MCED blood test boosts cancer detection in symptomatic patients

Abstract

A large observational study of patients in England and Wales evaluated the effectiveness of the methylation-based multi-cancer early detection (MCED) blood test Galleri to identify over 50 types of cancer in symptomatic subjects. According to the University of Oxford researchers, who led the SYMPLIFY study, never before has an MCED test’s performance been evaluated in patients suspected of having some form of cancer.

“This is the first study to examine the performance of this blood test in a symptomatic population,” says study co-author Mark R. Middleton, PhD, professor of experimental cancer medicine, a consultant medical oncologist, and head of the Department of Oncology at the University of Oxford in the United Kingdom. “The test has been developed for screening asymptomatic people in the general population. Looking at people with symptoms is important because most people with cancer have symptoms before they present.”

Dr Middleton, who is also director of the cancer research UKOxford Centre, adds, “Often, the symptoms are vague and family physicians have to work out whom to refer for invasive tests. We were interested in seeing how [this blood test] might perform here, to see if it could help sort out who needs more tests for what symptoms.” The study appears in Lancet Oncology (doi:10.1016/S1470-2045(23)00277-2).

Each subject was referred for imaging, endoscopy, or other diagnostic tests as follow-ups for suspected gynecological, lung, lower gastrointestinal, or upper gastrointestinal cancer and other non-specific but suspicious symptoms. Each subject also provided a blood sample for DNA testing.

According to the researchers, the most frequently reported symptoms were unexplained weight loss (1318, 24.1%), erratic bowel habits (1199, 22.0%), postmenopausal bleeding (875, 16.0%), and rectal bleeding (858, 15.7%). Other symptoms included abdominal pain (794, 14.5%), other pains (580, 10.6%), dysphagia (482, 8.8%) and anemia (390, 7.1%)

The researchers found that the MCED test detected cancer signals in 323 subjects. Of these, 244 were diagnosed with cancer, a positive predictive value of 75.5% and a negative predictive value of 97.6%. The investigators found that the test’s sensitivity for cancer detection increased with more advanced cancer stages, and that there was some variation in accuracy depending on referral pathway, symptom cluster, and symptoms. In essence, they wrote that the results illustrated the importance of clinical context in interpreting the results and urged careful evaluation of each individual case.

With these caveats, researchers found that the most common cancer diagnoses of the 368 cancers detected were colorectal cancer (137, 37.2%), lung cancer (81, 22.0%), uterine cancer (30, 8.2%), oesophago-gastric cancer (22, 6.0%), and ovarian cancer (14, 3.8%).

The overall sensitivity and specificity of the MCED test for uncovering the 368 cancers was 66.3% and 98.4%, respectively. Sensitivity increased with patient age, and dramatically by cancer stage, from 24.2% for stage I to 95.3% for stage IV.

The researchers noted that there are no screening programs for most cancers; instead, most patients diagnosed with cancer first approach their primary care physician with vague symptoms that can take time to accurately diagnose. The hope is that a broad-based blood test like this one can expedite symptomatic cancer diagnosis.

“I think it is an important study,” says Wafik El-Deiry MD, PhD, associate dean for oncologic sciences at the Warren Alpert Medical School and director of the Legoretta Cancer Center at Brown University in Providence, Rhode Island. “Even though it is observational and not a prospective randomized study, I think it is an advance in the field because cancer diagnosis can take quite some time when you start with non-specific symptoms or use the standard approach for all the different tumor types. These types of blood tests can produce results fairly quickly with impressive sensitivity and with epigenetics alterations. This makes them cost-effective and potentially very useful clinically.”

Dr El-Deiry adds, “In terms of accuracy—with a positive predictive value of 75.5% and a negative predictive value of 97.6%—the numbers aren’t bad. But, like many other tests in medicine, obviously, this isn’t perfect.”

Although MCED tests are used to detect 50 different types of cancer, Dr El-Deiry notes that, according to the study, they are not as effective for detecting certain types of cancers, including prostate cancer. “I think it has to be taken with the caveat that there’s certainly some variability across tumor types,” he says. “However, if the cost of the test is less of an issue, I think it could be well worth it.”

Dr Middleton believes there are important takeaways from the study. “[This specific blood test] performed as expected,” he says, “and at about the same [level of] diagnostic accuracy in the symptomatic population as [in] the asymptomatic [population].” He says that its current algorithm is not accurate enough to say with confidence that people whose family doctor refers them for tests to rule out cancer do not need to do so if the test is negative. “If we retune the algorithm this might be possible, but [it] would need a new trial to confirm.”

Dr Middleton also says that physicians could use the blood test to identify patients with vague symptoms (weight loss, for example) who might need scans or scopes but whose family doctors do not currently send them for these tests. “We do need a follow-up study to prove this, however,” he says. He also believes that because the blood test was accurate approximately 85% of the time in identifying where the cancer was in the body, it could be used by primary care doctors before pursuing other diagnoses.

Jeanne Tie, MD, MBChB, an associate professor in the Department of Medical Oncology at the Peter MacCallum Cancer Centre in Victoria, Australia, and the Walter and Eliza Hall Institute in Melbourne, Australia, by and large applauds the study in an accompanying editorial (doi:10.1016/S1470-2045(23)00288-7). She notes that there are still many issues that need to be resolved before there can be broad use of MCED testing in clinical settings.

“MCED tests need to be able to detect a low level of ctDNA [circulating tumor DNA] with high confidence and filter out single nucleotide polymorphisms and clonal haematopoiesis of indeterminate potential variants without previous knowledge of cancer-related genomic changes,” she says. She also notes that because of the varying cancer incidence and referral pathway in different countries, “the value of the multi-cancer early detection test will need to be assessed within the specific context in which it is being implemented.”