Multicancer Early Detection Test Fails to Reduce Cancers Detected at Late Stages

CHICAGO — Adding a multicancer early detection (MCED) blood test to standard of care did not reduce the number of cancers detected at stage III or IV across a dozen tumor types for which no screening tests are currently available, the randomized NHS-Galleri trial showed.

After three screening rounds, a nonsignificant 3% more stage III/IV cancers were detected with the addition of MCED (incidence rate ratio [IRR] 1.03, 95% CI 0.92-1.14, P=0.06), reported Charles Swanton, MD, PhD, of the Francis Crick Institute in London, during a press conference at the American Society of Clinical Oncology (ASCO) annual meeting.

However, looking at stage IV cancers only, the investigators found a reduction of 14% after three screening rounds (IRR 0.86, 95% CI 0.744-0.998), which included a reduction of 22% after the second screening round and a reduction of 26% after the third screening round. The larger reductions in stage IV cancers in incident screening rounds suggest that “with further follow-up, a benefit in terms of reduced stage IV cancers may increase,” Swanton said.

In addition, there was a 16% increase in the detection of stage I and II cancers after three screening rounds (relative risk 1.16, 95% CI 1.03-1.30).

Despite missing the primary endpoint, “the results are encouraging, nevertheless,” Swanton noted. “We observed a numerical reduction in the secondary outcome of stage IV cancers at the 12 sites, and a numerical increase in stage I and II cancers. With further follow-up, we will see whether these trends continue and how they will impact cancer mortality.”

At this time, “NHS-Galleri should not be included in cancer screening guidelines,” said moderator Julie Gralow, MD, ASCO’s chief medical officer. “ASCO’s guideline on liquid biopsies in cancer has been recently updated and will be published soon. It will continue to support the use of liquid biopsies in the monitoring and the defining of cancers that have already been diagnosed, but it will not yet recommend them for early cancer detection.”

The trial evaluated whether adding the Galleri test to standard-of-care screening in England would reduce the incidence of stage III/IV cancers responsible for about two-thirds of cancer deaths in the U.S. and U.K., for which early diagnosis is critical.

About 70% of cancer deaths in the U.S. are attributed to cancer types not covered by current screening methods. Stage IV cancers represent 18% of all estimated diagnoses but contribute to 48% of all estimated cancer-related deaths within 5 years.

“This highlights a major unmet need,” said Swanton. “We need to find cancers earlier before metastatic spread.”

“Integrating MCED testing on a population scale may in the future help reduce late-stage cancer burden and facilitate early detection of cancer at stages when treatment with curative intent may be possible,” he added.

For this study, 142,924 asymptomatic participants ages 50 to 77 were included and provided blood samples at up to three annual visits. They were then randomized 1:1 to the intervention arm (MCED test plus usual care) or control arm (usual care). The 12 prespecified cancer types were lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, esophagus, anus, lymphoma, ovary, and bladder.

Those with a positive MCED test result were unblinded and referred for NHS urgent referral pathways for diagnostic testing, outside of the trial.

Across all cancer types, 3,637 cancers were diagnosed in the MCED group, and 3,400 were diagnosed in the control group. In the MCED group, four times more cancers were found by screening than in the group that received only the recommended screening (1,173 cases vs 290 cases).

The specificity of the test was 99.55%, and the positive predictive value was 52%. The false-positive rate was fewer than 1 in 200. Fewer than 1 in 1,000 screened individuals experienced a nonphlebotomy-related adverse event over three rounds of screening.

“What we weren’t expecting when we modeled the statistical design of this study was the latent background stage III/IV cancers that are prevalent in an unscreened population,” Swanton said in offering a possible explanation for missing the primary endpoint. “There was a large burden of stage III/IV cancers that are asymptomatic and undiagnosed in this population of 50- to 77-years-olds, which was unexpected. As a result, we were on the back foot from the get-go.”

“This is why we are planning to do an extra year of follow-up for cancer incidence in the control arm,” he added. “We’ve got 237 extra cancers diagnosed in the screening arm that we expect to be diagnosed in the next year symptomatically, through clinical presentation or screening.”

Gralow noted that the calculus for incorporating an early cancer screening blood test into clinical practice might change with longer follow-up of participants in trials such as NHS-Galleri, if a significant reduction in late-stage cancers is found. She also highlighted the U.S. counterpart study — REACH/Galleri-Medicare study — that will add to the evidence base on the risks and benefits of MCED testing.