Molecular Characterization of Circulating Tumor Cells for the Earlier Detection and Longitudinal Monitoring of Hepatocellular Carcinoma

Abstract

Over ninety percent of cancer mortality is attributable to metastasis, most commonly due to the blood-borne dissemination of cancer cells from a primary tumor to secondary tissues. Circulating tumor cells (CTCs) offer both a window into understanding the fundamental underpinnings of this metastatic cascade and the potential to non-invasively and serially sample tumor-derived tissue for the early detection, longitudinal monitoring, and prediction of therapeutic response in cancer. Work by our laboratory has overcome some of the key challenges inherent to isolating these extraordinarily rare cells using microfluidic cell enrichment technologies. However, the ability to rapidly and accurately quantify and characterize CTCs has served as a major bottleneck toward broader clinical implementation of the technology. Recently, digital PCR (dPCR) has enabled the absolute quantification of a set of RNAs of interest, even while admixed with large excess of contaminating non-target RNAs. Combining dPCR with our microfluidic device facilitated the highly accurate tissue lineage-based cancer detection from blood specimens from patients with a variety of solid tumors. In hepatocellular carcinoma (HCC) a signature of 10 liver-specific transcripts was highly predictive for the presence of cancer in 9/16 (56%) newly diagnosed, untreated patients with HCC versus 1/31 (3%) controls with liver cirrhosis at high risk for developing HCC (P<0.0001). Similarly, in a group of 34 men with localized prostate cancer, a prostate-specific dPCR-CTC signature predicted the presence of microscopic dissemination with high specificity (P<0.001). In a cohort of 54 women with high risk localized breast cancer, elevated CTC-Scores greater than 3 weeks during neoadjuvant treatment predicted significant residual disease at the time of subsequent surgical resection (p < 0.047). Taking RNA-based quantitation of CTC-enriched bulk populations to the next analytic level, single cell RNA sequencing of the microfluidically enriched product defined the cellular origin of both individual CTCs and contaminating leukocytes. In HCC patients, novel RNA-based CTC biomarkers were identified from the inferred copy-number alterations present in cancer cells and validated in a separate cohort of HCC and CLD patients (p<0.03). This thesis serves as a proof of principle for the scalable, high-throughput use of single-cell RNA sequencing to inform digital PCR-enabled CTC detection of cancer, with potential applications for early detection, longitudinal monitoring, and prediction of therapeutic response.