The Promise of Liquid Biopsy Technology

Examining the potential of such diagnostic technologies in reshaping oncology trials.

Diagnosing cancers at earlier stages leads to better prognoses for patients,^1,2,3 yet such detective work is constrained by the mechanics and specificity of current methods. Patients’ medical histories, physical exams and results of screening tests, such as those involving body fluids, images and tissue samples, all help in the diagnosis of cancer. Each of these diagnostic standards has significant limitations, particularly given the dynamic ability of cancer biology to change within and among a patient’s cancer cells as the disease progresses, spreads and responds to treatment. Liquid biopsies are aggressively addressing such barriers and may help to revolutionize cancer research, patient treatment and survivor care.

Diagnosing cancer today: Focus on the physical

Traditional classification of cancer relies on designating the histological type and primary location of the originating tissue.⁴The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) groups cancers into six histological types: carcinoma, sarcoma, myeloma, leukemia, lymphoma and mixed types. ICD-O-3 also uses 10 location groups ranging from connective tissue to muscle to blood and lymphoid cells.

Cancers also are classified by their growth stage. The American Joint Commission on Cancer (AJCC) and the Union for International Cancer Control (UICC) created the universally used tumor–node–metastasis (TNM) staging system for solid tumors.⁵ Combination of the three scores creates an overall stage ranging from 0 to IV, with the possibility of additional clarifying parameters such as whether lymphatic vessels or veins are involved.

Of note, each cancer has a precise definition of T, N and M, so identical TNM scores for a patient with breast cancer and one with prostate cancer do not mean the cancers are of the same stage.

Solid tumors also are graded based on how abnormal their cells look microscopically, and usually are described as well-differentiated, undifferentiated or poorly differentiated, based on how closely the cells resemble normal tissue.⁶

Today, oncologists can go well beyond the physical aspects of cancer to help diagnose the nature of an individual patient’s disease. The body’s healthy cells can react and release proteins in the presence of cancer cells, which can serve as markers when found in body fluids or tissues. Some biomarkers occur for only certain tumors, while others are known to occur with several cancer types. The American Society of Clinical Oncology has published guidelines that outline what tumor markers may be used in the diagnosis of breast cancer, colorectal cancer and lung cancer, as well as other cancers. For example, blood levels of alpha-fetoprotein (AFP) are used to diagnose and determine treatment response in patients with liver cancer, and blood levels of chromogranin AFP can be elevated in patients with certain neuroendocrine tumors, small cell lung cancer or prostate cancer. The epidermal growth factor receptor (EGFR) is used as a marker for several cancers because of its role in cell division and high occurrence on the surface of cancer cells. When the protein known as epidermal growth factor attaches to EGFR, the resulting cascade of signals prompts cells to divide.

Getting more with markers

However, for more than 20 years, scientists have looked inside cancer cells to examine the utility of their genetic material. Somatic and inherited mutations found in precancerous or malignant tissue, but not present in healthy tissue, have been used as biomarkers, which has fundamentally changed clinical practices regarding patients with colon, breast, lung and other cancers. For example, the inherited mutations to the BRCA1 gene, which normally down-regulates cell growth, can increase the risk of developing breast, ovarian, prostate and other cancers.⁷

The application of genetic mutation information, rather than originating organ, is particularly helpful to address the treatment of patients with recurrent or advanced cancers. But the current processes to find these mutations have many limitations. Imaging tests (e.g., X-rays, CT, MRI, PET scans, mammography or ultrasound) can identify masses, but they cannot find microscopic metastases nor characterize a solid tumor’s cellular composition. For that, a sample of tissue is removed using a needle, endoscope or surgery and prepared, either as formalin-fixed paraffin-embedded (FFPE) or frozen samples. These tissue biopsies enable analysis histologically for cell shape, location and concentration, as well as genetically for mutation composition. But they are labor-intensive (even with computer assistance), with processes that involve the personal expertise of a pathologist and, hence, the possibility of reproducibility errors.

In an idyllic trial scenario, if a primary cancer has metastasized, physicians would take patient tissue biopsies at different locations and times, but this is not possible in many cases. Not all patients would agree to repeated use of such invasive procedures, and some tissue biopsies cannot be pursued if it is too risky for a patient’s health. Tissue sampling can cause complications for patients (e.g., prostate biopsies can result in fever, bleeding, infection and other complications). Moreover, not all cancers are readily accessible for biopsy, particularly brain tumors.

Tissue biopsies, by their nature, are a limited resource, as each test consumes part of the sample, creating an evidence supply issue for sponsors of long-term clinical studies. Significantly, each biopsy captures just one place at one moment and, therefore, individually do not represent the breadth of cancer heterogeneity possible within a patient, particularly if he or she has begun treatment. The evolution of cancer as it either metastasizes or responds to treatment yields changes genetically with end-effects on cellular, tissue or organ levels.

A single biopsy is simply “not representative of the mutational landscape of the entire tumor bulk,” as the authors of a milestone 2012 New England Journal of Medicine study reported.⁸ When comparing biopsies from kidney cancer tumors to complete tumor tissues after their surgical removal, the NEJM authors found a single biopsy had, on average, 70 mutations, which represented only 55% of all the mutations in the excised tumor from which the biopsy had been made. The biopsy missed nearly half of the potential genetic guideposts to patient treatment. Moreover, only 34% of identified mutations had distributed throughout the tumor. The authors concluded such single tumor-biopsy samples can lead to underestimating a tumor’s genetic composition, presenting “major challenges to personalized-medicine and biomarker development.”⁹

Processing and analyzing tissue biopsies are time-consuming and may genetically alter the tissue, causing erroneous interpretation if formalin fixation is used. The wait for results of a typical tissue biopsy can be a few days after the laboratory receives the sample, but longer if specialized handling is required. A recent study reported a median of 27 days from ordering to results of tissue biopsies from NSCLC cancer patients with acquired resistance vs. a median of 12 days for those newly diagnosed vs. a median of just three days from blood draw to results for liquid biopsies.¹⁰ Tissue type affects timelines, as hard tissues such as bone take more time due to treatments to remove minerals and restore softness to enable sample slicing for analyses. Notably, if the presiding pathologist seeks a second opinion to review the samples, more time will be added before the clinical team and patient receive results. Additionally, the costs of tissue biopsies can be significant, both in clinical trials and current clinical practice.

Ready for liquid biopsy upgrades

In contrast, liquid biopsies offer specificity, efficiency, scalability and are less invasive for patients. The technology also may aid in treatment selection during routine clinical care, monitoring medication effects such as drug resistance or tumor evolution, identifying recurrent or minimally residual disease and, ideally, finding cancers in their most nascent stages and informing prognoses (see chart). The same benefits are applicable in the clinical trial setting, including screening patients for trial enrollment. The potential of liquid biopsies to detect changes in tumor genetics well before imaging reveals changes in growth could enable therapy modifications or earlier second-line interventions.

Liquid biopsies can be taken and analyzed quickly-the cobas® EGFR Mutation Test v2 is reported to take less than four hours. Clinical trial teams can reassess patients’ responses to treatments with each blood draw, catching tumor progression earlier than current practices, which can involve waiting weeks after treatment to use imaging to determine tumor shrinkage. Moreover, this technology can look for large collections, often hundreds at a time, of diverse mutations. For trial sponsors, an ability to find “needles in haystacks” could result in finding more qualified candidates for trials, speeding recruitment and increasing the likelihood of trial success.

Hunting mutations with liquid biopsies

Almost 40 years ago, in an early study evaluating DNA in the blood of cancer patients, scientists predicted: “DNA in the serum may be an important tool for the evaluation of therapy or the comparison of different regimens.”¹¹ Today, liquid biopsy technology can leverage blood or other body fluids, such as urine, saliva, or cervical fluid, because all of the body’s cells emit genetic information.

These cells and particles may result from normal secretions, from the unorganized death of tissues that occurs during traumatic events such as stroke or heart attacks, or from apoptosis, the process of programmed, or organized, cell death. The resulting mix within a patient with cancer includes the debris of healthy or malignant cells in the form of circulating tumor cells (CTC), cell-free circulating DNA (cfDNA), exosomes, extracellular vesicles (EVs) and microRNA (miRNA).

Circulating tumor cells
Tumors cast off CTCs into the blood stream, potentially seeding metastasis. However, shedding frequency is low: in one milliliter of blood, among about 70 quadrillion white blood cells and 50 quadrillion red blood cells, one to 10 analyzable CTCs can be isolated.

Not all CTCs are genetically cancerous cells, so testing platforms must have refined search criteria to identify malignant cells, much like the immune system recognizes “foreigners.” Different technologies are improving the efficient selection of such cells, notably a microfluidic platform that can sort rare CTCs from whole blood samples. The advancement of such platforms has prompted great interest in CTCs, perhaps second only to shorter circulating tumor DNA (ctDNA) and some investigators apply both in studies.

The fragility of CTCs requires fast testing and precludes long-term storage. For example, Janssen’s CellSearch® CTC test, the first actionable CTC test approved by the FDA for use in patients with metastatic breast, prostate or colorectal cancer, requires sample processing within 96 hours of collection.¹²Cell-free circulating DNA/circulating tumor DNA
The bits of DNA (about 150 to 180 base pairs long) that comprise cfDNA have known utility in both cancer diagnostics and in non-invasive prenatal testing (NIPT) and transplantation. Only one cfDNA test, ColoGuard® from Exact Sciences, has received FDA approval to screen for cancer-and, though not technically a liquid biopsy, it functions in a similar way using stool, rather than blood, collecting genetic evidence of disease shed from tumors and adenomas as it travels through the large intestine.

In cancer patients, detection of shorter ctDNA fragments is correlated with more plentiful mutations, while greater quantities correlate with malignancy. Detectable ctDNA significantly varies by and within tumor type. Tumors release ctDNA as their cells die, usually because the malignancy’s rapid growth outpaces the ability of the blood supply to provide nutrients and oxygen. The dead cells are “cleaned up” by the phagocytosis of immune system cells, which then cast off the DNA snippets and other debris.

Mutant ctDNA have documented variations in concentrations, up to more than 100,000 copies in one milliliter (mL) of plasma, half-lives from 15 minutes to about two hours, but can be found in stored patient fluids, such as frozen plasma. A challenge for ctDNA analytics is to sort mutations from normal, prevalent cfDNA.

Also, analyses of ctDNA can detect a tumor’s original DNA changes, including mutations and abnormalities, but not the tumor RNA transcriptome or proteome. For a patient having received a definitive therapy to remove a tumor with curative intent, such as chemotherapy, radiation, immunotherapy or surgery, ctDNA has potential to identify occult disease, when minimal residual disease is present. Currently, determining if minimal residual disease exists involves monitoring up to five to 10 years and typically uses imaging diagnostics, which do not readily reveal microscopic disease, or protein biomarkers that may not be specific to a cancer type or status.

Exosomes and extracellular vesicles (EVs)
Exosomes are tiny sacks or vesicles that cells release and can stably transport a mix or individual pieces of RNA, DNA or proteins. Exosomes can be found in blood serum, plasma, saliva, urine, cerebrospinal fluid and other biofluids, promising applications beyond cancer such as for inflammatory, metabolic, cardiovascular and neurodegenerative diseases.

The documented roles of exosomes in cancer pathogenesis include tumor growth and angiogenesis stimulation and immune response suppression. Tumor cells cast off thousands of exosomes daily, creating plasma concentrations reaching 10 quadrillion per mL. Smaller than CTCs yet larger than cfDNA, exosomes typically are about 30 to 200 nanometers in diameter (about 1/200th to 1/20th of a small red blood cell). Diagnostic technologies are exploiting the surface proteins exosomes bear, which act like return tracking numbers to the originating cell, to distinguish the origins of different mutations present in the exosome cargo. An added benefit for analytics is that these particles can be extracted from frozen biofluids.

microRNA
In blood, bits of extracellular RNA alone and unpackaged are degraded instantly but remain protected when within an exosome or bound up with other entities, such as the Ago2 protein or high-density lipoprotein. These protected microRNAs, about 22 nucleotides long, increase in quantity when cells become dysregulated, as when cancer progresses. MicroRNAs are highly stable and have been used to distinguish men with prostate cancer from healthy patients; they are emerging as targets to help inform tumor origin and status, early detection and prognostication because of new sequencing methods.

MicroRNAs might be considered the most distant to market, compared to CTCs, ctDNA and exosomes. In 2013 the Common Fund of National Institutes of Health (NIH) established the Extracellular RNA Communication program “to discover fundamental biological principles about the mechanisms of extracellular RNA (exRNA) generation, secretion, and transport; to identify and develop a catalogue of exRNA in normal human body fluids; and to investigate the potential for using exRNAs as therapeutic molecules or biomarkers of disease.”¹³Current challenges: Hurdles for liquid biopsies

Liquid biopsies are so new, regulatory agencies are drafting the criteria for market clearances and only one liquid biopsy has received FDA approval as of early 2017, as a companion diagnostic. While dozens of companies are establishing liquid biopsy footholds while gathering in vitro and clinical data for regulatory submissions, clinical trial sponsors must consider how best to employ investigational liquid biopsy technology to complement-or replace, in the post-approval future-the current gold standards of tissue biopsies and imaging.

Sponsors intending to use liquid biopsies know patient safety is paramount and protocols must generate evidentiary data to document certainty and reproducibility of the new technology performance and relationship to patient outcomes. In contrast to the regulatory environment for in vitro diagnostics which use standards of sensitivity and specificity relative to a gold standard, the FDA has set expectations of measuring utility of liquid biopsy by improvement in overall clinical outcomes of patients, such as improvements in overall survival. Currently, redundancies with using tissue biopsies, imaging technology and other diagnostics are necessary, which can impact trial timelines and resources. One strategy for ongoing long-term trial sponsors or those with near-term study launches might be to amend their approved protocols and patient consent to add experimental endpoints that use liquid biopsies, even if retroactively on stored samples. Several other aspects of planning future trials using liquid biopsies include the following:

Selecting targets
The capacity of liquid biopsy technology to enable real-time monitoring requires choosing targets most useful for tracking real-time tumor transformations and activities. Research suggests mutation targets might include those least likely to be responsive to the treatment. Though costly, some sponsors might want to consider employing multiple mutation panels that permit patient-population and patient-specific screening. The National Cancer Institute’s Cancer Genome Atlas and the International Cancer Genome Consortium have extensive mutation data, which can aid in protocol development.

Composing tumor boards
Clinical trial sponsors need to understand how incorporating liquid biopsies into their protocols might affect the use of traditional medical tumor boards and may need to consider creating molecular oncology boards for patient evaluations. The use of such boards in clinical care today-traditionally composed of medical, surgical and radiation oncologists, pathologists, and radiologists with extensive cancer expertise-is particularly helpful to guide treatment of patients whose cancers are rare, difficult or treatment-resistant. Such boards must also consider how to care for those patients for whom no treatment is yet genetically indicated or available.

With the potential uptake of liquid biopsies, boards will need to take on interpreting more diagnostic data, such as the relative effect of mutations, to guide treatment decisions. Such discussions may take more time, slowing the pace of patient reviews, and many oncologists may need more genetics training to engage fully in discussions. The same is true for the clinical trial setting, so sponsors employing liquid biopsies will have to ensure all clinical trial staff understand the design, use and significance of liquid biopsies and may consider supplementing their boards with experts in bioinformatics, bioethicists and geneticists. Moreover, any results that lead to recommending treatments must also be verified in a laboratory certified under the Clinical Laboratory Improvement Amendments (CLIA) regulations.

Counseling treatment decisions
Sponsors must state in their protocols, so that internal review boards, study investigators, site teams and trial participants understand, the role of results from the investigative liquid biopsy vs. other diagnostic methods. The protocol must include the procedures and processes for determining the significance and resolution of result discrepancies between biopsy technologies, such as when a tissue sample tests negative but a liquid biopsy is positive.

The future

The track record of success for liquid biopsies is growing beyond the ability to merely screen patients for the presence of mutations and to augment tissue biopsies for treatment decisions. In early July 2016, investigators successfully applied the technology to significantly determine the prognoses of patients with stage II colon cancer that had not metastasized, a form with a high post-surgery cure rate. By using liquid biopsies during the two-year study, investigators found that among patients with the target ctDNA after surgery, 79% relapsed and at a median of 27 months. In contrast, relapses occurred in only 9.8% of the patients without identifiable target cfDNA (p<0.001).¹⁴ The findings demonstrate how liquid biopsies might help clinicians prioritize patients in need of post-surgical treatment because of their increased risk of recurrence, while reassuring others of a very low likelihood of relapse. Such outcomes data are exactly what regulators and clinicians need to move forward confidently in adopting liquid biopsy technology.

To enhance the clinical use of ctDNA, a working group of the NIH Foundation’s Biomarkers Consortium Cancer Steering Committee is developing a project to establish qualitative and quantitative controls for performance metrics of technology, reagents and laboratory practices to assure the validity of liquid biopsy results for ctDNA tests. The project outcomes should be applicable, independent of technology, to basic and clinical research, clinical care and the needs of CLIA for quality assessment, the FDA for regulation, and payers for reimbursement decisions.

Looking distantly, liquid biopsies may be used to screen asymptomatic people, whether at-risk populations or the worried well. Such applications, experts report, have significant medical, regulatory, financial, and ethical hurdles. Identifying early pre-malignancies via liquid biopsies will need to reveal the origin certainly and define growth aggressiveness and necessity of treatment in terms of improved patient outcomes and acceptable pharmacoeconomics.

Joy Yucaitis is Senior Director, Oncology Strategy, Novella Clinical

References

1. Diaz, L.A. “Detection and Monitoring of Pre-Malignant Disease and Therapeutic Response. Liquid biopsies and the early diagnosis of cancer.” AACR Annual Meeting Plenary Session. April 20, 2016. Accessed at http://webcast.aacr.org/console/player/32242?mediaType=audio&. @ 1 min.
2. World Health Organization. Early detection of cancer. http://www.who.int/cancer/detection/en/
3. Rubin G, Vedsted P, Emery J. Improving cancer outcomes: better access to diagnostics in primary care could be critical. The British Journal of General Practice. 2011;61(586):317-318. doi:10.3399/bjgp11X572283.
4. National Cancer Institute. Cancer Classification. https://training.seer.cancer.gov/disease/categories/classification.html
5. American Joint Committee on Cancer (AJCC). What is Cancer Staging? 2017. https://cancerstaging.org/references-tools/Pages/What-is-Cancer-Staging.aspx
6. National Cancer Institute. About Cancer. Diagnosis and Staging. Prognosis. Tumor Grade. May 3, 2013. https://www.cancer.gov/about-cancer/diagnosis-staging/prognosis/tumor-grade-fact-sheet
7. Whole paragraph adapted from National Cancer Institute. About Cancer. Diagnosis and Staging. Diagnosis. Tumor Markers. https://www.cancer.gov/about-cancer/diagnosis-staging/diagnosis/tumor-markers-fact-sheet
8. Gerlinger, M. “Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.” N Engl J Med. 2012 Mar 8;366(10):883-92. doi: 10.1056/NEJMoa1113205. http://www.ncbi.nlm.nih.gov/pubmed/22397650
9. Gerlinger, M. “Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.” N Engl J Med. 2012 Mar 8;366(10):883-92. doi: 10.1056/NEJMoa1113205. http://www.ncbi.nlm.nih.gov/pubmed/22397650
10. Sacher, A., et al. “Prospective Validation of Rapid Plasma Genotyping for the Detection of EGFR and KRAS Mutations in Advanced Lung Cancer.” JAMA Oncol. Published online April 07, 2016. doi:10.1001/jamaoncol.2016.0173
11. Leon SA, et al. “Free DNA in the serum of cancer patients and the effect of therapy.” Cancer Res. 1977 Mar;37(3):646-50. http://www.ncbi.nlm.nih.gov/pubmed/837366. Identified in Yong as reference 3.
12. National Institute of Health. Extracellular RNA Communication. May 19, 2016. http://commonfund.nih.gov/Exrna/index
13. Tie, J., et al. “Circulating tumor DNA analysis detects minimal residual disease and predicts recurrence in patients with stage II colon cancer.” Science Translational Medicine. Jul 2016: Vol. 8, Issue 346, pp. 346ra92. DOI: 10.1126/scitranslmed.aaf6219.
14. Janssen’s CellSearch System Brochure.