient lymph node by subcutaneous propagation in nod scid gamma (NSG) immunodeficient mice. Established PDXs were implanted into 2 flanks of NSG mice, and mice were either untreated (parental) or treated with cetuximab using modulated dosing dependent upon tumor growth (0.04-0.2 mg 1-2 times weekly) with the goal of PDX growth at the maximum dose. Tumor histology and short tandem repeat (STR) profiling were evaluated for HNSCC tumors and paired PDX models. Cell lines were derived from PDX tumors grown in NSG hypoxanthine-guanine phosphoribosyltransferase (Hprt)-deficient mice, allowing negative selection of mouse fibroblasts in vitro. Cell lines were retrovirally transduced to express either green fluorescent protein (GFP) or telomerase (HTERT). Cell lines were evaluated for presence of human cells using a PCR-based assay to detect human amelogenin alleles (sex id assay). Candidate gene mutations and protein/phosphoprotein levels will be evaluated in parental and cetuximab-selected PDXs using Ion Torrent and immunoblotting technologies, respectively. Results: Patient HNSCC tumors and paired parental PDXs displayed similar histologies and consistent STR profiles. Four of 14 PDXs derived from the tongue HNSCC displayed growth in the presence of cetuximab at the maximum dose, and of 18 PDXs derived from lymph node, 3 grew in the presence of cetuximab at the maximum dose. Average tumor volume at treatment initiation was 134 mm3 and average duration of treatment was 8 months. For tumors that eventually grew in presence of maximum dose cetuximab, initial tumor volume and duration of treatment were 156.5 mm3 and 9 months, respectively. Cetuximab-resistant tumors were heterogeneous regarding their resistance to cetuximab in a subsequent passage. Of the two cetuximab-resistant tumors STR evaluated, 1 was consistent with the patient and parental PDX and 1 could not be genotyped. In vitro cell cultures were established from parental PDX tumor cells transduced with HTERT and GFP, and these cells were confirmed to be of human origin by the sex id assay. Molecular profiling and establishment of cell lines derived from cetuximab-resistant, genotype confirmed PDXs are currently ongoing. Conclusions: It was possible to obtain PDX tumors grown under conditions of chronic treatment with cetuximab. However, dose modulation was required and propagation of the trait of cetuximab-resistant growth upon PDX passage was not guaranteed. The acquisition of cetuximab-resistant PDX tumors was a lengthy process that required care regarding dosing modulation and genotype validation. Once established, these models will provide valuable insights regarding molecular and phenotypic changes associated with acquired cetuximab resistance. Citation Format: Pei Zhou, Hua Li, Sarah Wheeler, Jennifer R. Grandis, Laura A. Stabile, Ann Marie Egloff. Generation of head and neck cancer patient-derived xenografts with in vivo acquired cetuximab resistance. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Drug Sensitivity and Resistance: Improving Cancer Therapy; Jun 18-21, 2014; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(4 Suppl): Abstract nr B39.