乌鸦传媒

Could a new study that suggests colonoscopy is less effective than expected thwart efforts in the U.S. to get more people screened for colorectal cancer?

Medical associations and physicians are weighing in on the subject, noting that they are worried that the results from the randomized NordICC trial, published earlier this week in the , are being interpreted in a way that may discourage people from undergoing screening.

The study, led by Michael Bretthauer, MD, PhD, of the University of Oslo in Norway, showed that individuals invited to undergo screening colonoscopy had a reduced risk of colorectal cancer compared with those who did not undergo screening, although at a lower rate than researchers had expected. Furthermore, they found no significant difference in colorectal cancer-related mortality between the two groups.

"We've all been trying to boost colorectal cancer screening, and what we don't want is to derail that effort or have our patients think that screening is not effective and forego it," Aasma Shaukat, MD, MPH, of the NYU Grossman School of Medicine, told 乌鸦传媒. "We should mitigate any impression that this should deter, demotivate, or discourage people from undergoing colorectal cancer screening."

Shaukat suggested that the results are not being interpreted deeply enough to understand the study's limitations and what it means in the context of other available evidence on the topic that "overwhelmingly supports that colorectal cancer screening is effective in reducing the incidence of colon cancer and reducing deaths, and a lot of it is much more relevant because it comes from U.S. populations."

George J. Chang, MD, of the University of Texas MD Anderson Cancer Center in Houston, pointed out that "this is not data saying nobody should get a colonoscopy," adding that endoscopy remains the only screening modality that actually prevents cancer because it can include polypectomy.

"Those are key messages," he told 乌鸦传媒. "However you get screened, get screened. But, I know that endoscopy does allow you to prevent cancers and I would not steer away from it based on this work."

Concerns about the study seemed to focus on three main points -- the fact that less than half of individuals randomized to colonoscopy actually underwent the procedure, the question of operator quality, and a follow-up that was probably underpowered to accurately determine the risk of colorectal cancer-related mortality.

Participation Rate

The NordICC trial involved 84,585 healthy men and women ages 55 to 64 from population registries in Poland, Norway, Sweden, and the Netherlands from 2009 to 2014. Participants were randomly assigned 1:2 to an invitation to undergo a single screening colonoscopy or to usual care (no screening).

However, only 42% of those invited to undergo colonoscopy actually underwent the procedure. "It's hard to know the value of a screening test when the majority of people in the study didn't get it done," said William Dahut, MD, chief scientific officer of the American Cancer Society, in a .

In Norway, where participation in the trial was higher, at 60.7%, a greater reduction in cancer incidence was seen.

Raymond N. DuBois, MD, PhD, director of the Hollings Cancer Center at the Medical University of South Carolina in Charleston and a former president of the American Association for Cancer Research, put the 42% rate into perspective by comparing it with in which 58% to 87% of people invited were actually screened.

"So that's a much higher rate of participation," DuBois told 乌鸦传媒. "And it could be that lower participation [in the current study] could have skewed the results."

In fact, among patients who actually underwent colonoscopy, the risk of colorectal cancer and cancer-related mortality was substantially reduced. In an adjusted per-protocol analysis, Bretthauer and colleagues found that if all those who were randomly assigned to screening had actually undergone colonoscopy, it would have reduced the incidence of colorectal cancer by 31% and the risk of colorectal cancer-related death by 50%.

Thus, "that's message number one," Chang said. "Colonoscopy is not effective if you don't do it, and so the effect was probably a lot smaller than expected because of the low compliance rate."

Folasade May, MD, PhD, of the UCLA Jonsson Comprehensive Cancer Center in Los Angeles, told 乌鸦传媒 she didn't find the 42% rate surprising. She noted that NordICC took place in countries without national colonoscopy guidelines, which means the conversation about screening is going to be different than in the U.S. where guidelines suggest that average-risk adults be screened as early as age 45.

"So layer on top of that the fact that now it's an invitation, potentially from your non-provider, in a place that already sees screening as optional," she said. "So, I'm not surprised that only 42% of people participated. And to me, that's the most interesting finding of the entire study, which is that if we put out an invitation for something that we know is likely lifesaving, it's not enough. We need to do more than just invite people to participate in this potentially lifesaving mechanism."

Operator Quality

In a statement issued by the American Society for Gastrointestinal Endoscopy, Bret Petersen, MD, the society's president, noted that the quality of colonoscopies varies across the globe, depending on physician skillset and training, as well as availability of endoscopy tools and technology.

"Nearly 30% of the endoscopists who were included in the NordICC trial did not meet the adenoma detection rate (ADR) of 25% that is recommended in the U.S. -- while the ADR average in the U.S. is above 40%," Petersen noted.

That brings up concerns about the general quality of the examinations in the NordICC study, and whether or not the endoscopists in the study missed some adenomas, said DuBois. "We expect a much higher percentage of an adenoma detection rate."

Chang pointed out that "colonoscopy is operator dependent and we want to make sure that we have an expert who adheres to guidelines and has ... an appropriate adenoma detection rate. I'm not saying that the quality of the endoscopy was poor in the study, but this is another factor that consumers in the U.S. probably should consider in thinking about their choices for screening and the interpretation of this data."

Follow-Up

The authors of the NordICC trial did not see a reduced risk of dying from colorectal cancer, "and that's a little disappointing, but not unexpected," said Shaukat, adding that the study had a median follow-up of 10 years, which means that it is underpowered to accurately assess the risk of colorectal cancer-related mortality.

DuBois noted that it takes about 15 to 20 years to develop colorectal cancer. The authors "will do another analysis of the data in 5 more years, so it could be also that it was just too short of a duration, of observation, to see a significant impact. I would hate for people to jump off the screening bandwagon now because it might have been premature. The study needed more time to make these observations."

What to Tell Patients

Since the study was published in "a high-profile journal, with a high-profile topic, people are going to see it," said David M. Poppers, MD, PhD, of NYU Langone Health in New York City.

"They shouldn't jump to conclusions [based on the study] and should not shy away from what we have already established in multiple trials throughout the years -- that colorectal cancer screening is lifesaving," Poppers told 乌鸦传媒. This message is particularly important since many people are only just starting to reestablish care in the aftermath of the COVID-19 pandemic, he added.

"We are actually detecting more cancer, and more cancer in an advanced stage, than we would have otherwise, in part due to the pandemic and delayed screening," he pointed out. "We don't want to take our foot off the accelerator. Our goal is cancer prevention, and that starts with good quality screening, and colonoscopy remains the gold standard because it is unique in being able to directly visualize the lining of the colon."

Shaukat noted the importance of giving patients "our best advice -- which is to complete colon cancer screening, and also letting them know there are options other than colonoscopy, because at the end of the day, the best test is the one the patient is willing to complete. If they don't want to do a colonoscopy, but are willing to undergo screening, we have stool-based tests and imaging-based tests."

She also suggested there are generalizability issues with the study, since it took place in countries with overwhelming white populations. "In the U.S., we've been trying to boost screening in multiple racial and ethnic groups where the risk of cancer is particularly high," she said. "It is important that those groups are encouraged to undergo screening starting at age 45."

Study Author Also Weighs In

In an email, Bretthauer said that deciding on the type of colorectal cancer screening test should involve "informed decision making."

For colonoscopy, "the per-protocol analyses show a benefit of about 30% for incidence and 30% to 50% for mortality," he said. "I think this is a very credible benefit; few other cancer screening tests are that good."

Furthermore, Bretthauer said that he and his colleagues were transparent in their study, as far as giving the absolute risks and absolute risk reductions in the intention-to-treat and per-protocol analyses, adding that if one believes that reducing the risk of colorectal cancer from 1.2% in the usual care group to 0.98% in the intention-to-treat group or 0.84% in the per-protocol analysis demonstrates benefit, "then get a colonoscopy."

"If you think this benefit is too small to bother, then maybe don't get one," he added. "It is up to each person to decide."

Comment