Suzette Delaloge, MD, MSc, oncologist, breast cancer specialist, and director of the individualized cancer prevention program (Interception) at the Gustave Roussy Institute in Villejuif, France, looks into these “liquid biopsies” and shares her reservations about their potential marketing, especially to the organized care plans.
Question: What are the general principles underpinning these MCED tests?
Suzette Delaloge, MD, MSc: Despite their specificities, the general idea is to detect certain cancer markers in various body fluids (blood, urine, saliva, etc.), for example, molecules released by cancer cells (cytokines, inflammatory proteins, leptin, etc.) or distinctive features of the DNA in tumor cells. In blood, these molecules can be found in plasma or in serum. In urine, it’s more about detecting kidney, bladder, and urinary tract cancers.
Q: What sort of time frame are we looking at for these MCED tests to be used in routine practice?
Dr. Delaloge: They first appeared around 10 years ago. Development of these tests has intensified in recent years. There are numerous research laboratories, both public and private, that are developing different early-detection tests for cancer.
Some of these development processes are about to come to an end and are expected to be in regular, concrete use within 5-10 years. For the most advanced developments, the main biologic material researched and analyzed is DNA from cancer cells. We all have fragments of DNA from dead cells in our plasma (apoptosis), but cancer cells release more of these than others, and most importantly, their DNA has distinctive characteristics. The idea is to develop tests capable of detecting these characteristics.
Liquid biopsies based on genomic biomarkers could make MCED a reality, especially for cancers for which there is no standard screening process. But at this stage of the research, there are limitations, including low sensitivity for detecting stage I cancers in validation studies and an increased risk for overdiagnosis.
Q: What specific set of characteristics are the most advanced approaches based on?
Dr. Delaloge: They’re based on the analysis of DNA methylation, a biological process by which CH3 methyl groups are added to the DNA molecule and that determines gene expression. This phenomenon differs depending on whether the cell is cancerous. Among the tests currently under development making use of this specific characteristic is the Galleri test, which is the most advanced of them all.
A previous British National Health Service study, SYMPLIFY, which was published in 2023 by researchers at the University of Oxford, was conducted in symptomatic patients attending a health center. It offers promising results in a diagnostic situation. It has nothing at all to do with screening here. A large, randomized English study, NHS-Galleri, is underway, this time involving the general population, with the aim of assessing the potential benefit of the same test as screening in 140,000 people between ages 50 and 77 years.
In the SYMPLIFY study, which was carried out in symptomatic patients attending a health center, the Galleri MCED test had a positive predictive value of 75.5%, a negative predictive value of 97.6%, a sensitivity of 66.3%, and a specificity of 98.4%. Sensitivity increased with age and cancer stage from 24.2% at stage I to 95.3% at stage IV. For cases for which a cancer signal was detected in patients with cancer, the prediction of the original site of the cancer by the MCED test was accurate in 85.2% of cases. This large-scale prospective evaluation of an MCED diagnostic test confirms its feasibility in a symptomatic population but is not yet sufficiently accurate to “confirm or rule out the presence of cancer.” According to the authors, “in cases in which the MCED test detects a cancer signal in this context, the probability of a diagnosis of cancer being made is considerably higher and may identify cancers at sites other than those suspected during the initial referral phase, thus reducing delays in diagnosis.” A negative test means a lower likelihood of cancer but not so low that proper investigation can be ruled out. Further tests will be needed to optimize use of a negative predictive value.