Will liquid biopsies be used routinely to screen for cancer?

Click here to read the Cover Story, “Oncologists must be ‘very cautious’ as liquid biopsy use expands.”

Yes.

Genomic analysis of plasma cfDNA has been adopted widely in the care of patients with cancer. Indeed, the rapid and widespread adoption is striking and a testament to how convenient and informative these blood tests can be.

What is next for this technology? Investigators are intuitively asking whether such a test could be used to screen for occult cancer. Remember, clinicians already use plasma DNA sequencing routinely to screen for fetal aneuploidy among pregnant women, so we know the technology is robust and scalable. Cancer is a disease of the genome, so genomic analysis should similarly be able to detect the presence of cancer.

The available clinical data suggest that blood tests have some features that could make them quite ideal for cancer screening. First, there is potential to detect a wide range of cancers — over 20 cancer types have been detectable with some tests. Second, in addition to detecting the cancer signal, these blood tests can predict what kind of cancer is present — some of the most compelling data have used methylation signatures to predict the tissue of origin. Third, the false-positive rate is quite low compared with imaging tests, with recent data reporting a false-positive rate below 1%, meaning an abnormal result could be more likely to represent cancer as opposed to a false positive. Finally, these tests seem to preferentially detect dangerous cancers that need treatment and might be less likely to detect “overdiagnosis” cancers that are slow growing. In an era when treatments are improving, liquid biopsy approaches could be ideal for finding deadly cancers early and getting them treated quickly.

It is important to acknowledge that not all cancers shed DNA into the blood, so liquid biopsy approaches will not have perfect sensitivity. I suspect a blood test is unlikely to be as sensitive as imaging for finding small tumors, and I do not think a blood test will replace established screening approaches. But a blood test is certainly logistically easier than imaging and could permit testing for cancers that are hard to detect with imaging, such as pancreatic and ovarian cancers.

In short, I do see a day in the future when liquid biopsy approaches are used to screen for cancer, but important questions remain. What populations would be screened with such a test, and how frequently would the testing be done? How do you manage a positive test — what is the diagnostic odyssey from positive blood test to cancer diagnosis and treatment? We need to better quantify the risks and benefits of this approach before blood-based cancer screening will be routine, but the good news is this research is well underway and these tests are rapidly becoming a reality.

References:

Oxnard GR, et al. Abstract LBA77. Presented at: European Society for Medical Oncology Congress; Sept. 27-Oct. 1, 2019; Barcelona, Spain:

Oxnard GR, et al. Abstract 1545. Presented at: ASCO Annual Meeting; May 31-June 4, 2019; Chicago.

Geoffrey Oxnard, MD, is associate professor of medicine at Harvard Medical School. He can be reached at geoffrey_oxnard@dfci.harvard.edu. Disclosure: Oxnard reports consultant/advisory roles with and honoraria from Grail, Guardant Health and Illumina.

No.

Although liquid biopsies promise the ability to detect minute amounts of tumor cells and tumor DNA in a noninvasive and potentially low-cost way, using liquid biopsies to screen the general population for cancer presents a number of problems that must be carefully thought through.

Liquid biopsies have sensitivity and specificity issues, resulting in both false-positive and -negative results. Therefore, their routine use to screen for cancer could lead to an unnecessary amount of procedures and patient anxiety and angst. Along with this is the cost, not only of performing complex genetic and laboratory testing on millions of patients, but of the subsequent physician visits, diagnostic biopsy procedures, complications resulting from these biopsies, time needed off from work and other less tangible costs of performing routine screening.

Another issue for routine screening is technical. Not all cancers share the same genetic mutations or appear morphologically similar under a microscope. To be an effective screening tool, the test must be easily reproducible and apply to a broad segment of the population. Also, there must be a clinically relevant action that can be taken, such as a lumpectomy for a breast mass detected on mammogram. Although a blood test may detect cancer cells or cancer DNA, it remains unclear exactly how this information would be acted upon, especially as these genetic tests do not necessarily inform us about where the cancer is (localized to one organ or metastatic), and because there are many genetic changes that appear across multiple tumor types. If a screening CT scan or ultrasound fails to detect the cancer present on a liquid biopsy, it is not clear what potentially morbid intervention would be performed. Enhanced screening with more frequent visits and screening tests could be pursued, but it nonetheless remains unclear if this will indeed save lives.

Despite these drawbacks as a screening test, there are very promising uses for liquid biopsies. Currently, liquid biopsies are being used as a tool for discovering new genetic targets in patients with known cancer, to monitor for the development of drug resistance and to identify new therapies with a high chance of working. In lung cancer, for example, a cfDNA test can monitor for the development of resistance to certain oral EGFR inhibitors, and in prostate cancer a circulating tumor cell test can predict resistance to next-generation hormonal therapies. Indeed, these more “targeted” uses of liquid biopsies are more compelling and likely to yield more clinically actionable information.

Additionally, a recent study showed promise for personalizing liquid biopsies to detect recurrence of a known cancer. Christensen and colleagues evaluated patients with localized bladder cancer who had their bladder removed and needed to be surveyed for recurrence. Typically, that is done with repeated CT scans or urine tests, but those do not detect very small amounts of residual cancer. In this study, the researchers sequenced the bladder cancer tissue of patients to find the unique mutations present in each patient’s tumor and created a profile. They then obtained serial blood samples over time after surgery and found that mutations were detectable only among the patients who recurred, meaning that all the patients with detectable mutations after surgery relapsed, whereas those who remained mutation-free also remained cancer-free. This personalized surveillance strategy to detect recurrences can potentially lead to better clinical outcomes, for example by allowing providers to starting chemotherapy, radiation or immunotherapy earlier, and this will need to be evaluated in studies going forward. Additionally, this use of liquid biopsy to predict recurrence is not limited in principle to bladder cancer but could be applied across most of oncology.

Putting this together, the technology that allows us to find very small amounts of cancer via a liquid biopsy has significant potential. These areas of potential deserve more focus and more trials and, if successful, will improve outcomes for patients. Although detecting cancers earlier in asymptomatic populations is a noble goal, there are a number of major challenges in implementing this strategy. This technology is not yet ready for it, and efforts should concentrate on focused uses with tangible and relevant clinical outcomes.

Reference:

Christensen E, et al. J Clin Oncol. 2019;doi:10.1200/JCO.18.02052.

Terence Friedlander, MD, is chief of the division of hematology/oncology at Zuckerberg San Francisco General Hospital, associate director of Helen Diller Family Comprehensive Cancer Center and associate clinical professor in the division of hematology/oncology at University of California, San Francisco. He can be reached at terence.friedlander@ucsf.edu. Disclosure: Friedlander reports a consultant/advisory role with Foundation Medicine and research funding from Epic Biosciences.