Abstract 990: Single-cell profiling of small cell lung cancer circulating tumor cell-derived xenograft models reveals intratumoral heterogeneity among mediators of chemoresistance

Abstract

Abstract Small cell lung cancer (SCLC) accounts for 14% of lung cancer diagnoses in the United States and is characterized by rapid onset of chemoresistance and poor clinical outcomes. Once considered a homogeneous disease, recent analyses of SCLC have identified intratumoral heterogeneity (ITH) with respect to NOTCH signaling, ASCL1/NEUROD1 balance and MYC amplification - all of which are potential mechanisms underlying SCLC's aggressive and refractory biology. Unfortunately, patient-derived models of SCLC with which to better characterize the molecular profiles of refractory SCLC are scarce. To address this, we generated circulating tumor cell-derived xenograft (CDX) models from liquid biopsies of patients with treatment-naïve or relapsed SCLC. Each CDX model underwent pathological review to confirm tumors were consistent with SCLC based on histology and standard immunohistochemical markers (e.g., TTF1, chromogranin A, synaptophysin, NCAM). Sequencing of these models revealed mutations typical of SCLC (e.g. TP53, RB1), which were maintained in vivo over multiple passages. Importantly, each model's in vivo response to cisplatin matched the patient's platinum response at the time of CDX generation. At the proteomic level, platinum-resistant models exhibited mTOR activation, increased SOX2 and ATM, and reduced E-cadherin, suggesting a shift toward EMT and cancer stem cell expansion may contribute to resistance. To investigate ITH, we analyzed single-cell gene expression profiles by RNAseq using a droplet-based Chromium Single Cell system that analyzed a filtered subsample of 2000 cells per tumor. Consistent with SCLC, all CDX models contained large numbers of cells expressing neuroendocrine-specific genes (SYP, CHGA). However, Principle Component Analysis revealed that cells from chemosensitive CDX models had distinct expression profiles from resistant models. Using our published EMT gene signature, we found that resistant models had higher proportions of mesenchymal (vs. epithelial) cells. Several other distinctions between sensitive and resistant models were detected at the single-cell level but not in bulk RNA and protein analyses, suggesting that single-cell resolution can identify occult platinum-resistant subpopulations. For example, higher proportions of ASCL1- and DLL3-expressing cells were associated with platinum sensitivity, whereas a shift toward predominant NEUROD1-expression was observed with resistance. Cells expressing each of these three genes were identified across all tumors, suggesting platinum-sensitive and resistant subpopulations are ubiquitous but that even subtle shifts in the fractional distribution of these subsets can exert significant impact on response. These data support further use of single-cell analysis to explore the role of ITH as a driver of drug resistance in SCLC. Citation Format: C. Allison Stewart, Carl M. Gay, Yuanxin Xi, Siva V., Junya Fujimoto, Pan Tong, Lixia Diao, Lerong Li, Mohan Bolisetty, Neda Kalhor, Patrice Lawson, Mayra Vasquez, Hai Tran, Ignacio I. Wistuba, Bonnie Glisson, Jianjun Zhang, Stephen G. Swisher, Jack A. Roth, John V. Heymach, Paul Robson, Jing Wang, Lauren A. Byers. Single-cell profiling of small cell lung cancer circulating tumor cell-derived xenograft models reveals intratumoral heterogeneity among mediators of chemoresistance [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 990.