Liquid Biopsy Solutions for Noninvasive Molecular Profiling

Liquid biopsy offers the potential to understand and treat cancer in ways never before possible. The concept is simple: collect tumor DNA after it enters the peripheral blood circulation and identify treatable biomarkers. This has been attempted for decades with varying degrees of success. In the past few years, technologies focused on enrichment, next-generation sequencing (NGS) and bioinformatics have pushed the liquid biopsy method closer to the clinic. This article focuses on the benefits, challenges and methodology for implementing liquid biopsy in the oncology setting.

Benefits of liquid biopsy

Cancer is a disease of progressive mutations; even tumors of the same origin (e.g., breast, lung, prostate) have remarkable heterogeneity. Knowing a tumor’s molecular profile can be critical to aligning the patient with a therapy from which he or she is likely to benefit. Many patients start with an initial tissue biopsy acquired through surgery or needle biopsy. This tissue is then profiled using sequencing or other molecular methods to analyze genes with known relationships to therapeutic response.

Adding to the complexity is that the tumor cells from an individual patient can change over time because they evade first-line therapies and the disease progresses. Serial tissue biopsies are not performed routinely for many reasons, such as lack of primary tumor tissue to access, difficulty in accessing the tumor or metastases, patient trauma and expense. The primary advantage of liquid biopsy is its ability to access tumor content at all critical inflection points, such as at the start of a new therapy or temporal monitoring of an existing course of treatment. This is facilitated by a conventional blood draw that makes the sample collection process more tolerable for the patient.

Challenges in tumor DNA enrichment and extraction

Tumor DNA enters the peripheral blood circulation via intact cells (circulating tumor cells, CTCs) and as DNA fragments (cell-free DNA, cfDNA), both of which originate from the solid tumor. These sources of tumor DNA are in low abundance in the peripheral blood. CTCs exist as low as 1 per billion blood cells, and a single tube of blood (10 cc) contains from 0 to a few hundred CTCs. Detection of cell-free DNA in plasma also poses a significant challenge as tumor DNA is heavily diluted by DNA from normal cells to abundances that can be as low as 0.01%.

The high level of wild-type background in these samples presents a challenge for genomic analysis. Most NGS workflows optimized for tissue analysis maintain 1–5% cutoffs for calling mutations. This presents two key challenges for the analysis of liquid biopsy samples: 1) enrichment of the tumor content to the highest level possible, and 2) optimization of the molecular analysis workflow to improve detection sensitivity. Most efforts in the CTC field have centered on optimizing the tumor cell enrichment process, while in cfDNA the focus has been on the detection workflows. As these techniques become more refined, it is possible to combine enrichment and analysis workflows to increase detection sensitivity.

CTC enrichment

Figure 1 – IsoFlux system for CTC analysis.

CTCs are widely believed to be the cells involved in the spread of cancer to distal sites in the body, known as metastases. They enter the blood circulation from the solid tumor and are present in very low concentrations. Much effort has been spent trying to enrich CTCs to levels that can be analyzed with modern genomic tools. The IsoFlux system from Fluxion Biosciences (South San Francisco, Calif.) (Figure 1) automates the enrichment process of CTCs from peripheral blood and prepares them for genomic analysis.

The system utilizes immunomagnetic enrichment to preferentially separate CTCs based on the surface expression of both epithelial and mesenchymal markers. Antibodies for EpCAM, EGFR, N-cadherin, Her2 and other markers defined by the user are added to magnetic beads that are mixed with the blood sample. A combination of antibodies is used to cast a wider net for all types of CTCs, since these cells are known to alter their expression of individual markers over time.

Figure 2 – IsoFlux microfluidic cartridge.

The sample is loaded into a microfluidic cartridge that allows the cells to flow through the capture zone in a laminar flow profile (Figure 2). As the cells pass through the capture zone, the magnetic-bead labeled CTCs are captured by a neodymium magnet that sits atop the cartridge (Figure 3). The unlabeled white blood cells continue along the direction of the flow profile into a waste reservoir. When the sample has fully passed through the channel (about 40 minutes), the CTCs are collected from the removable cap with a pipet and are retained for molecular profiling.

Figure 3 – Schematic of the CTC capture zone.

DNA is collected from the CTCs using the IsoFlux NGS DNA kit. The cells are placed across a purification column that further depletes leukocytes from the sample and collects the lysate of the retained cells. The lysate can then either be purified directly for gDNA or amplified, depending on the input requirements of the NGS workflow.

CTC analysis using next-generation sequencing

The final purity of the CTC samples is in the 5–25% range of tumor content. The samples can then be analyzed with conventional sequencing hardware and content panels. Targeted sequencing panels are used in clinical research because they provide an optimal balance of gene coverage, cost and throughput.

In these workflows, the CTC gDNA is enriched for specific regions of the genome known to contain cancer-associated mutations, such as KRAS, EGFR and BRAF. The enriched product then undergoes standard library preparation and indexing to permit many samples to be run simultaneously.

IsoFlux CTC samples can be sequenced using instruments such as the Illumina MiSeq (San Diego, Calif.) and Ion Torrent PGM (Personal Genome Machine, Life Technologies, South San Francisco, Calif.). The sequencing workflow is optimized for low-frequency allelic detection, and entails running the samples with deep coverage using a bioinformatic workflow tuned to detect mutations at the <1% level.

Analytical performance of NGS workflow

Spiked samples were prepared to characterize the NGS workflow performance using known concentrations of target mutations. The MDAMB- 231 cell line was added to healthy donor peripheral blood tubes (10 cc) in concentrations of 0, 3, 7 and 15 cells/mL. Chosen for its mesenchymal-like properties, including very low EpCAM expression, to more closely approximate clinical samples, this cell line harbors mutations in the KRAS, BRAF and TP53 genes. The analytical samples were enriched using the IsoFlux system and gDNA collected with the IsoFlux NGS DNA kit. gDNA was quality controlled using qPCR measurements (RNase P Kit, Life Technologies). Parallel samples at each concentration were analyzed using fluorescence microscopy to enumerate the number of CTCs recovered and to determine the expected allelic frequency.

Target enrichment was performed with the TruSight tumor panel (Illumina), which is optimized for low-frequency mutation detection and contains 174 amplicons across 26 cancer-associated genes. Following library preparation and indexing, the eight samples were sequenced on a single lane using the MiSeq sequencer. This yielded 16,000× mean coverage, with 98.4% of the panel above 3200× coverage (20% of the median) and 99.7% above 2000× (lower limit for 1% detection to have 20 reads on the mutant allele).

All three variants were detected in each concentration of the spiked samples, with no false positive readings and no variants detected in the negative controls (0-cell spike in). The resulting allelic frequency of the homozygous TP53 mutation aligned well with the expected tumor cell purity measured in the enumeration sample (Figure 4).

Figure 4 – Analytical NGS results using the TruSight tumor panel and MiSeq.

 

Liquid biopsy for clinical use

As we learn more about tumor biology and treatment-resistant mechanisms, the need for timely, patient-specific information grows. Potential applications for CTC molecular analysis are:

  • Predicting response to molecular anticancer therapies
  • Detecting treatment-resistant molecular profiles (i.e., patient stratification)
  • Monitoring treatment to evaluate efficacy
  • Guiding prognosis
  • Improving biological understanding and new drug target identification.

Conclusion

Liquid biopsy offers timely availability of tumor tissue, which has limited the utility of molecular analysis techniques such as NGS. Despite the easy access to peripheral blood in the clinic, these types of samples present challenges in their subsequent enrichment and analysis. The IsoFlux system delivers enriched CTCs that maintain the level and purity required for use with conventional NGS hardware and content panels. CTC enrichment followed by NGS analysis creates a powerful workflow that can be used to better track cancer patients.

Additional reading

  1. Alix-Panabières, C.; Pantel K. Circulating tumor cells: liquid biopsy of cancer. Clin. Chem. 2013 Jan, 59(1), 110–8.
  2. Harb, W.; Fan, A. et al. Mutational analysis of circulating tumor cells using a novel microfluidic collection device and qPCR assay. Transl. Oncol. 2013 Oct, 6(5), 528– 38.
  3. Krebs, M.G.; Hou, J.-M. et al. Circulating tumour cells: their utility in cancer management and predicting outcomes. Ther. Adv. Med. Oncol. 2010 Nov, 2(6), 351–65.

Michael Schwartz is VP marketing, Fluxion Biosciences, Inc., 385 Oyster Point Blvd. #3, South San Francisco, Calif. 94080, U.S.A.; tel.: 650-241- 4777; e-mail: [email protected]www.fluxionbio.com

Comments