Debating the Clinical Trial Upending Colonoscopy Practices

F. Perry Wilson, MD, MSCE; David A. Johnson, MD; Kenneth W. Lin, MD, MPH


November 28, 2022

This transcript has been edited for clarity.

F. Perry Wilson, MD, MSCE: Hello, and thank you for joining us today for what promises to be a lively discussion about screening for colon cancer.

My name is Perry Wilson. I’m an associate professor of medicine and director of the Clinical and Translational Research Accelerator at the Yale School of Medicine. My new book, How Medicine Works and When It Doesn't: Learning Who to Trust to Get and Stay Healthy, is available for pre-order now anywhere that books are sold.

I'm joined by two wonderful experts. Dr David Johnson is a professor of medicine and the chief of gastroenterology at the Eastern Virginia School of Medicine. He is the past president of the American College of Gastroenterology. And I'm very encouraged to see that he's won a Distinguished Educator Award for his efforts in gastroenterology.

I'm also joined by Dr Kenny Lin. He's a frequent contributor to Medscape and WebMD. He's a family physician and public health consultant from Lancaster, Pennsylvania, and deputy editor of the American Family Physician journal. He's also a teacher of residents and students at Lancaster General Health and the Penn Medicine Family Medicine Residency program.

So, we have two great educators with us today to hopefully help teach us something about colon cancer and colon cancer screening. Thank you for joining me today.

David A. Johnson, MD: Thanks for having us.

Kenneth W. Lin, MD, MPH: Good to be here.

Wilson: Colon cancer is the second leading cause of cancer mortality in the United States. A little over 50,000 people die every year in the United States due to colon cancer.

A month ago, I would have said that there was a pretty broad consensus, at least from my perspective, that people should be getting colonoscopies. That's certainly what we tell our patients.

Then a paper came out in The New England Journal of Medicine (NEJM), a very prestigious journal, that has caused a lot of consternation online and led to my receiving a lot questions from patients and their family members. Today, I’d like to talk about this randomized trial of screening colonoscopy for colon cancer, and why it has caused so much—perhaps—confusion, calls for change, and concern out there.

Dr Johnson, can you give us a brief overview of what this trial was about?

Johnson: This was a randomized trial looking at screening colonoscopy versus no screening test whatsoever. They looked at the outcomes of prevention of cancer and the prevention of colon cancer-related death.

The short answer was that it was disappointing as it relates to colonoscopy. The study looked at patients from four European countries, with data from three of them (Norway, Poland, and Sweden) ultimately analyzed in this report in NEJM. It got a lot of attention because it surprised a lot of people by saying maybe colonoscopy wasn't quite as good as we thought it was.

They tried to correct that by only looking at the numbers of patients who got their colonoscopy screening, which still showed value, but it was less than that we've seen before. There’s lots of reasons for that, which we’ll discuss shortly.

An Invitation to a Screening

Wilson: This was a bit of an interesting trial design. I think I'm correct, Dr Lin, that this was the first randomized trial of screening colonoscopy. But they didn't really randomize people to get a colonoscopy versus not get a colonoscopy. Can you tell us why this differed from that study design, which I’d have thought would be simpler way of assessing this?

Lin: It's definitely an important point to highlight about the study. What investigators did was randomize patients to receive an invitation to get a screening colonoscopy. When the trial was set up, they randomized people before they were asked whether they wanted to participate in the study. If you did it the other way around, by first asking them whether they wanted to be in the study and then randomizing them, you would have been assured that more of them probably would have gotten the colonoscopy.

But in this case, they were more interested in figuring out the real-life results of having a national program that invited patients to receive screening colonoscopy. Because we know that everyone that you recommend to get a colonoscopy doesn't necessarily want to do that, forgets to do it, or something happens that prevents their actually getting it.

When it comes to measuring the effectiveness of the colonoscopy, it perhaps wasn't the greatest type of study to do that. But I think it did provide some information about what would happen if you invited people to get colonoscopy, in terms of how many would do it and the results overall for that population.

Lower Participation Numbers Than Excepted

Wilson: Dr Johnson, the data show that 42% of people who were in that invitation arm followed through and got their colonoscopy. You're a gastroenterologist. Does that seem low or about right? Do about half of people who should get a colonoscopy end up getting one?

Johnson: No, it's low. In the United States, those numbers are probably in the 70% range. Certainly, the test doesn't work for people who don't get the test performed. So, if 42% of those randomized to receive an invitation to get the colonoscopy got one, that really means the majority of patients never got the test.

Wilson: Certainly, we wouldn't expect impressive results if they don't get the test. But on the other hand, I imagine that people who choose to get the test when they're invited are sort of a different breed. Perhaps they’re more health conscious or living in other healthy ways. Is that something we should worry about when we look at these results?

Johnson: I don't think you can stratify based on this study. Factors like ethnicities and diet weren't really explained. The key element that will hopefully have the major take-home impact is quality. It's not just the test. It's how the test is done.

The Key Results

Wilson: Let's start with the big picture. This was a study looking at everyone invited; not the subgroup of people who got the colonoscopy, but the real randomized study population.

Dr Lin, the study did show that the invited group had a lower risk of colon cancer over the next 10 years. That's a good thing, I imagine.

Lin: I think that's a significant benefit. Initially in the first few years, they had more colon cancers diagnosed. But that's probably because those were cancers that were already existing and couldn't be prevented by the test.

But then over the years the curves crossed, and by the end of the average follow-up of 10 years, there was a significantly lower rate of colon cancers being detected. That’s as you would expect, because you're finding polyps and removing them before they became colon cancer.

Wilson: Dr Johnson, is that the natural history of colon cancer? It starts out as a polyp that maybe can be easily removed and doesn't require more therapy. Is that why screening colonoscopy is helpful?

Johnson: The ultimate goal of screening is prevention of cancer, rather than detection of cancer. That occurs by identification and complete removal of the polyps that we find that are precancerous. The key is, first, detection, and second, resection. Adequate resection comes down to some very significant issues of quality, which are questions that I'd raised about this study, and we can talk about momentarily.

Wilson: Absolutely. Let me first go through the two other big findings in this study.

The fact that there were fewer cases of colon cancer over 10 years seems good. But colon cancer mortality was not significantly different in the two groups. Now, of course, we know that not everyone got a colonoscopy. I would have expected though, if you had less colon cancer, you'd have less death from colon cancer.

Dr Lin, what might explain this disconnect?

Lin: I think there are a couple of possible explanations.

One explanation is that they just didn't follow the people long enough. Colon cancer takes a long time to go from an adenoma to cancer, and from cancer to something that would cause the patient's death. You may need to follow them for longer than the 10 years that most of these patients were followed to see that benefit. I think there probably will be benefit after a while, because if you are removing colon cancers that otherwise would have progressed and metastasized, you often see a benefit.

We also have to consider the other possibility that not all the polyps removed necessarily were going to progress to advanced cancer. Therefore, you weren't seeing the death benefit because not every polyp that was removed was necessarily going to cause health consequences.

In Colonoscopy, Quality Is Key to Success

Wilson: You're removing things and have no way of knowing in advance which are the bad ones and which aren't.

Dr Johnson, you’ve mentioned several times now that the quality of colonoscopy matters here. So, I'm intuiting that it's not one-size-fits-all, that it's not all the same. What do you mean by quality of colonoscopy, and what was it in the NEJM study?

Johnson: Quality colonoscopy is the quality of the whole process. It starts with the warm-up, if you will, and the clean out for the procedure. That allows the colonoscopist to be able to identify precancerous polyps, which we call adenomas (there are other precancerous polyps called sessile serrated lesions).

The identification of adenomas is extremely important. Even a small increase in the detection of those precancerous polyps has benefits. Well-performed studies looking at large databases show that a small, 1% increase in the adenoma detection leads to a 3% decrease in colon cancer and a 5% decrease in colon-cancer related death. There’s a huge array of effect when we talk about small increases in the adenoma detection rate.

Now, let's go back to this study in NEJM.

If we base quality on the physician performing the colonoscopy, and say that the colonoscopy is achieving the act of getting all the way around the colon, but not all physicians in the study were able to do that, it starts to raise the question about quality, because adenoma detection is so important. Earlier reports from this group [Nordic-European Initiative on Colorectal Cancer Study Group] have shown that the adenoma detection rates have been way below the national thresholds. So, this raises the question of whether they found the polyp, and then whether they resected the polyp. They also don't tell us where these cancers were. It is about the colonoscopy quality. It's not the instrument. It's the process.

An Overview of Other Screening Tools

Wilson: Dr Lin, colonoscopy, which requires prep and anesthesia, is not the only colon cancer screening method we have. In fact, there are a bunch. I think we're on board saying it's probably better to detect colon cancer early than not detect it. But what are our other options aside from colonoscopy that can allow for early detection of colon cancer?

Lin: For most of my career, there were three options that I presented patients with. The first was the fecal test, which used to be in the form of initial hemoccult tests. These have been mostly replaced by fecal immunochemical testing. But they're both just basically looking for the presence of blood in the stool. Anyone who has a positive test would be referred for a diagnostic colonoscopy.

The other test besides colonoscopy, which has been largely phased out in the United States, although it is still very much used in Canada and much of Europe, is flexible sigmoidoscopy. Until this study, the tests supported by randomized controlled trials were the fecal tests and flexible sigmoidoscopy.

Interestingly, there was a recent systematic review of flexible sigmoidoscopy looking at four trials and their effects over 15 years. They showed not only a reduction in colon cancer, but also a reduction in colon cancer mortality, and even a small reduction in all-cause mortality.

I believe three out of the four trials were done where the patients were consented and then randomized, so they had a higher uptake of the procedure.

But when you compare this with the colonoscopy trial, it really isn't that impressive. You would expect a much larger benefit, because obviously you're looking at the entire colon. But you really didn't see that. It was, at best, maybe equivalent to sigmoidoscopy, but not a whole lot better.

Johnson: Perry, you mentioned sedation. It’s important to understand that this particular cohort of patients are from Norway, Sweden, and Poland, where it’s very much the norm to not get sedation for your colonoscopy. Any of the listeners who have had colonoscopy will tell you that they are not ones to say, “Don't give me sedation.” The rate of sedation is around 11% in Norway, maybe 23% in Sweden, and around 45% in Poland. So, the examiner and the patient were never really super comfortable.

I've done 50,000 colonoscopies in my career, and many non-sedated. We know that taking time increases the finding of polyps and the adequate identification and resection. So, that ability to perform at a high quality is very much impacted when the patients aren't comfortable.

Wilson: Dr Johnson, we brought up flexible sigmoidoscopy. For the patients watching whose doctors are talking to them about screening colonoscopy, what's the difference?

Johnson: Flexible sigmoidoscopy is just a short scope examination, in which you see about one third of the colon. I’ve been in the field for 45 years, and during that time we've seen that there's a progressive increase in the development of cancers above that bottom third of the colon to the higher end, the two-thirds of the colon that you would miss without doing a full colonoscopy. Also, flexible sigmoidoscopy typically does not get covered for sedation.

Again, if you do the exam and find something, then you're going to have to come back and do an adequate resection with a colonoscopy. So, one-stop-shopping colon cancer screening is not about detection of cancer, it’s about prevention of cancer, and that's what colonoscopy does.

Patients Want Convenience, But at What Cost?

Wilson: Dr Lin, how are your patients in your family practice handling this study? Have conversations changed around colon cancer screening? What are people asking about these days?

Lin: I don't think the conversations have changed in my practice that much. When patients ask about this study, we do discuss the limitations, that it wasn't designed to assess the maximum benefit of getting a colonoscopy because the majority of people assigned to that group didn't get colonoscopy.

But I think it is an opportunity in primary care to consider the way we present the options to patients. Because I would guess that a majority of primary care physicians, when they present the options, would say colonoscopy is the gold standard and recommend their patients get it. And they only offer fecal testing to patients who don't want the colonoscopy or really refuse.

That hasn't been my practice. I'm usually more agnostic, because there are both harms and benefits. If you get a fecal test, the chance of you having a complication from colonoscopy is automatically lower because most of those people will not get colonoscopy. Now obviously, the complications with colonoscopy are pretty rare and usually self-limited, but they do exist. If you're doing lots and lots of these, eventually you’ll see them. Probably all primary care physicians have patients who've had a complication from colonoscopy and may or may not have regretted it depending on how information was presented.

But I feel like this study reinforces my feeling that we ought to be presenting these, and not saying one is superior or inferior to the other. Instead, I’d base it on what the patient's priorities are. Is your priority finding every single cancer? Do you want to know exactly what the benefit is? I think with colonoscopy, we're still trying to figure out exactly what the benefit is. Whereas we can say it pretty confidently for fecal tests because we have those randomized trials.

Wilson: Dr Johnson, I think patients who are watching need to know, first of all, that if they do the fecal test route, a positive fecal test does lead to colonoscopy. In some sense, all roads lead to colonoscopy once you have a positive screening test. So, I can certainly see the value of just sort of skipping to that point. But what about this risk-vs-benefit relationship? Colonoscopy, albeit a relatively safe procedure, is still a procedure. There is some risk associated with it. If we can get the same benefit from yearly fecal immunochemical testing, is that a better choice potentially, at least for patients at average risk?

Johnson: The stool-based testing is really more effective for detection of cancer. That's not screening, where the entire goal is the prevention of cancer. The fecal-based testing, including the stool-based DNA testing, misses the majority of precancerous polyps. And the fecal immunochemical tests, which Dr Lin just mentioned, misses virtually all of them. We really want to get to the prevention of cancer, meaning identification and removal of polyps, not just screening for cancer.

Wilson: Do you see anything on the horizon that could unseat colonoscopy as, to quote Dr Lin, the potential gold standard for screening for colon cancer?

Johnson: I think not on the horizon for identification and removal of polyps. That's really the gold standard. Technology continues to advance. We'll see what happens. But on the short and intermediate horizon, colonoscopy is going to be needed.

We are finding that some patients are starting to acquiesce to stool-based testing because they can do it at home. Maybe they don't have to do a prep. We're talking about screening only here, not about the follow-up of patients who have a family history, patients who have colitis, patients who have had colon polyps, or other reasons. Stool-based testing is not an option for the follow-up of those patients.

Convenience testing, in the face of COVID, also has thrown a wrench into things. Patients may have wanted to stay home and do these tests. Again, we need to be proactive, not reactive. We want to prevent cancer, not detect it.

Changing Advice in the Face of Younger Screening Thresholds

Wilson: Dr Lin, I’m 42 years old. I don't believe I’m at any increased risk of colon cancer based on my family history or other risk factors. I'm 3 years away from when the US Preventive Services Task Force tells me I should potentially consider starting to screen for colon cancer. That recommendation has recently been moved down from 50 years old to 45 years old. So, it's on my mind as I approach that age. What do you advise younger patients approaching 45 right now in terms of screening for colon cancer?

Lin: For patients with the risk factors that Dr Johnson mentioned, I would recommend screening colonoscopy as the initial test.

Assuming you don’t have those risk factors, I present it as we have a couple of different fecal tests. There's the traditional one that just looks for blood. Then there's the newer one that also adds DNA, which is more sensitive for colorectal cancer, but a little less specific, which is a problem just because there are more false positives.

But you need to compare that with colonoscopy, which you only need to get done ideally every 10 years if there are no findings. That is more complete. And theoretically, as we've been talking about, it would also prevent as well as detect early cancers.

So, I think it's really down to your preference in terms of how the various factors that come into play, such as convenience of the test and your level of concern about cancer. I do tell patients that family history of cancer is not terribly predictive of whether you get it or not. A lot of people unfortunately who develop colorectal cancer have no previous family history. Diet will come into play to some extent. There are some things that point to increased risk for colorectal cancer if you have a diet high in red meat and things like that. But ultimately, it really is up to the patient. I lay out the options, and whatever they choose, I'm happy to pursue.

But the most important thing is that they do some test, because doing no test is not going to help anyone. I do agree with the notion that the best test is the test that gets done.

Wilson: Absolutely. I think the NEJM study supports that, even when we're talking about colonoscopy.

Dr. Johnson, you've had some criticisms about the NEJM study, and I think they make sense. At the same time, as this is the first randomized trial of colonoscopy, it’s kind of the only data we have. Are we going to get better data? Are there other studies going on out there that might help shed some light on what's turning out to be a complicated issue?

Johnson: Yes, there are ongoing studies. They’re not taking place within the United States, because you couldn't get through a no screening option trial. There are comparative studies that are probably still 5 years away looking at stool-based testing.

But again, we have to recognize that if you do these alternative tests that were eloquently discussed by Dr Lin, and not the colonoscopy, which would be every 10 years with high-quality performance, that you have to annualize or do them in sequence. It's important that you follow up on those with regularity. It's not just a one-time test every 10 years for these individual tests.

And any of the time that those tests are ordered, the patient should be instructed that if it's positive you need a colonoscopy. We're seeing a lot of slippage on that front for the stool-based testing. Convenience is not the answer. It's getting the job done.

Wilson: Would you agree, Dr Johnson, that for patients that really don't want to do the colonoscopy for one reason or another, and you've done your best in explaining what you think the risks and benefits are, that you'd rather have them get something than nothing?

Johnson: Absolutely. It comes down to what I recommend and then what you decide. But I still make the point explicit: if we've gone through those checkpoints and it’s positive, we agree that you understand that colonoscopy is the next step.

Final Take-Home Messages

Wilson: Dr Lin, I’ll turn the last word over to you, as the person who is probably discussing the choice of screening modalities more than any of us, before someone would get referred to someone like Dr Johnson. What's your final take-home message about the NEJM study and the state of colon cancer screening in the United States?

Lin: My take-home points about the study are that there were some limitations, but it is good to finally have a randomized trial of colonoscopy screening two decades after we really started doing that in the United States. It won't immediately change—nor do I think it should—the way we practice and discuss different options. I think that some of Dr Johnson's points about making sure that whoever's doing the colonoscopies for your practices is doing it in a high-quality way are really important. Just as it's important, if you're doing the fecal tests, to make sure that all patients who have positives get expeditiously referred for colonoscopy.

Johnson: Perry, I’d like to make one concluding comment as the gastroenterology expert in this discussion. I've had countless questions about this study from my patients and my peers. I tell them the following: don't let the headlines mislead you.

When you look at this study, the instrument is not so much the question. We know that getting the test is the first step in colon cancer screening. But we also know that getting the test done, with the highest-quality providers and the best-quality performance, is really the key to optimizing the true value of colonoscopy for colon cancer prevention.

So please don't lose sight of this when reading the headlines in the media around this study. We really need to analyze the true characteristics of what we call a quality performance, because that's what drives success and that's what prevents colon cancer.

Wilson: Dr Johnson and Dr Lin, thank you very much. I appreciate you spending time with me here today and wish you all the best.

I guess I'll sum up by saying that if you're getting a colonoscopy, make sure it's a good one. But do get screened.

This video originally appeared on WebMD on November 28, 2022

Follow Medscape on Facebook, Twitter, Instagram, and YouTube