We're going to interrupt our programming to bring you a very, very important announcement by the president of ASGE, Brett Peterson's here with us to give us some important update and perspective on the recent Nordic trial, which I'm sure you all have read and have questions about. I hope most of you are familiar with a recent publication in New England Journal of Medicine that came out in press on October 9th, Sunday evening, October 9th, and we'll talk a little bit about the publication, some contrasting data, and what it might do to our practices and what we should all be doing to counter some of the messages. So this is regarding what's called the Nordic trial of screening colonoscopy. The first author was Michael Brethauer from Oslo, Norway, a relatively young but distinguished investigator in his own right, and the senior author is from Warsaw, Poland, who's most well known for being the first to correlate adenoma detection rate with subsequent cancer rates, showing that it makes a difference. So this was a population-based study, that's the important component, and the first randomized trial published of colonoscopy versus usual care for screening. And in this study, there was an invitation to screening colonoscopy, randomized versus no screening or usual care. The endpoints included colorectal cancer incidence and mortality at 10 and 15 years. And the way this came out was with an abstract using an intention to screen analysis with risk of cancer at 10 years being reduced by 18 percent, and the risk of cancer death at 10 years being reduced by 10 percent, but with a wide confidence interval demonstrating no significance compared to the non-screened group. The conclusion was politely worded as the risk of colorectal cancer at 10 years was lower among participants who did undergo screening than among those who did not. Top end sentence. But of course, the media is more savvy than that, and the only story they heard was that this doesn't decrease colorectal cancer mortality at 10 years. And within about two minutes of the official release of this, Bloomberg News came out with their headline, Screening Procedure Fails to Prevent Cancer. And I think within 12 to 18 hours, this Bloomberg message from Bloomberg was in most of our local newspapers, many of our television stations, and certainly network news and morning reports. So there are a number of challenges with this study, and the first is the use of the intention to screen analysis with associated very low uptake. So the intention to screen, or in non-screening studies, we would call it intention to treat, is a very important methodology that we pretty much expect in randomized trials because of its ability to reduce bias. In this case, with the screening study, it's very useful for population analysis and planning on a large scale. Unfortunately, only 42% of patients who were invited to undergo screening actually did do that. If we compare that to a per-protocol analysis, which examines those who actually do get screened, this approach in a large randomized trial such as this does carry some risk of bias and contamination of the enrolled groups who actually elects to go forward versus doesn't feel it's necessary. But this is the analysis that I would say is most valuable for conversation with our individual patients. The long and short of it is there's really no benefit unless you actually do undergo the screening procedure. I think the study was well done. The presentation of it might have been different for our preferences, but one could question whether the media frenzy really, really should have been anticipated. Was that an oversight or was that actually planned by the publisher? I think the publisher takes some credit for what might be damaging our practices going forward. If we do compare the intention to screen and per-protocol analysis, you can see that the reduction in cancer incidence goes from 18% to 31% and the reduction in mortality goes from a non-significant 10% up to as high as 50% in this particular study. There are lower estimates also provided in the article because they used two or three different statistical methodologies to reduce the bias inherent in a per-protocol analysis and whether they were pertinent and important to this study, I can't comment. But in any regard, the per-protocol analysis does provide statistically significant data. A second concern was the relatively early 10-year analysis, and some might call it premature. In an author's discussion of the same trial in 2016, they discussed their intent for a 15-year follow-up, which they will come forward with eventually. We know that progression from polyp to cancer is a relatively slow process. And when one compares this with a more recent pooled sigmoidoscopy study looking at four randomized trials, we can see that the differences between intervention and observation diverge beyond 10 years. The authors did note in conversation just yesterday that, surprising to them, their observed cancer mortality in the non-screened group was quite a bit lower than they expected to find. And in planning the study, their number of enrollees was all planned based on these expectations. So they feel that this may be an underpowered study to show what we all believe, that colonoscopy is useful. So this is the pooled results published two days after the New England Journal study. The same author, first author of the New England Journal study, is a collaborator on this study. This is sigmoidoscopy alone, the only randomized trials that exist, and it shows that overall colorectal cancer mortality starts to diverge only after 10 years. Distal mortality also diverges only after 10 years. In this set of four trials, there was no protective effect for proximal cancer, although some sigmoidoscopy non-randomized trials have suggested proximal benefit on the basis of the subsequent full colonoscopy if distal polyps are found. So if we look closer at that study just published in the last 10 days, their attendance was about 58 to 84%, significantly better. They also did use an intention to treat analysis, but this was at 15 years, and their incidence and mortality were both decreased by 21 and 20%, as you can see. Equivalent or better than the results of this New England Journal study doing a full colonoscopy. So go figure, you just can't bring those two together and understand them in the same light, given one's a full exam and one's a half exam. So there were also questions about polyp detection in this study, and frankly the authors took a little bit of offense at the raising of these issues. The ADR rates, however, included 28% of endoscopists with ADRs below 25%. Now they may have been low-volume endoscopists and not contributed significantly to the study itself. Overall, their ADR rate was satisfactory at 30%, and so one wonders about the spread amongst the endoscopists and was there disparate performance among them. A second point they made is that although Sweden's contribution to the study was quite small, about 10%, their polyp prevalence is really quite low, and so a much lower ADR was found and would be expected to be found in that smaller contribution to the study. Secondly, CEQL intubation rates. About 17% of endoscopists had CEQL intubation rates below 95%. Again, overall, per country, CEQL intubation rates were satisfactory by current modern standards, but not apparently related to no sedation or low sedation. There was a wide divergence in countries using sedation or not, but that didn't influence their polyp detection rate, and apparently not correlated with patient tolerance, but that's a tough thing. How well can you examine the right colon in the midst of a squirming patient? I know in our country that's challenging. Well, what has the U.S. response been? On the following day, the 10th, the ASGE released an e-blast to members, which hopefully you all received and became aware of so that you could start conversations with patients. We've heard from many members that patients were calling in within 24 hours saying, I don't think I want to come in for that colonoscopy, or do you really think I need that colonoscopy? New England Journal says, maybe I don't. So this is becoming an issue for all of us. All relevant organizations, not only the professional societies, but the American Cancer Society and multiple others have made statements cautioning against change in our current practices regarding colorectal cancer risk and whether we should be screening. There have been many media interviews by local cancer and endoscopy experts around the country and by others representing these other organizations, and the media have stepped back. I think it took 48 hours for that to happen, but National Public Radio, even Fox and CNN have softened their message and started to have longer stories looking at the underside of this whole article. So they're now questioning the data vis-a-vis what we already know and what our standard of practice is in North America. So we did have this webinar that I mentioned with the authors yesterday morning. It's now available, and I encourage your viewing of it. And it will certainly fall on all of us to have ongoing education with families and patients to avoid slippage in our current rates. And this is what the site looks like. If you go to the ASGE website, it'll be readily available to you. For providers, well, there are a lot of issues about this study. We've talked about one that we didn't mention is the dated entry. 28 to 13 years ago in our world in North America, that's a long, long time. And the quality of our exams, our sedation, our cleansing is all remarkably improved, I would wager, since 2010. There are issues with follow-up, low uptake, disparate quality parameters, and prevalence of disease. So a lot of concerns. The per-protocol analysis, I think, is the biggest concern relative to talking to our individual patients. Prior studies do suggest that this underestimates the benefits, and hopefully their 15-year analysis will demonstrate more than we've seen so far for benefit. We all believe that colonoscopy remains the most accurate means of detecting and treating colorectal neoplasia. And so this should not change our advised and commonly embraced approach of sequential offerings for screening, beginning with colonoscopy and then the other screening tests, which are all legitimate and useful. And the best screening test, of course, is the one that the patient will actually accept or tolerate. So thank you very much.

Brett Peterson

ASGE

Nordic trial

screening colonoscopy

colorectal cancer