Developers of an investigational multi-target screening test for colorectal cancer and its precursors are aiming to replace the widely used conventional fecal immunochemical test.
The new investigational assay, multi-target fecal immunochemical test, or mtFIT, appears to outperform FIT on endpoints that measure both positivity and detection rates for colorectal cancer and precancerous growth, according to results from a clinical trial that enrolled more than 13,000 participants in the Netherlands in 2022.
The findings were published Feb. 9 in The Lancet Oncology.
“The present clinical trial was the first prospective intervention study in the real-world setting of a well-organized, population-based colorectal cancer screening program,” said Raymond DuBois, director of the Medical University of South Carolina Hollings Cancer Center and executive chairman of the board of directors at the Mark Foundation for Cancer Research.
DuBois, who is not directly involved in the study, is an expert in colorectal cancer and the role of inflammatory mediators in the progression of colon cancer.
As director of a cancer center that focuses on mitigating health inequities in its catchment area, DuBois sees a potential for mtFIT to provide an inexpensive screening option that is more sensitive than FIT.
The Mark Foundation is an investor in the development of mtFIT.
The ideal colorectal cancer screening test would not just detect more colorectal cancers at an early stage, but actually detect more high-risk lesions before these have become invasive, and thus a full-blown cancer.
Gerrit Meijer
“[The trial results] confirmed that incorporating additional protein biomarkers into stool testing improves our ability to increase the rate of detection of high-risk pre-cancerous lesions,” DuBois said to The Cancer Letter. “These polyps can be easily removed before they turn into cancer, as a cancer prevention and interception strategy.”
The performance of mtFIT is evidence that single-disease tests continue to have a role in oncology, experts say—even as the field is investing heavily in the development of multi-cancer detection tests, or MCDs, also known as multi-cancer early detection tests, or MCEDs (The Cancer Letter, Jan. 12, 2024; Nov. 17, 2023).
In the Dutch trial, mtFIT—which detects two protein biomarkers in addition to hemoglobin—demonstrated:
- About double the positivity and detection rates for advanced neoplasia, at 9.11% and 2.27% respectively, compared to 4.08% and 1.21% for traditional FIT,
- About double the detection rates for advanced adenoma and advanced serrated polyps, at 1.64% and 0.43% compared to 0.86% and 0.17% with FIT, and
- A higher detection rate for colorectal cancer, at 0.20% compared to 0.17% with FIT.
These findings are based on valid test results received from 13,187 participants aged 55-75 who were provided with both the mtFIT and FIT tests.
mtFIT detects a greater number of larger polyps without a significant increase in false-positive results, thus avoiding unnecessary colonoscopies, study authors say.
The new modality has not been compared with FIT-DNA.
“mtFIT detects more advanced adenomas than FIT, which—based on the mathematical modeling—over the course of full participation in a screening program would lead to 21% colorectal cancer incidence reduction and 18% mortality reduction when compared to screening programs that use a high FIT cut-off, like in the Netherlands,” Gerrit Meijer, principal investigator of the mtFIT trial and senior author of the Lancet paper, said to The Cancer Letter.
mtFIT has been developed for over a decade in Meijer’s laboratory at the Netherlands Cancer Institute (NKI), where he is the head of research and innovation at the Department of Pathology. Meijer is also a professor of oncologic pathology at University Medical Center Utrecht.
“mtFIT showed a significantly higher sensitivity for advanced neoplasia than did FIT (42·9% vs 37·3%), while maintaining an equally high specificity of 96·6%,” the authors wrote, citing a 2021 retrospective study of a prospective controlled colonoscopy-controlled cohort of 1,284 participants.
Many countries have introduced population-based screening programs that rely on FIT to successfully diagnose early-stage colorectal cancer and reduce mortality—including in Europe, Asia, and Australia.
“FIT is good in detecting colorectal cancer in a single screening round, and individuals who participate in all rounds of a screening program would have multiple opportunities to have a cancer being detected before it becomes symptomatic, and then often be at a late stage,” Meijer said. “However, the ideal colorectal cancer screening test would not just detect more colorectal cancers at an early stage, but actually detect more high-risk lesions before these have become invasive, and thus a full-blown cancer.
“By far, the most common high-risk precursor lesions are advanced adenomas, which in practice, often are large polyps. In most cases, these lesions can still be removed during colonoscopy, so a surgical resection is not needed anymore,” Meijer said. “This is a particular spot where the current FIT leaves room for improvement, and where indeed mtFIT does better than FIT, as you can read in the paper.”
At last year’s annual meeting of the American Association for Cancer Research, experts agreed that it is too early to abandon proven single-cancer tests, Meijer said.
“MCED is starting to get over the top of the hype cycle and still quite a number of outstanding questions remain, e.g., related to the WHO Wilson and Jungner criteria,” Meijer said, noting that NKI also has a large research portfolio on MCEDs.
“I would argue that for early detection of colorectal cancer, currently, disease-specific tests prevail, but when well executed, clinical utility studies would demonstrate non-inferiority, Meijer said. “Then, of course, nothing would be against considering that option.”
At the moment, it’s too early to predict the clinical utility of MCEDs, given that careful validation is needed, DuBois said.
“Cancer screening tests are currently evolving. MCEDs are being evaluated as an option to detect multiple cancers in patients using a single sample,” DuBois said. “They may prove to be useful for colorectal and other cancers. One concern is that MCED tests might be so sensitive, if positive, it could be difficult to detect the actual location of the cancer due to current limitations of medical imaging techniques. Thus, it might be difficult to take any meaningful action, other than to increase patient concern.
“However, the mtFIT test has been shown to be a marked improvement over FIT for colorectal cancer and the projected relative low cost of this test might make it useful in underserved and underinsured patient populations,” DuBois said. “If the mtFIT test is positive, then it is clear what action needs to be taken in terms of a diagnostic work up, which would be for the patient to undergo a colonoscopy to determine the presence or absence of colorectal cancer.”
The improved performance of mtFIT, compared to FIT, holds up across three different scenarios that the study authors used to evaluate different cutoff points for the diagnostic test.
“The three scenarios described by the authors are related to specific rules about the way the FIT test is used in different countries—essentially, the choice of a cutoff value where you call a test ‘positive,’” DuBois said. “The authors tested three cutoff values that represent the range of likely choices if the test were to be implemented in the current system.
“Testing all three scenarios is important, because they didn’t just cherry-pick the cutoff that gave the most favorable outcome. In all scenarios, the mtFIT test is predicted to improve detection of early lesions and prevent death from colorectal cancer.”
Comparing recommended stool tests
The United States Preventive Services Task Force recommends screening for colorectal cancer with high certainty of substantial benefit (grade “A”) in all adults aged 50-75 years, and with moderate certainty of substantial benefit (grade “B”) in adults aged 45-49 years (The Cancer Letter, June 22, 2016; Oct. 9, 2015).
Although the U.S. currently relies primarily on colonoscopy to screen for colorectal cancer, there are many barriers—including cost, access, and patient willingness—that leave a huge unmet need, DuBois said.
“Fecal testing in the form of FIT and FIT-DNA test have begun to make inroads in the U.S.—but is much more widely adopted in Europe, Asia and Australia—and this mtFIT study shows further promise for a highly sensitive, non-invasive, inexpensive testing option,” DuBois said. “Given the findings of this study, I strongly support the continued research and development of the mtFIT test.”
Of the screening modalities recommended by the task force, stool-based tests include high-sensitivity gFOBT (guaiac fecal occult blood test), FIT, and sDNA-FIT, or FIT-DNA.
FIT is the easiest test to administer, as it only requires a single stool sample, compared to multiple samples for gFOBT or a full bowel movement for sDNA-FIT.
While mtFIT is more sensitive than FIT, it has not been compared to FIT-DNA, which also outperforms FIT. However, Mark Foundation officials say they expect that mtFIT would cost less than FIT-DNA screening.
“mtFIT has not been compared head-to-head with Cologuard or any FIT-DNA test, and any attempt to make a claim about relative performance would be speculative,” Becky Bish, head of discovery and preclinical research at The Mark Foundation, said to The Cancer Letter. “It would be very interesting to do that study.”
For a positive FIT, FIT-DNA, or mtFIT test, the standard workup is colonoscopy to detect and treat polyps and other lesions.
According to DuBois, the advantage of mtFIT is that it combines the ease of administration of FIT with better detection rates.
Said DuBois:
- Traditional FIT tests use an antibody-based method to measure a single protein, i.e. hemoglobin, in the stool. Hemoglobin is a protein derived from blood, and the presence of blood in the stool can be a symptom of cancer. FIT has logistical advantages in that it requires only a small sample of stool that can be shipped by postal mail.
- The mtFIT test uses an antibody-based method to measure three proteins in the stool: hemoglobin (like the FIT test), plus two additional proteins (calprotectin and SERPINF2). The technology is similar to the FIT test (which helps keep the cost low and makes implementation easier) but increases sensitivity by measuring three biomarkers instead of one.
- FIT-DNA tests combine a FIT test with measurements of DNA biomarkers in the stool—for example, testing for genes that are typically altered in colorectal cancer. The FIT-DNA test involves multiple types of assays (for proteins and nucleic acids), which increases the cost. In addition, the test requires a large sample of stool (whole bowel movement) collected in a bucket, which logistically is more demanding. This test is marketed as Cologuard by Exact Sciences Corp. in the U.S. [The Cancer Letter, March 28, 2014]. Cologuard is not used in the large population-based screening programs organized in many countries outside the U.S.
Commercializing mtFIT
Calprotectin and SERPINF2 were identified as promising biomarkers by Meijer’s lab.
“Sample logistics and analytical logistics are the same as for the current FIT, facilitating the adoption of the new test in FIT-based screening programs,” DuBois said.
“Professor Meijer and Dr. Meike De Wit at NKI have founded a spin-off company called CRCBioscreen to develop and commercialize the [mtFIT] test,” DuBois said. “The Mark Foundation made a critical investment in CRCBioscreen to support the clinical trial.
The mtFIT test has been shown to be a marked improvement over FIT for colorectal cancer and the projected relative low cost of this test might make it useful in underserved and underinsured patient populations.
Raymond DuBois
“This investment was key in obtaining matching funds from the Dutch government to cover the remaining costs of the trial. This is a truly unique model of a non-profit foundation, Mark Foundation, coming together with private investors and government support to get a high-impact trial done quickly for the benefit of cancer patients.”
The study was funded by KWF Dutch Cancer Society, Health Holland, Maag Lever Darm Stichting, Stand Up To Cancer’s (SU2C) and CRCbioscreen, supported by The Mark Foundation for Cancer Research and Human+.
“The next step will be to develop an [European Union] In-Vitro Medical Device Regulation-certified clinical grade test, to follow-up on the research-use-only test that was used in the study,” Meijer said. “We anticipate that this development, including the clinical validation involved, will take approximately four years.
“To this end, the CRCbioscreen company has been formed as a spin-off of the Netherlands Cancer Institute to develop this test commercially, with funding from the Mark Foundation and others.”
Meijer is a co-founder and the chief scientific officer of CRCbioscreen, which holds the exclusive license to develop mtFIT. NKI owns the intellectual property rights, and The Mark Foundation is an investor in CRCbioscreen.
“Through the investment side of our organization, The Mark Foundation provides venture capital funding to companies with rigorous research programs that have transformative potential in cancer,” Ryan Schoenfeld, CEO of The Mark Foundation, said to The Cancer Letter. “Any returns realized from our investment portfolio are put back into the foundation to enable additional support for cutting edge cancer research in the form of new grants or investments.”
Once a commercial clinical grade mtFIT becomes available, it would be logical for clinicians to switch to mtFIT over FIT for screening indications, DuBois said.
“mtFIT also is likely to be the most cost-effective choice compared to other stool tests,” he said. “Rates of colorectal cancer are dramatically increasing in younger adults, but this population generally hasn’t been recommended to undergo screening colonoscopies.
“Increased adoption of fecal testing as a screen for colorectal cancer has tremendous promise for early detection and prevention, and the increased sensitivity of mtFIT for precursor lesions would be especially beneficial in this context.”
Developing mtFIT
As the leader of the Diagnostic Translational Oncology section at NKI, Meijer has dedicated his scientific pursuits over the past decade to developing mtFIT, with the goal of improving early detection of colorectal cancer.
Also, he serves on committees for the implementation of the Dutch national colorectal cancer screening program.
“A first impetus was that early detection and intervention is the most realistic approach to prevent colorectal cancer death rates and at the same time reduce colorectal cancer incidence and associated morbidity,” Meijer said. “Yet, the speed of development in early detection research versus research focusing on late-stage disease is out of balance.
“More specifically, most widely used current colorectal cancer early detection approaches, i.e., FIT and colonoscopy, are already several decades old, and these are not perfect, so there is an evident need to improve on these.”
Meijer’s efforts to develop mtFIT—prior to the recently completed prospective interventional clinical trial comparing mtFIT to FIT—can be summarized in two stages:
- In a discovery and pilot study, mass spectrometry was used to compare stool samples from patients with colorectal cancer vs. controls, and 27 candidate protein biomarkers were identified. Antibody-based assays—of the same type used in traditional FIT testing—were developed for a few of the most promising new biomarkers. This study was published Nov. 21, 2017 in the Annals of Internal Medicine.
- In a second retrospective study, the most promising combination of the 27 candidate protein biomarkers was identified, using a series of 1,284 biobanked stool samples from cancer patients and healthy controls. This clinical validation study demonstrated that the combination of hemoglobin, calprotectin and SERPINF2 yields the best diagnostic performance. This 3-protein combo was coined multitarget FIT, or mtFIT. In this study, mtFIT detected 35% more high risk CRC precursor lesions—advanced adenomas and large polyps—than traditional FIT at a specificity of 95%. These results were published July 20, 2021.
Meijer’s remarks on his research journey follow:
In the past, we have investigated DNA methylation markers in stool, but methylation-based stool tests require large sample sizes. The logistics of these large samples is not compatible with the logistics of large-scale population-based CRC screening programs.
That is why, in the Dutch Cancer Society/StandUp2Cancer-funded MEDOCC project, we have investigated 1) whether new technologies would allow us to reliably measure DNA methylation markers in smaller stool samples. In parallel, we 2) investigated whether by looking at multiple proteins rather than just hemoglobin (=FIT) we could improve the performance of current FIT.
Track 1) did not work out in the end, but track 2) has been successful and resulted in the recently published study.
First, we performed a discovery study using shotgun mass spectrometry looking into about 300 biobanked stool samples of individuals with or without CRC or advanced adenomas during colonoscopy. This yielded some 27 promising candidate protein markers.
In a second study, we selected 10 candidates out of these 27 to further investigate. To that end, we shifted from shotgun mass spectrometry to antibody-based assays, as are being used in FIT-based screening.
In addition, we moved from large samples to the small samples that are being used in current FIT-based CRC screening and tested these over 1,200 FIT samples, from a screening population enriched for CRC and advanced adenomas with known colonoscopy outcomes.
The technology is similar to the FIT test (which helps keep the cost low and makes implementation easier) but increases sensitivity by measuring three biomarkers instead of one.
Raymond DuBois
Long story short, this yielded the combination of hemoglobin, calprotectin and SERPINF2, referred to as multitarget FIT. In this study, mtFIT demonstrated higher sensitivity for in particular advanced adenomas than FIT, without losing specificity (which was 97% for both FIT and mtFIT).
The Lancet Oncology study was the third stage of this development journey. Here, we validated mtFIT in a real world CRC screening setting, as described in the paper and the press release. It is important to note that only participants with either a positive mtFIT and/or a positive FIT (cut-off 47 µg/g) were referred for colonoscopy.
So, the endpoint was to see whether at equal positivity rates, mtFIT would detect more true positives, in particular individuals with advanced adenomas.
mtFIT detects more advanced adenomas then FIT, which—based on the mathematical modeling—over the course of full participation in a screening program would lead to 21% colorectal cancer incidence reduction and 18% mortality reduction when compared to screening programs that use a high FIT cut-off, like in the Netherlands.
Since in the approved study design we could not head-to-head compare mtFIT to FIT at a low cut-off (= equally high positivity rate as mtFIT), we explored two workarounds.
In scenario 2, we recalibrated the mtFIT decision tree algorithm to arrive at a positivity rate of 4% just like FIT. In that scenario, mtFIT still detected 17% more advanced adenomas than FIT, translating into a predicted extra 5% colorectal incidence reduction and 4% mortality reduction compared to FIT, showing that the effect in scenario 1 was not just caused by the difference in positivity rates.
Because increasing the cut-off of mtFIT could provide a disadvantage for mtFIT, we also explored scenario 3, where we estimated the positivity rates of FIT at a low cut-off based on historical data from earlier prospective screening studies. Evidently, this carries a higher level of uncertainty, but the findings are at least in line with the other two scenarios.
