The risk for colorectal cancer (CRC) in young adults is actually lower than has been estimated, because previous studies did not differentiate between colorectal adenocarcinoma and the histologically different carcinoid tumors, which are incidental findings, say experts.
New estimates for the risk of CRC in young adults, which differentiate colorectal adenocarcinoma from other types, are reported in a study published December 15 in the Annals of Internal Medicine.
They are important because this finding has implications for CRC screening, say a trio of experts in an accompanying editorial.
Reports of an increase in the incidence of CRC in younger adults have led to changes in screening for this cancer in the United States. The age for starting CRC screening has been lowered to 45 years (instead of 50 years) in recommendations issued in 2018 by the American Cancer Society (ACS), and also more recently in preliminary recommendations from the United States Preventive Services Task Force (USPSTF).
However, that 2018 ACS recommendation to lower the starting age to 45 was based to a large extent on a report of a higher incidence of CRC in younger adults from a 2017 study that used the SEER (Surveillance, Epidemiology, and End Results) database. (CA Cancer J Clin. 2017;67:177-193.)
But that SEER-based study considered “colorectal cancer” as a homogeneous group defined by topology, the editorialists point out.
The new study, say the editorialists, uses that same SEER database but has ‘disentagle[d] colorectal adenocarcinoma, the target for screening, from other histologic CRC types, including neuroendocrine (carcinoid) tumors, for which screening is not recommended.”
The study authors explain that adenocarcinoma is a target for prevention through screening because it arises from precancerous polyps. Those growths can be detected and removed before cancer develops. That doesn’t apply to carcinoid tumors, which are frequently incidental findings on flexible sigmoidoscopy or colonoscopy, they say.
These carcinoid tumors typically are indolent, with a better prognosis than most other cancer types, the editorialists add. “Most likely, the majority of carcinoid tumors identified by screening represent incidental findings with little
health benefit from detection. In fact, many may be characterized as overdiagnosed tumors, which by definition increase the burden and harms of screening without the balance of additional benefit.”
This new analysis shows that 4% to 20% of the lesions previously described as CRC were not adenocarcinoma but carcinoid tumors, the editorialists point out.
This figure rises even higher in the subgroup of findings pertaining to the rectum, the colonic segment with the largest reported increase in early-onset CRC. Here, up to 34% of lesions (depending on patient age) were carcinoid tumors rather than adenocarcinoma, they note.
The three editorialists — Michael Bretthauer, MD, PhD, and Mette Kalager, MD, PhD, University of Oslo in Norway, and David Weinberg, MD, MSc, Fox Chase Cancer Center, Philadelphia — call for action based on the new findings.
“The ACS’s 2018 estimate of about 7000 new CRC cases among persons aged 45 to 49 years in the United States (the justification for screening) needs to be adjusted downward on the basis of the new evidence,” the trio write.
They conclude that “caution is warranted when promoting the benefits of CRC screening for persons younger than 50 years.”
However, the senior author of the new study, Jordan Karlitz, MD, Tulane University School of Medicine, New Orleans, Louisiana, strongly disagrees.
Contrary to the editorialists, Karlitz told Medscape Medical News that he and his colleagues firmly believe that colorectal cancer screening for average-risk patients should begin at age 45 and that their new research, despite its clarification about carcinoid tumors, provides evidence for that.
“There are a number of other studies that support screening at age 45 as well,” he said. “This [new] finding supports the presence of a large preclinical colorectal cancer case burden in patients in their 40s that is ultimately uncovered with screening initiation at age 50. Many of these cancers could be prevented or diagnosed at an earlier stage with screening at age 45.”
“This is the first study to analyze early-onset colorectal cancer by specific histologic subtype,” Karlitz also pointed out.
“Although colorectal carcinoids are increasing at a faster rate than adenocarcinomas, adenocarcinomas constitute the overwhelming majority of colorectal cancers in people in their 40s and are also steadily increasing, which has implications for beginning screening at age 45,” he said.
Adenocarcinomas also make up the “overwhelming majority” of colorectal cancers in patients under 50 overall and “are the main driving force behind the increased colorectal cancer burden we are seeing in young patients,” Karlitz added.
Furthermore, “modeling studies on which the USPSTF screening recommendations were based [which recommended starting at age 45] were confined to adenocarcinoma, thus excluding carcinoids from their analysis,” he said.
Steepest Changes in Adenocarcinomas in Younger Groups
In their study, Karlitz and colleagues assessed the incidence rates of early colorectal cancer, using SEER data from 2000 to 2016, and stratifying the data by histologic subtype (primarily adenocarcinoma and carcinoid tumors), age group (20-29, 30-39, 40-49, and 50-54 years), and subsite.
A total of 123,143 CRC cases were identified in 119,624 patients between the ages of 20-54 years during that time period.
The absolute incidence rate in the younger age groups (20-29 and 30-39 years) were very low compared with those aged 40-49 and 50-54 years.
The greatest 3-year average annual incident rates changes in adenocarcinoma (2000 to 2002 vs 2014 to 2016) for any age group or subsite were for rectal-only cases in the 20-29 years group (+39%), as well as rectal-only cases in those aged 30-39 years (+39%), and colon-only cases in the age 30-39 group (+20%).
There was also significant increase in rectal-only adenocarcinoma in individuals aged 50-54 years (+10%). A statistically significant increase in the annual percentage change for adenocarcinomas was observed for all age groups, except for colon-only cases in the 20-29 years group (0.7%) and for both colorectal (0.2%) and colon-only cases (–0.1%) in those aged 50 to 54 years.
Even though the absolute carcinoid tumor incidence rates were lower than for adenocarcinoma in all age groups and subsites, a statistically significant increase was observed in the 3-year average annual incidence rate of combined-site colorectal carcinoid tumors in all age groups from 2000–2002 and 2014–2016. This increase was largely the result of increases in rectal carcinoid tumors, the authors note.
The authors also highlighted the results in the 40- to 49-year age group “because of differing opinions on whether to begin average-risk screening at age 45 or 50 years.”
They report that rates of rectal and colon adenocarcinoma are increasing “substantially,” whether measured by changes in 3-year average annual incidence rate or by annual percentage changes. The change in average annual incidence rate of colon-only adenocarcinoma for persons aged 40-49 years was 13% (12.21 to 13.85 per 100,000), and that of rectal adenocarcinoma was 16% (7.50 to 8.72 per 100,000). Corresponding annual percentage changes were 0.8% and 1.2%, respectively. “These significant increases in adenocarcinoma incident rates add to the debate over earlier screening at age 45 years,” they comment.
Calls for Next Steps
The editorialists emphasize restraint when promoting the benefits of colorectal screening for persons younger than 50 years.
They point out that the USPSTF released a provisional update of its CRC screening recommendations about lowering the age to initiate screening to 45 years, as reported by Medscape Medical News.
“No new empirical evidence has been found since the USPSTF update in 2016 to inform the effectiveness of screening in persons younger than 50 years,” they write, adding that similar to the American Cancer Society in 2018, the task force has relied exclusively on modeling studies.
This new data from Karlitz and colleagues “should prompt the modelers to recalculate their estimates of benefits and harms of screening,” they suggest. “Revisiting the model would also allow competing forms of CRC screening to be compared in light of new risk assumptions.
“Previous assumptions that screening tests are equally effective in younger and older patients and that screening adherence will approach 100% may also be reconsidered,” the editorialist comment.
The study authors conclude somewhat differently.
“In conclusion, adenocarcinoma rates increased in many early-onset subgroups but showed no significant increase in others, including colon-only cases in persons aged 20-29 and 50-54 years,” the investigators write.
They also observe that “rectal carcinoid tumors are increasing in young patients and may have a substantial impact on overall CRC incident rates.”
Those findings on rectal carcinoid tumors “underscore the importance of assessing histologic CRC subtypes independently,” the researchers say.
This new approach, of which the current study is a first effort, “may lead to a better understanding of the drivers of temporal changes in overall CRC incidence and a more accurate measurement of the outcomes of adenocarcinoma risk reduction efforts, and can guide future research.”
The study had no outside funding. Karlitz reports personal fees from Exact Sciences, personal fees from Myriad Genetics, and other fees from Gastro Girl and GI OnDEMAND, outside the submitted work. Bretthauer reports grants from Norwegian Research Council, grants from Norwegian Cancer Society for research in colorectal cancer screening. Weinberg and Kalager have disclosed no relevant financial relationships.