June 14, 2012 — Controversy is still brewing over optimal strategies for colorectal cancer screening, according to a letter to the editor in the New England Journal of Medicine regarding an editorial, and theeditorialists' response, both published in the May 31 issue.
Worldwide, colorectal cancer is the third most common cancer, with a lifetime risk of approximately 5% in the United States. Colonoscopy and fecal immunochemical testing (FIT) are both accepted strategies for colorectal cancer screening in the average-risk population, but issues regarding these tests include sensitivity and specificity for detecting adenomas and cancers, compliance with screening, and cost.
"Further research with FIT and new blood and fecal DNA tests is being done, and more data on sensitivity for advanced adenomas, cancer, and acceptance of testing over multiple rounds will be forthcoming," James E. Allison, MD, clinical professor of medicine emeritus at the University of California San Francisco and San Francisco General Hospital in the Division of Gastroenterology, and adjunct investigator in the Kaiser Division of Research, told Medscape Medical News in an email.
"The FIT data are already very compelling and led to a recent change in the American Cancer Society's recommendation for using [this test]," Dr. Allison said. "The idea that only structural exams find advanced adenomas is being put to rest in the United States, as is evident from recent statements from the National Cancer Institute and the American College of Physicians."
As previously reported by Medscape Medical News, 2 studies published February 23 in the New England Journal of Medicine addressed some of the issues regarding colorectal cancer screening.
The first study was a retrospective analysis from the National Polyp Study (NPS) by Ann G. Zauber, PhD, and colleagues, showing that colonoscopic removal of adenomatous polyps was associated with a 53% reduction in colorectal cancer mortality during a mean follow-up of 15.8 years.
The second study was a large, randomized controlled trial by Enrique Quintero, MD, PhD, and colleagues from the COLONPREV Study, showing that real-world screening rates remain low. Among participants aged 50 to 69 years, only 24.6% accepted once-only screening with colonoscopy. About one third (34.2%) agreed to the first of 5 biennial screenings with FIT, and the investigators and editorialists anticipated that compliance with FIT would decrease over time.
Although colonoscopy and FIT were equally effective for detecting colorectal cancer, colonoscopy had significantly better diagnostic yield for adenomas, which are a strong predictor of cancer risk. Advanced adenomas were detected in 1.9% of the colonoscopy group and in 0.9% of the FIT group (odds ratio, 2.30; P < .001). Nonadvanced adenomas were detected in 4.2% of the colonoscopy group and in 0.4% of the FIT group (P < .001).
Aims of Screening
In an accompanying editorial, Michael Bretthauer, MD, PhD, and Mette Kalager, MD, emphasized the importance of removing precursor lesions (adenomas) because clinical symptoms do not develop until late in the course of colorectal cancer.
"Screening tests for colorectal cancer can be broadly divided according to two distinctly different aims; cancer prevention vs early detection of cancer," Dr. Bretthauer told Medscape Medical News in an email. "Preventive screening methods target the disease before it becomes malignant, by detecting and removing precursor lesions. ... Reduced incidence will, as a consequence, result in reduced mortality (fewer people will get the disease, so fewer can die as a result of it)."
In the COLONPREV study, the diagnostic yield for adenomas (which is the target for preventive colorectal cancer screening tests) was much lower for FIT than for colonoscopy, both in the "as-screened" analysis and in the intention-to-treat analysis (which is affected by screening adherence).
"This means that, despite the lower adherence for colonoscopy, colonoscopy is still a much better test for adenoma detection and thereby for cancer prevention," said Dr. Bretthauer, who is professor of medicine at the Institute of Health and Society in the Department of Management and Health Economy at the University of Oslo and in the Department of Transplantation Medicine in Gastroenterology at Oslo University Hospital in Norway.
Regarding the COLONPREV study, the editorialists note that the low yield for adenomas in the FIT group indicates that FIT is not a good test for detecting adenomas. Although the FIT group had fewer complications, they suggested that this finding was likely to change with more rounds of testing.
What Defines the "Best" Screening Test?
In considering which screening test is "best" to detect and prevent a condition such as colorectal cancer, considerations include sensitivity to diagnose the condition, specificity or lack of false-positive results, compliance with screening, and complications and costs of screening.
A letter to the editor by Dr. Allison questioned the editorialists' conclusion that FIT is not a good test for detecting adenomas. He noted that results in the COLONPREV study were reported after only 1 screening round and that FIT is recommended as a program of repeated screenings every other year.
Dr. Allison also pointed out that only a small number of all polyps ever become fatal cancers, with annual progression rates of 0.25% to 1% (Allison and Selby 2001; Clark et al 1985; Ransohoff 2002). This transformation usually occurs over a period of years, during which time there would be many opportunities for FIT testing to discover polyps.
"It is true that not all adenomas are progressing into cancer, and only a small minority does," Dr. Bretthauer said. "However, we believe that the vast majority of cancers develop from adenomas, and because we don't know which adenomas will grow into cancer, we need to remove as many as possible. But to twist this around...the majority of large, high-risk adenomas and cancer arise from once small adenomas, [so] detection and removal of adenomas (even the small ones) is important to prevent cancer."
Before the availability of FIT, guaiac fecal occult blood testing (FOBT) used small stool samples to detect microscopic bleeding. Compared with colonoscopy, the difference in detecting advanced adenomas and colorectal cancers appears to be much greater with guaiac FOBTs than with FIT, according to Dr. Allison. He added that FIT is actually more sensitive than anticipated in detecting advanced adenomas.
Because the sensitivity of FIT for detecting advanced adenomas is a little less than one third of its sensitivity for detecting cancer, a significant number of advanced adenomas may be missed on first-round screening and detected only in the second round.
"As several very large randomized trials have shown, fecal testing is a reasonably good cancer detection test and reduces cancer death, but not incidence, because this needs adenoma detection," Dr. Bretthauer said.
In a response to Dr. Allison's letter, Dr. Bretthauer and Dr. Kalager suggest that the low yield for adenomas in the first FIT round reduces the potential value of FIT screening for adenoma detection, particularly because the time needed for progression of an adenoma into cancer is unknown.
"One round of FIT may be an odd comparator for colonoscopy," Dr. Bretthauer said. "Although this is true, we do not believe [COLONPREV's] results make a strong case for FIT as an adenoma test, because one may expect a decline in compliance with repetitive FIT rounds, as shown in several studies; and with increasing screening rounds there will be an increase in complications and adverse events (due to colonoscopy follow-up of FIT tests). Adenoma bleeding is highly variable over time and only very small amounts of feces are collected, which makes it unlikely that blood from adenomas will be detected by taking one tiny FIT stool sample every other year."
Although adenoma detection could be improved by using more than 1 stool sample, shorter screening intervals, and lower cutoff levels for test positivity, these strategies would most likely increase costs and reduce compliance, Dr. Bretthauer said.
Dr. Allison cited an Italian study and editorial published in the June issue of Clinical Gastroenterology and Hepatology (2012;10:570-572, 633-638) regarding the outcomes of 4 rounds of biennial screening with a quantitative FIT. Even though compliance was far from perfect, hardly any interval carcinomas were observed, and detection of advanced adenomas remained high over the 4 rounds of screening.
Advanced adenomas are removed during colonoscopy after a positive FIT result, and they therefore can no longer progress to colorectal cancer. In addition to detecting colorectal cancer at an earlier stage, FIT screening can help prevent colorectal cancer via removal of advanced adenomas, the editorial notes.
Screening Compliance
"It is interesting that the editorialists criticize a FIT-based program for unsubstantiated poor compliance, when there are no data I am aware of showing as good or better compliance to every-10-year colonoscopy," Dr. Allison said. "Given the very poor acceptance of initial screening colonoscopy by patients offered colonoscopy in the Spanish study, it would be unlikely that colonoscopy screening compliance would be as good."
He added that although few studies have addressed compliance over multiple rounds of FOBT, there appears to be a relatively small dropout rate from the original cohort of those who accepted screening, with about 85% of the original cohort continuing over 3 rounds of biennial testing.
"Low adherence for both tests, and for colonoscopy in particular, has been a problem in the [COLONPREV] study," Dr. Bretthauer said. "However, adherence is modifiable, as many studies have shown, by behavior and preferences research and by information about effectiveness of the tool and the significance of the disease. Adherence is a more modifiable barrier than test characteristics, which are basically fixed."
Screening Compliance
"It is interesting that the editorialists criticize a FIT-based program for unsubstantiated poor compliance, when there are no data I am aware of showing as good or better compliance to every-10-year colonoscopy," Dr. Allison said. "Given the very poor acceptance of initial screening colonoscopy by patients offered colonoscopy in the Spanish study, it would be unlikely that colonoscopy screening compliance would be as good."
He added that although few studies have addressed compliance over multiple rounds of FOBT, there appears to be a relatively small dropout rate from the original cohort of those who accepted screening, with about 85% of the original cohort continuing over 3 rounds of biennial testing.
"Low adherence for both tests, and for colonoscopy in particular, has been a problem in the [COLONPREV] study," Dr. Bretthauer said. "However, adherence is modifiable, as many studies have shown, by behavior and preferences research and by information about effectiveness of the tool and the significance of the disease. Adherence is a more modifiable barrier than test characteristics, which are basically fixed."
Clinical Implications and Costs
In their editorial, Dr. Bretthauer and Dr. Kalager suggested a strategy of using colonoscopy as a triage screening, to be performed once for everyone at age 60 years. Those with no adenomas detected would be classified as low risk of developing colorectal cancer and would not need further screening, whereas those with advanced or other adenomas would be classified as high risk and undergo strict surveillance.
"Colonoscopy programs should be 'cost-effective,' which implies that we have established a good estimate for effectiveness of screening in reducing CRC [colorectal cancer] incidence or death, and that costs are reasonable," Dr. Bretthauer said. "Each society must discuss what in the context of screening or healthcare in general is 'cost-effective.' Many countries have adopted the UK NICE [United Kingdom National Institute for Health and Clinical Excellence] threshold of about 40,000 to 50,000 Euro for one quality-adjusted life-year saved as cost-effective, which may be a reference for a colonoscopy screening program."
However, a still-unanswered question is how a population-based colonoscopy screening program would be funded.
"The costs, monetary and human, are huge, and such a program is unlikely to be adopted in any country with limited healthcare resources," Dr. Allison said. "Since as of today that even includes the United States, it seems unlikely to happen."
Dr. Allison pointed out that some US commercial insurance plans spend more every year on colonoscopies than on cardiac bypass and hip and knee surgeries combined, and that gastroenterologists spend as much as 50% of their practice time doing colonoscopy.
"One can only wonder how well patients with serious gastrointestinal diseases are being cared for, or how easy it is for them to get a timely appointment with their gastroenterologist," he added.
The editorial by Dr. Bretthauer and Dr. Kalager noted that the NPS highlights the importance of long-term surveillance for patients after the initial removal of adenomas. With strict surveillance of these patients, colorectal cancer mortality was similarly low in the adenoma cohort and in a concurrent nonadenoma cohort during the first 10 years of follow-up. However, the risk subsequently increased in the adenoma cohort, when surveillance was not organized by the investigators.
According to Dr. Allison, the assertion in the editorial that FIT "is not a good one for detecting adenomas," may have unforeseen implications because most adenomas do not directly lead to death from colon cancer.
"What is sad and possibly not apparent to non-US citizens is that such statements influence US primary care physicians and nongastroenterologist subspecialists, the US print and television media, and the general patient population in the US," Dr. Allison said. "As was shown clearly after publication of their editorial, these groups take this to mean FIT is not a good screening test for colorectal cancer. This has a huge impact on our uninsured/underserved population who can't afford colonoscopy and understandably don't want to be offered a 'second best' test.
"It certainly doesn't help our CDC [Centers for Disease Control and Prevention] and AGA [American Gastroenterological Association] supported program at San Francisco General Hospital, where we are trying to increase CRC screening rates of the uninsured underserved patient population," Dr. Allison concluded. "Can you imagine how they would feel if they thought they were getting the 'not best' test?"
Dr. Bretthauer's group has received research funding from Olympus, Ferring, and Fujinon. Dr. Allison has disclosed no relevant financial relationships.
Comments
Post a Comment