MCED Testing Enters New Diagnostic Realms

Evidence is mounting that a single blood test can effectively screen for dozens of tumor types simultaneously, even in individuals with suspicious symptoms or in “real-world” cases involving asymptomatic populations.

The findings highlight the potential of multicancer early detection (MCED) assays to greatly expand the reach of cancer screening and detection across a wide range of patient care situations. “This is meaningful,” says Jeffrey Venstrom, MD, chief medical officer of GRAIL, maker of Galleri, the first and only commercially available MCED test in the United States.

However, some researchers worry about the widespread applicability of MCED tests—particularly in high-risk populations that are not representative of clinical cohorts.

Galleri works by analyzing methylation patterns in cell-free DNA and then using machine-learning algorithms to detect more than 50 cancer types. Sold since 2021 with a prescription, the screening tool is available as a laboratory-developed test. It is not yet cleared or approved by the FDA, so few insurance companies cover its $949 sticker price. Still, that hasn't stopped tens of thousands of people from ordering it—and data reported at the 2023 American Society of Clinical Oncology (ASCO) Annual Meeting in June indicate that Galleri is largely working as intended in real-world cases (J Clin Oncol 41, 2023 [suppl 16; abstr 10519]).

Consider the percentage of positive results. Among the first 53,744 customers, the test flagged 510 potential cancer cases—a signal detection rate of around 1%, on par with what researchers had documented previously in large clinical trials.

It's too early to say how many of the 510 positive test results truly correspond to cancer. But at the ASCO meeting, Candace Westgate, DO, of Adventist Health in Saint Helena, CA, and her colleagues presented data on the first 152 evaluable individuals: 72 had their diagnosis confirmed—a positive predictive value (PPV) similar to the 43% rate observed in a large prospective trial. Among those 72 individuals, the blood test correctly identified the tumor's tissue of origin in about 91% of them, consistent with earlier studies.

The real-world performance “speaks to the validity of the test,” Westgate says. “It is functioning in the same way” that it did in controlled trials, “which is really, really exciting.”

However, some clinicians suggest caution may be warranted, especially for individuals in high-risk professions.

Take firefighters, for example, a group regularly exposed to smoke, soot, and other carcinogens as part of their jobs. Vershalee Shukla, MD, from the Vincere Cancer Center in Scottsdale, AZ, has given the blood test to thousands of first responders. Although Galleri detected some cancers, according to Shukla, a co-author of the ASCO report, it failed to identify many others evident in whole-body scans. “It's not ready for prime time yet,” she says, citing the need for additional data from younger people facing occupational hazards.

More data are also needed on people with symptoms that might indicate cancer, such as unexplained weight loss, abdominal pain, or changes in bowel habits. Initial insights into Galleri's performance for such individuals come from the SYMPLIFY trial in which people seeking medical care in the UK for nonspecific symptoms received both standard diagnostic assessments and the Galleri test.

According to data presented at the ASCO meeting and subsequently published, 323 of 5,461 participants (about 6%) tested positive on the Galleri test. Of these, 244 were confirmed as true positives (Lancet Oncol 2023;24:733–43). That translates to a PPV of approximately 75%, a degree of accuracy that could prompt physicians to further evaluate patients with vague symptoms that may not initially warrant extensive diagnostic workups.

“If you had a point-of-care test that you could run along with basic labs and other initial lightweight workup for vague symptoms, then certainly you would reduce diagnostic delays in many, many patients that actually do have cancer,” says Sana Raoof, MD, PhD, of Memorial Sloan Kettering Cancer Center in New York, NY, who consults for GRAIL and other MCED developers.

Moreover, tests such as GRAIL's could “speed up investigations” by helping to predict the likely site of a cancer for clinicians to confirm via other diagnostic procedures, notes Mark Middleton, MD, PhD, of the University of Oxford, UK, who led the SYMPLIFY trial.

But as the trial data show, Galleri frequently missed some cancers detectable through imaging, endoscopy, or other diagnostic tests. It performed well when it came to identifying esophageal and gastric cancers, but overall, it failed to catch 124 of the 368 cancers found by other methods.

A negative test result from GRAIL's assay alone would therefore be insufficient to rule out the need for further investigation, says Jeanne Tie, MD, of the Peter MacCallum Cancer Centre in Melbourne, Australia, who penned a commentary article (Lancet Oncol 2023;24:710–11).

The assay's moderate sensitivity “may be sufficient in the screening setting,” she says. “However, for MCED testing to be useful in patients who present with symptoms suspicious of cancer, a high test sensitivity is critical to avoid missing a cancer diagnosis that would otherwise be picked up by the standard diagnostic pathway.”

The algorithms that power the test could be tweaked to boost sensitivity and maximize the probability of true negative results for symptomatic patients. Such a recalibration, says Middleton, “might be good enough to rule out the need for invasive tests in some settings”—“provided,” he adds, “that we can turn the test around fast enough to make clinical decisions within a week or two.” –Elie Dolgin

For more news on cancer research, visit Cancer Discovery online at https://aacrjournals.org/CDNews.