The Increase of Early-Onset Colorectal Cancer: New Insights and Emerging Hypotheses

Abstract

Rising rates of early-onset colorectal cancer have been noted worldwide. This commentary addresses possible causes of this trend and opportunities for intervention.

Rising rates of early-onset colorectal cancer (CRC), also known as young-onset colorectal cancer (YO-CRC), have become a growing concern for the public health community worldwide. ¹ Since the American Cancer Society (ACS) initially recommended screening for CRC in the 1970s, its age-adjusted annual incidence rate in the US has declined by 49%, from a peak of 66.2 cases per 100,000 in 1985 to 33.5 per 100,000 in 2022. ² This decrease, predominantly in populations over 50 years, has been ascribed to more widespread implementation of colonoscopy-based screening, which allows for the removal of precancerous polyps. ³ Indeed, by 2023, 72.6% of adults aged 50-75 years in the United States (U.S.) had received CRC screening – a sharp increase from the 38.2% of adults screened in 2000. ⁴

However, a distinct and worrying trend has emerged over the past 3 - 4 decades: a dramatic rise in CRC incidence in those below age 50 years. From a nadir of 4.5 cases per 100,000 in 1987, the incidence rate in this population has more than doubled, reaching 9.4 per 100,000 in 2022. Younger adults are included in this trend; among those in the 15-39 age range, the incidence rate rose 47.5% in the same timeframe, with an annual percent change (APC) of 2% per year. ²

CRC’s rise to prominence in the young, coupled with its decline in older populations, has resulted in a substantial shift in the CRC patient population. Of the 147,931 new colorectal cancers diagnosed in the United States in 2022, 20,422 cases (13.8%), were in patients younger than 50 years. ⁵ By 2030, YO-CRC is projected to account for 11% of colon cancers and 23% of rectal cancers worldwide, affecting both men and women. While approximately 20% of cases are linked to hereditary cancer syndromes, most arise without an identifiable monogenic predisposition. ⁶ Though updated ACS guidelines have moved the starting age for screening from 50 to 45 years, many of the patients with YO-CRC would still not be screened as per these updated guidelines, necessitating further research into the possible causes of this phenomenon.

Recent research has helped characterize the trend of YO-CRC on a global scale. An analysis of cancer registry data by Sung et al found that between 2013 and 2017, YO-CRC incidence rates climbed in over half the countries studied, rising most swiftly in New Zealand, Chile, Norway, Puerto Rico, and England. While some countries noted concurrent increases in older-onset CRC, the majority, including many European, North American, Oceanic, Asian, and Latin American countries, found a decreasing or unchanged incidence in older age groups. ⁷ This evidence indicates that rising YO-CRC rates are a global problem, and it suggests that the driving forces behind this trend are present and spreading throughout various geographic regions.

Any exposure proposed to explain this trend would need to fulfill multiple criteria: it must be associated with the development of CRC, globally widespread, and ostensibly only present in young people. ⁸ In the US, certain non-modifiable risk factors were found to be predictors of YO-CRC when compared with late-onset CRC, such as male sex (OR, 1.44; 95% CI, 1.11–1.87), Black or Asian race (OR, 1.73; 95% CI, 1.08–2.65; OR, 2.60; 95% CI, 1.57–4.15, respectively), family history of CRC excluding known heritable CRC syndromes (OR, 2.87; 95% CI, 1.89–4.25), and personal history of inflammatory bowel disease (IBD) (OR, 2.97; 95% CI, 1.16–6.63). ⁹ However, apart from IBD (which is increasing in incidence), none of these risk factors could contribute to the observed temporal trend. While certain modifiable environmental exposures associated with YO-CRC, such as obesity, low levels of physical activity, alcohol use, and cigarette smoking, have been well-described in the literature to date, these exposures are hardly unique to young people, or even to colorectal cancer. ¹⁰ Hypotheses focusing on these factors alone are incomplete.

There is likely considerable overlap and interplay between various contributing factors, especially emerging ones like increasing obesity, Western dietary patterns, and intestinal dysbiosis. Numerous, incompletely explored pathways link obesity to cancer pathogenesis, including derangements in the insulin-like growth factor-I (IGF-I) system, sex hormone biosynthesis, and adipokine production in visceral adipose tissue (VAT). ¹¹ Intestinal microbiome alterations in obesity can both increase the intra-luminal production of toxic metabolites and compound on the chronic inflammation generated by obesity.^10,11 Indeed, one model has estimated that as much as 28-30% of YO-CRC incidence in the US can be explained by obesity. ¹² Western dietary patterns, including meals high in red and processed meats, sugar- and syrup-sweetened beverages, and highly refined grains, have also been implicated in CRC development. ¹⁰ Using data from the Nurses’ Health Study II (NHS II), Wang et al reported that subjects in the fifth quintile of caloric intake from ultraprocessed foods were 45% more likely to develop young-onset colorectal adenomas than those in the first quintile, even after adjusting for possible confounders. ¹³ Western diets can alter the composition of the gut microbiome, including increasing the number of bacteria species that synthesize the DNA-damaging genotoxic compound hydrogen sulfide. ¹ Additives such as emulsifiers, which are common in processed foods, are hypothesized to promote carcinogenesis given their ability to cause colitis in murine models. ¹⁴ Overuse of antibiotics, particularly in early life, may also have long lasting effects on commensal microbiota. ¹

The gut microbiome represents a possible mediator through which many risk factors may cause YO-CRC carcinogenesis. ¹⁴ Stool RNA and metagenomic sequencing analyses performed in a population of Shanghai residents found significant enrichments of Escherichia, Parvimonas, Fusobacterium, Flavonifractor and morganii genera in YO-CRC patients. ¹⁵ Other studies have found that higher concentrations of Bacteroides, Peptostreptococcus, Streptococcus, Prevotella, and Escherichia are associated with elevated risk of developing CRC; and emerging research indicates that Akkermansia and Bacteroides are specifically associated with YO-CRC. ¹⁶ Causative relationships between specific bacteria and CRC development are difficult to establish, and some bacteria may in fact be indicators of a tumor microenvironment rather than contributors. However, colibactin, a mutagen produced by species of E. coli and Enterobacteriaceae, has recently been implicated in CRC development. In a study of CRC genomes across multiple geographic regions, Díaz-Gay et al demonstrated that the mutational signatures SBS88 and ID18 (a single-base substitution and insertion-deletion, respectively), which are known to be caused by colibactin, were significantly more prevalent in younger CRC patients. Furthermore, SBS88 and ID18 were more prevalent in countries with greater overall CRC incidence rates. ¹⁷ Machine learning models have detected and attributed mutations to colibactin in 12% of metastatic colorectal cancers in a large clinical dataset. ¹⁸ The virome and mycobiome, too, are increasingly recognized players in CRC development. Meta-analysis of fecal metagenomic DNA showed that several species of fungi were consistently enriched, such as Aspergillus rambellii, or depleted in patients with CRC. ¹⁹ Another study of fecal metagenomic DNA demonstrated that, compared to controls, CRC cases exhibited greater diversity of intestinal bacteriophages, such as orthobunyavirus. ²⁰ Increased colonic inflammation and mutagen exposure secondary to intestinal dysbiosis represent important areas for further study into the mechanisms by which YO-CRC develops.

The impact of exposures that may increase incidence rates of CRC in older individuals is difficult to assess, as so much of our data comes from Western countries where CRC screening has been widely adopted. However, data from countries outside of North America, Oceania, and Europe, where there is little population screening for CRC, could provide insight into more recent incidence trends in older age groups. In sub-Saharan Africa, for instance, CRC incidence rates have increased by 12% between 2000 and 2020, while lifestyle, diet, and physical activity levels have become more akin to those of Western countries.^21-23 In our previous commentary on this subject, we noted that in Uganda, the annual incidence rate of CRC in male populations below age 50 rose from 1.40 cases per 100,000 in 1995 to 2.39 cases per 100,000 in 2015; and in male populations age 50-74, CRC incidence rose from 31.96 to 43.36 cases per 100,000. ²³ In the same time frame in Germany, CRC incidence in men below age 50 increased from 8.00 to 9.76 per 100,000, but decreased from 187.30 to 157.61 per 100,000 in men age 50-74. ²³ Newer data reiterate this contrast. When examining countries that have well-established national screening programs, such as Canada, Australia, Japan, England, and the Netherlands, many show a positive total percent change (TPC) in YO-CRC incidence rates between 1990 and 2023, but a negative TPC in CRC incidence rates in older populations (age 50-74) (Table 1). Countries where screening is not yet routine at a national level, on the other hand, almost universally show increasing CRC rates in all age groups (Table 1).

Table 1.

Total Percent Change (TPC) in Colorectal Cancer Incidence Rate by Age Group in Countries With and Without Established Population-Based CRC Screening Programs

Countries with established population-based CRC screening programs	Age group (years)	Total percent change in incidence rate from 1990-2023	Countries without CRC screening programs	Age group (years)	Total percent change in incidence rate from 1990-2023
Australia	15-49	0.32	Chile	15-49	1.39
	50-74	−0.34		50-74	0.71
Canada	15-49	0.49	China	15-49	0.23
	50-74	−0.13		50-74	0.16
England	15-49	0.56	Colombia	15-49	1.35
	50-74	−0.07		50-74	0.94
Japan	15-49	0.27	Costa Rica	15-49	2.00
	50-74	0.42		50-74	1.20
Netherlands	15-49	0.04	India	15-49	0.60
	50-74	−0.30		50-74	0.32
United States	15-49	0.47	Uganda	15-49	0.76
	50-74	−0.29		50-74	0.60

Source: https://vizhub.healthdata.org/gbd-results?params=gbd-api-2023-permalink/5c2bf026eb8ca4c81b23495985b08b7a.

Rationale for country selection: The countries on the left were chosen because they have national, population-based screening programs, updated colorectal cancer registries, and a high proportion of the eligible population successfully screened. While the United States does not have a nationalized screening program (no central invitation system, insurance-driven access), it has a high proportion of the eligible population screened and robust cancer incidence data. Countries on the right were chosen because they currently do not have national, population-based screening programs and have a low proportion of the eligible population successfully screened, but they have established, high-quality cancer registries.

While the countries with CRC screening programs and those without differ from one another in a multitude of ways, such as environmental exposures, climate, and population genetics, their contrasting CRC incidence trends support our hypothesis: the exposures driving YO-CRC in Western countries are contributing to a gradual rise that affects both young people and older populations. ²³ Recent trends also point to a cohort effect that may be mediating the observed rise in YO-CRC. The environmental exposures and changing lifestyles believed to be at the root of increasing CRC rates have become much more prevalent in successive generations, in particular in cohorts born after 1950 in the US. ⁷ The implication of this is that, while younger generations may currently outpace older ones in increasing CRC incidence, as they age, older-onset CRC will increase in incidence as well.

Using the Bayesian Age-Period-Cohort Analysis model, the global burden of YO-CRC has been forecasted to increase, with the age-standardized incidence rate (ASIR) projected at 6.46 per 100,000 by 2030. ²⁴ If a similar rise were to occur in the 65+ age group, who continue to account for most new cases of CRC (the US incidence rate as of 2022 is 146.2 per 100,000), we would expect the overall incidence of CRC to increase dramatically. However, a continued rise in CRC screening and polyp removal in this group may drive down the incidence rate by preventing many new potential cancers. ²⁵

We can leverage our evolving understanding of CRC and its risk factors to support the expansion of CRC screening and to direct further investigation into the mechanisms by which CRC develops. The spread of Western diets, obesity, reduced exercise patterns, antibiotic use and related intestinal dysbiosis in new geographic regions may help explain current global CRC incidence trends. These exposures are potential targets for new prevention efforts, and they are opportunities to greatly improve our ability to risk-stratify patients. Risk scores incorporating numerous features directly from the electronic medical record, for instance, might aid clinicians in identifying patients who would benefit the most from early screening, beyond those with heritable cancer syndromes or family histories of early CRC.

While US guidelines have recently been updated to recommend CRC screening for those aged 45–49, future initiatives to reach this younger population should not de-prioritize efforts to screen older individuals, who remain the most high-risk group. ²⁶ In the US, 64% of cancer screening costs in 2021 were due to CRC alone. Although current recommendations are considered cost-saving due to the high price of treating CRC, colonoscopies still accounted for over $23 billion in screening costs annually. ²⁷ In the US, further lowering the screening age while continuing to rely primarily on colonoscopy may inappropriately shift limited resources away from older populations who need them most. However, a recent large-scale cohort study has shown that, compared to screening starting at age 50, fecal immunochemical testing (FIT) starting as early as age 40 correlates with a significant decrease in CRC mortality over a mean follow-up time of >17 years. ²⁸ Therefore, new approaches could involve using stool-based testing like FIT, which is less resource-intensive, specifically for younger individuals given their lower overall incidence, while reserving gold-standard colonoscopy for older patients. Guidelines must balance potential benefits with the cost they impose on health systems, especially those in countries with publicly funded screening programs. Though we may one day succeed in mitigating CRC risk factors at a societal level, the more immediate future of CRC incidence may be determined primarily by how well we can screen for and prevent the disease in our most vulnerable populations.

ORCID iDs

Isaac Lee https://orcid.org/0000-0003-1371-8364

Ezra Bernstein https://orcid.org/0000-0003-1911-1969