Molecular biomarker analyses using circulating tumor cells

Abstract

Background: Evaluation of cancer biomarkers from blood could significantly enable biomarker assessment by providing a relatively non-invasive source of representative tumor material. Circulating Tumor Cells (CTCs) isolated from blood of metastatic cancer patients hold significant promise in this regard.

Methodology/principal findings: Using spiked tumor-cells we evaluated CTC capture on different CTC technology platforms, including CellSearch and two biochip platforms, and used the isolated CTCs to develop and optimize assays for molecular characterization of CTCs. We report similar performance for the various platforms tested in capturing CTCs, and find that capture efficiency is dependent on the level of EpCAM expression. We demonstrate that captured CTCs are amenable to biomarker analyses such as HER2 status, qRT-PCR for breast cancer subtype markers, KRAS mutation detection, and EGFR staining by immunofluorescence (IF). We quantify cell surface expression of EGFR in metastatic lung cancer patient samples. In addition, we determined HER2 status by IF and FISH in CTCs from metastatic breast cancer patients. In the majority of patients (89%) we found concordance with HER2 status from patient tumor tissue, though in a subset of patients (11%), HER2 status in CTCs differed from that observed in the primary tumor. Surprisingly, we found CTC counts to be higher in ER+ patients in comparison to HER2+ and triple negative patients, which could be explained by low EpCAM expression and a more mesenchymal phenotype of tumors belonging to the basal-like molecular subtype of breast cancer.

Conclusions/significance: Our data suggests that molecular characterization from captured CTCs is possible and can potentially provide real-time information on biomarker status. In this regard, CTCs hold significant promise as a source of tumor material to facilitate clinical biomarker evaluation. However, limitations exist from a purely EpCAM based capture system and addition of antibodies to mesenchymal markers could further improve CTC capture efficiency to enable routine biomarker analysis from CTCs.

Figure 1. Evaluation of EpCAM as a marker for capturing CTCs.

A. EpCAM expression in breast cancer cell lines grouped by molecular subtype. B. Expression of EpCAM in relation to other epithelial and mesenchymal markers in breast cancer cell lines. C. EpCAM expression in different tumor types and in white blood cells (WBC). D. Spike-in CTC recovery on CellSearch® (left panel) in EpCAM high and EpCAM low cells when analyzed at 24 h or 48 h post spike-in at Genentech or two reference labs (Ref Lab). In the EpCAM-high (24 h) group, samples are graphed by individual spike-in cell count (10–500 cells) and pooled together for all other groups.

Figure 2. EGFR IF and HER2 FISH in CTCs.

A. mRNA expression of EGFR (diamonds) or EpCAM (bars) in NSCLC cell lines. IHC scores for EGFR from tissue microarrays are indicated below. B. EGFR immunofluorescence (IF) scoring criteria for CTCs. For each scoring level, the range of high and low expression are shown. C. EGFR IF scoring of spiked tumor cells isolated from blood. The weighted H-score from CTC analysis and corresponding IHC score for that cell line is listed below for each sample. D. HER2 FISH assay in captured SKBR3 cells on the OncoCEE microchannel platform. Cells are stained with anti-cytokeratin antibody (green), DAPI (blue), FISH probes against HER2 (red dots) and a centromeric probe, CEP17 (green dots).

Figure 3. qRT-PCR assay for molecular subtyping of breast cancer in CTCs.

A. 100 cells were spiked into normal donor blood from a luminal (T47D), HER2+ (SKBR3) and a basal-like breast cancer cell line (HCC70) or negative control (WBC) and were isolated using the CellSearch® platform, followed by qRT-PCR analysis with a panel of genes specific for the three corresponding breast cancer subtypes. Heatmap shows hierarchically clustered z-score normalized Ct values for each gene.

Figure 4. KRAS mutation detection assay in CTCs.

A. Schematic of digital PCR based mutation detection assay. B. Results of the KRAS G12C mutation assay on Digital PCR arrays starting from DNA isolated from KRAS mutant tumor H2122 cells spiked into whole blood in the indicated numbers.

Figure 5. EGFR expression and HER2 status in CTCs from metastatic cancer patients.

A. EGFR scoring in CTCs from metastatic lung cancer patient samples, scored according to criteria described in Figure 2B. B. CTC counts from 39 metastatic breast cancer patient samples listed by subtype. C. HER2 IF scoring criteria showing the high and low range of expression. D. Quantification of HER2 scoring in CTCs. H-score for HER2 IF in CTCs, HER2 status in patient tumor, CTC counts and HER2 FISH results in CTCs are listed in table below. HER2 FISH was performed on CellSearch (CS) or on the OncoCEE microchannel (OncoCEE). For HER2 FISH, (+) indicates a positive FISH result, (-) indicates a negative result and (f) indicates that the FISH failed because no CTCs were detected.