Liquid Biopsies to Diagnose Cancer

Liquid biopsies use blood—not tumor tissue—to diagnose cancer

A blood sample being held with a row of human samples for analytical testing including blood, urine, chemistry, proteins, anticoagulants and HIV in lab
Andrew Brookes / Getty Images

Typically, tumors are examined using tissue biopsies. A small sample is taken from the tumor and genotyped, or analyzed for genetic makeup. The problem with this approach is that biopsying tumors can be challenging. Furthermore, a tumor biopsy provides only a snapshot of the tumor.

Writing in Discovery Medicine in 2015, Labgaa and co-authors state the following about conventional tumor biopsy:

For obvious reasons, it is difficult to monitor tumor evolution by sequential biopsies. Also, biopsy only mirrors a single spot of the tumor and is therefore unlikely to represent the whole spectrum of somatic mutations in large tumors. An alternative would be to obtain multiple biopsies for the same tumor, but this option seems neither realistic nor accurate.

Liquid biopsy involves the measurement of circulating DNA (ctDNA) and other tumor byproducts in blood samples obtained from patients with cancer. This emerging diagnostic approach promises to be rapid, noninvasive, and cost-effective.

History of Liquid Biopsy

In 1948, Mandel and Métais, a pair of French researchers first identified ctDNA in the blood of healthy people. This discovery was ahead of its time, and it wasn’t until decades later that ctDNA was further explored.

In 1977, Leon and colleagues first identified increased amounts of ctDNA in the blood of cancer patients. By 1989, Stroun and colleagues identified neoplastic (i.e., cancer) characteristics in the blood. After these discoveries, several other groups identified specific mutations in tumor suppressors and oncogenes, microsatellite instability, and DNA methylation, which proved that ctDNA is released into the circulation by tumors.

Although we know that ctDNA derived from tumor cells circulates in the blood, the origin, rate of release, and mechanism of release of this DNA are unclear, with research yielding conflicting results. Some research suggests that more malignant tumors contain more dead cancer cells and release more ctDNA.

However, some research suggests that all cells release ctDNA. Nevertheless, it seems likely that cancerous tumors release increased levels of ctDNA into the blood, making ctDNA a good biomarker of cancer.

Because of heavy fragmentation and low concentrations in the blood, ctDNA is difficult to isolate and analyze. There is a discrepancy of ctDNA concentrations between serum and plasma samples. It seems that blood serum rather than blood plasma is a better source of ctDNA. In a study by Umetani and colleagues, ctDNA concentrations were found to be consistently low in the plasma compared to the serum due to possible loss of circulating DNA during purification, as coagulation and other proteins are being eliminated during specimen preparation.

According to Heitzer and colleagues, here are some specific issues that need to be resolved to harness the diagnostic potential of ctDNA:

First, preanalytical procedures need to be standardized …. Selection of an isolation method that ensures extraction of a sufficient amount of high-quality DNA is critical and it has been shown that preanalytical factors of blood sampling and processing can strongly affect DNA yield …. Second, one of the most important issues is the lack of harmonization of quantification methods. Different quantification methods, ... produce different results because these measurements target either total or only amplifiable DNA …. Third, less is known about the origin and the detailed mechanism of ctDNA release, and in most studies confounding events that might also contribute to the release of ctDNA.

Targeted vs. Untargeted Approaches

Currently, there are two main approaches taken when analyzing blood plasma (or serum) for ctDNA. The first approach is targeted and looks for specific genetic changes indicative of tumors. The second approach is untargeted and involves a genome-wide analysis looking for ctDNA reflective of cancer. Alternatively, exome sequencing has been used as a more cost-effective, untargeted approach. Exomes are the portions of DNA that are transcribed to make protein.

With targeted approaches, serum is analyzed for known genetic mutations in a small set of driver mutations.

Driver mutations refer to mutations in the genome that promote, or “drive,” the growth of cancer cells. These mutations include KRAS or EGFR.

Because of technological advances in recent years, targeted approaches to the analysis of the genome for small amounts of ctDNA have become feasible. These technologies include ARMS (amplification refractory mutation system); digital PCR (dPCR); beads, emulsions, amplification, and magnetics (BEAMing); and deep sequencing (CAPP-Seq).

Even though there have been advances in technology that make the targeted approach possible, the targeted approach only targets a few positions of mutations (hotspots) and misses a lot of driver mutations such as tumor suppressor genes.

The main benefit of untargeted approaches to liquid biopsy is that they can be used in all patients due to the fact that the test doesn’t rely on recurrent genetic changes. Recurrent genetic changes don’t cover all cancers and aren’t specific cancer signatures. Nevertheless, this approach lacks analytic sensitivity and comprehensive analysis of tumor genomes is not yet possible.

Of note, the price of sequencing an entire genome has substantially dropped. In 2006, the price of sequencing the whole genome was approximately $300,000 (USD). By 2017, the cost had dropped to approximately $1,000 (USD) per genome, including reagents and the amortization of sequencing machines.

Clinical Utility of Liquid Biopsy

Initial efforts to use ctDNA were diagnostic and compared levels in healthy patients with those of cancer patients or those with benign disease. Results of these efforts were mixed, with only some studies showing significant differences indicating cancer, disease-free status, or relapse.

The reason why ctDNA can be used only some of the time to diagnose cancer is because variable amounts of ctDNA are derived from tumors. Not all tumors “shed” DNA in the same amount. In general, more advanced, widespread tumors shed more DNA into the circulation than do early, localized, tumors. Additionally, different tumor types shed different amounts of DNA into the circulation. The fraction of circulating DNA that is derived from a tumor is widely variable across studies and cancer types, ranging from 0.01% to 93%. It is important to note that, in general, only a minority of ctDNA is derived from the tumor, the rest of it coming from normal tissues.

Circulating DNA could be used as a prognostic marker of disease. Circulating DNA could be used to monitor changes in cancer over time. For example, one study showed that the two-year survival rate in patients with colorectal cancer (i.e., the number of patients still alive at least two years after diagnosis with colorectal cancer) and the KRAS hotspot mutations was 100 percent in those without evidence of corresponding circulating DNA. Moreover, it’s possible that in the near future, circulating DNA can be used to monitor precancerous lesions.

Circulating DNA could also be used to monitor response to therapy. Because circulating DNA proffers a better overall picture of the genetic makeup of tumors, this DNA likely contains diagnostic DNA, which can be used instead of diagnostic DNA attained from tumors themselves.

Now, let’s take a look at some specific examples of liquid biopsy.


Guardant Health developed a test that uses next-generation sequencing to profile circulating DNA for mutations and chromosomal rearrangements for 73 cancer-related genes. Guardant Health published a study reporting the utility of liquid biopsy in oncology. The study used blood samples from 15,000 patients with a combined 50 tumor types.

For the most part, the results from the liquid biopsy test aligned with gene alterations observed in tumor biopsies.

According to the NIH:

Guardant360 identified the same critical mutations in important cancer-related genes like EGFR, BRAF, KRAS, and PIK3CA at frequencies very similar to what had previously been identified in tumor biopsy samples, statistically correlating to 94% to 99%.

Furthermore, according to the NIH the researchers reported the following:

In a second component of the study, the researchers evaluated nearly 400 patients—most of whom had lung or colorectal cancer—who had both blood ctDNA and tumor tissue DNA results available and compared the patterns of genomic changes. The overall accuracy of the liquid biopsy in comparison with results from the tumor biopsy analyses was 87%. The accuracy increased to 98% when the blood and tumor samples were collected within 6 months of each other.

Guardant360 was accurate even though the levels of circulating DNA in the blood were low. Oftentimes, circulating tumor DNA only made up 0.4 percent of the DNA in the blood.

Overall, using liquid biopsy, the Guardant researchers were able to identify tumor markers that could direct treatment by physicians in 67 percent of patients. These patients were eligible for FDA-approved treatments as well as investigational therapies.

ctDNA and Lung Cancer

In 2016, the FDA approved the cobas EGFR Mutation Test to be used for the detection of EGFR mutations in the circulating DNA of patients with lung cancer. This test was the first FDA-approved liquid biopsy and identified patients who may be candidates for treatment with targeted therapies using erlotinib (Tarceva), afatinib (Gilotrif), and gefitinib (Iressa) as first-line treatment, and osimeritinib (Tagrisso) as second-line treatment. These targeted therapies attack cancer cells with specific EGFR mutations.

Importantly, because of the high number of false-negative results, the FDA recommends that a tissue biopsy sample also be taken from a patient who has a negative liquid biopsy.

ctDNA and Liver Cancer

The number of people dying of liver cancer has increased during the past 20 years. Currently, liver cancer is the second leading cause of cancer death in the world. There are no good biomarkers available to detect and analyze liver, or hepatocellular (HCC), cancer. Circulating DNA could be a good biomarker for liver cancer.

Consider the following quotation from Lagbaa and co-authors about the potential of using circulating DNA to diagnose liver cancer:

Hypermethylation of RASSF1A, p15, and p16 have been suggested as early diagnostic tools in a retrospective study including 50 HCC patients. A signature of four aberrantly methylated genes (APC, GSTP1, RASSF1A, and SFRP1) was also tested for diagnostic accuracy, while methylation of RASSF1A was reported as a prognostic biomarker. Subsequent studies analyzed ctDNA in HCC patients using deep sequencing technologies .... Strikingly, aberrant DNA copy numbers were detected in two HBV carriers without previous history of HCC at the time of blood collection, but who developed HCC during follow-up. This finding opened the door to evaluate copy number variation in ctDNA as a screening tool for early HCC detection.

A Word From Verywell

Liquid biopsies are an exciting new approach to genomic diagnosis. Currently, certain liquid biopsies, which offer comprehensive molecular profiling, are available to physicians to complement genetic information gained from tissue biopsy. There are also certain liquid biopsies that can be used in lieu of tissue biopsy—when tissue biopsies are unavailable.

It’s important to keep in mind that many liquid biopsy trials are currently ongoing and more research needs to be done to flesh out the therapeutic utility of this intervention.


Blood Test for Genetic Changes in Tumors Shows Promise as Alternative to Tumor Biopsy. NIH.

Heitzer E, Ulz P, Geigl JB. Circulating Tumor DNA as a Liquid Biopsy for Cancer. Clinical Chemistry. 2015; 61:112–123. doi: 10.1373/clinchem.2014.222679

Lagbaa J, Villanueva A. Liquid biopsy in liver cancer. Discovery Medicine. 2015;19(105):263-73.

Liquid Biopsy: Using DNA in Blood to Detect, Track, and Treat Cancer. NIH.

Umetani N, et al. Higher amount of free circulating DNA in serum than in plasma is not mainly caused by contaminated extraneous DNA during separation. Ann N Y Acad Sci. 2006;1075: 299-307.

Wellstein A. General Principles in the Pharmacotherapy of Cancer. In: Brunton LL, Hilal-Dandan R, Knollmann BC. eds. Goodman & Gilman's: The Pharmacological Basis of Therapeutics, 13e New York, NY: McGraw-Hill.