Challenges around diagnosis of early onset colorectal cancer, and the case for screening

Introduction

Colorectal cancer is the third most diagnosed cancer worldwide with nearly 2 million cases diagnosed in 2020, and the second most common cause of cancer death resulting in over 900 000 deaths in 2020.¹ Over the past few decades, the incidence of early onset colorectal cancer (EOCRC) defined as CRC diagnosed under the age of 50 has been rising across many parts of the world, with the trend being independent of those seen in overall incidence of CRC, occurring across countries where overall CRC incidence is stable or declining.^{2, 3} It is estimated that 13% of CRC cases diagnosed in USA in 2023 will be in those younger than 50 years,⁴ with other predictions estimating by 2030 10% of colon and 25% of rectal cancers will be in those under 50 years.⁵ Causes for this increase are unknown however most cancers are sporadic and as such likely driven by the interaction of an individual's gut microbiota in combination with lifestyle and/or environmental factors.^{6, 7}

Patients diagnosed with CRC under the age of 50 tend to have a higher proportion of left sided and rectal cancers, and more commonly present with later Stage 3 or 4 disease.⁷ Recent studies report the proportion of EOCRC patients who present with stage 4 disease is over 36%.⁸ Research has shown that younger patients are more likely than older patients to experience delays to diagnosis.⁹ Regardless of the cause and until methods of prevention can be developed, improved detection should be a major focus to reduce the impact of this rising problem, with the largest predictor of prognosis currently being stage at diagnosis.^{7, 10}

Earlier detection of cancers via screening programmes is proven to be an effective way to improve outcomes in bowel cancer,^{11, 12} but has historically been reserved for those over the age of 50¹³ There is however increasing evidence of its efficacy in those aged under 50 years and, as discussed in this review, screening may be the most effective way to improve outcomes in those with EOCRC. Moreover, it is crucial that extension of any screening programmes is done in a way that ensures equitable access and outcomes for all.

Limitations of focusing on work up of symptomatic patients

The triage of patients with symptoms suggestive of CRC presents a logistical challenge. Symptoms such as rectal bleeding, abdominal pain, and altered bowel habit are non-specific, and very common in the general community with the vast majority due to benign pathology especially in younger patients. As such, the significance of these symptoms in patients with colorectal cancers are easily overlooked. There are also many barriers to patients in accessing primary health, with these being most pronounced in certain populations including younger patients,¹⁴ as well as certain ethnic and lower socioeconomic status groups.¹⁵ There is also evidence these populations are less likely to be referred on for investigation or assessment by specialist services,¹⁶ driving inequities in the diagnosis of bowel cancer.

For those who are referred for investigation, colonoscopy remains the gold standard procedure used to diagnose neoplastic lesions. However, the availability of endoscopy time in most health systems is limited, necessitating triaging of referrals. Recent research from the USA assessing the diagnosis of over 5000 cases of EOCRC found that the presence of one or more ‘red flag’ symptoms (abdominal pain, rectal bleeding, diarrhoea and iron deficiency anaemia) were associated with increasing risk of CRC and that in their cohort, 68.6% of patients had presented with one or more of these red flag symptoms between 3 months and 2 years prior to diagnosis.¹⁷ This is evidence that early recognition may indeed aid timely diagnosis of EOCRC.¹⁷ The non-specific and common nature of these symptoms however presents an enormous logistical challenge in terms of health systems being able to offer all patients who present with these red flags a colonoscopy. A recent study calculated, based on the likely prevalence of these symptoms in the general American population, and the prevalence of EOCRC that 59 856 colonoscopies would be required to diagnose just two cases of EOCRC.¹⁸

Timely workup and diagnosis of these symptomatic patients is clearly important, given prognosis depends on stage of disease, and substantial delays are likely to be associated with later stage. This is particularly likely to occur in EOCRC given evidence that diagnostic intervals (time from symptom onset to diagnosis) are longer in younger patients. One study reported that the median time from symptom onset to starting treatment in those under 50 years of age was 217 days compared to 29.5 days in over 50 years of age.⁹ This delay however, was not associated with worse stage or survival, consistent with systematic reviews of over 120 primary studies which found no significant association between diagnostic and treatment intervals and outcomes.^{19, 20} Variability between stage at presentation may also be a reflection that tumour biology rather than diagnostic interval has a greater role in the prognosis of EOCRC.¹⁹ Therefore improving diagnostic intervals will only lead to an improvement in outcome in a minority of patients.

There is also a significant psychological cost of perceived delays in the diagnostic interval. While it is impossible to know the true impact of any delay in diagnosis on any individual case, trying to come to terms with the ‘what if’ in those patients now facing a terminal diagnosis after a delayed diagnostic interval is undeniably something which will be difficult to come to terms with. Studies have indeed shown that perceived delays to cancer diagnosis are associated with significantly increased levels of distress, and decreased quality of life, with this effect being greatest in younger patients.²¹

Screening for early onset colorectal cancer

Bowel cancers diagnosed by screening prior to symptoms developing have far better outcomes, with studies showing a 74% lower CRC specific mortality compared to symptom-detected cancers.¹¹ There are several different methods available for bowel cancer screening,²² with the most widely used currently the faecal immunochemical test (FIT).²³ This test, which has largely replaced the less sensitive faecal occult blood test,¹⁶ and measures the quantity of haemoglobin present in the stool. The evidence for the effectiveness of this test as a screening tool is very well established.²⁴ Moreover, the FIT test has the advantage that a threshold that determines a positive test can be set. As such, while sensitivity may decrease as the threshold is raised, specificity increases making it an ideal test for screening.²⁵

Most countries that have implemented bowel cancer screening programmes do not include patients under the age of 50.²³ However, in countries which are screening those below age 50, most are not seeing an increase in EOCRC (Table 1). A review of international trend in EOCRC incidence rates across five continents found only three countries where the incidence rates were decreasing.²⁶ Two out of these three countries (Italy and Austria) have been screening patients in their 40's since the 1980's.²³ Austria, which has been screening from the age of 40 since 1980, has seen overall decreasing incidence of EOCRC in recent years (2003–2016). Interestingly, while they are not seeing the rise in EOCRC incidence, Austria is still seeing an increase in CRC rates in those under the age of 40, which remains outside their screening program.²⁷ Italy and Japan have been screening for bowel cancer below the age of 50 since 1982 and 1992, respectively. While the incidence of EOCRC is falling in Italy, it has remained stable in Japan.²⁶ This is in contrast to China, which has also been screening from the age of 40 for a significant period of time²³ but also reporting that EOCRC incidence is increasing.²⁸ This may in part reflect screening in some but not all provinces coupled with the impact of rapid urbanization and westernization of diet in China over the past four decades and the associated increase in CRC risk may have overshadowed any benefit gained from screening (Table 1).²⁹

As an individual's risk of CRC increases with age, the efficacy of screening with FIT decreases with age due to reducing prevalence resulting in lower positive predictive value, therefore necessitating larger numbers of screening tests to be done per cancer found.³⁰ Previously, due to the low incidence of EOCRC, it was not cost effective to screen those aged under 50 years. However, due to the dramatically rising incidence of EOCRC studies have now shown that screening is cost effective down to at least the age of 45,³¹ and possibly 40 years.³² Modelling from Canada found that screening from age 40 resulted in 18 135 fewer CRC cases, 7988 fewer CRC deaths, and 150 373 Quality adjusted life years (QALYs) compared to age 50. This was over a 40 year period at a cost of $CAD 2622 per QALY.³² To put this number into perspective, a report commissioned for the Canadian government in 2018 determined the threshold for a health intervention to be considered cost effective was over $28 000 per QALY gained, over 10 times more than the estimated cost of screening from age 40.³³ This is a somewhat conservative estimate with other studies quoting a cost effectiveness threshold in Canada of over $44 000 per QALY.³⁴ The assumed costs calculated in their modelling were comparable to estimates used in a cost-efficiency analysis of the New Zealand bowel cancer screening pilot,³⁵ and therefore is likely to be applicable to screening in New Zealand. Likewise, modelling performed by the Cancer Council Australia, indicated that lowering bowel cancer screening to 40 years would ‘increase the health benefits of screening and cause limited increases to the costs, resource demand, and potential harms of screening’ and found it was cost effective from 40 years.³⁶ These figures are aside from the massive human benefits of early cancer detection.

These changing trends are being reflected in several major organizations in the USA (the National comprehensive cancer network (NCCN),³⁷ US Multi Society Task Force on Colorectal Cancer,³⁰ and US Preventative Services Task force)²⁵ now recommending screening start from age 45. American health insurance companies are now required by law to fund colorectal cancer screening in their patients from age 45.³⁸ There have been similar calls to lower the screening age to 45 across Europe,³⁹ but as yet, no countries have followed Italy and Austria in screening below the age of 50. Australia has also recently started population screening from the age of 45 in July 2024, following the recommendation of updated guidelines, endorsed by the National Health and Medical Research Council (NHMRC),³⁶ with publicly funded screening available to those from age 40 if they request it from their family doctor.³⁶ It is too early to see the effects of these changes in the USA and Australia will impact on EOCRC incidence but this will provide good insights in the future.

Recently published data from New Zealand found that in the 5 years leading up to 2020, 71% of all EOCRC were diagnosed in those from 40 to 49 years, with incidence rates dropping dramatically under the age of 40.³ Therefore, we believe starting screening from age 40 would improve outcomes in the vast majority of these EOCRC cases, while remaining cost-effective.

Considerations around ethnicity and equity

Screening has the potential to worsen inequities with evidence that there is uneven access to screening programmes in many minority groups.^40-44 It is therefore vital that screening programmes are designed to address this. Despite having lower incidence rates of CRC, Māori in New Zealand are disproportionately affected by EOCRC with 18% of all CRC in Māori occurring under the age of 50 years, compared to 8.5% in the general population.³ This finding is considered largely due to their different age structure and to the growing awareness that Māori, once diagnosed with CRC, are more likely to die from their disease than non-Māori.⁴⁵ To address this, the current New Zealand colorectal cancer screening programme (which starts screening non-Māori from age 60) has now started screening Māori from 10 years earlier.⁴⁶ While this is a positive first step, it will also be vital to identify any inherent systemic barriers that might limit young Māori with regards accessing bowel cancer screening, as further exemplified within Aboriginal and Torres Strait Islander communities.⁴⁷ Research indicates that a tailored approach to increasing engagement is effective,^{40-42, 46, 48} suggesting this approach could be more widely considered to ensure similar equity for those under 60 who don't identify as Māori. It is not clear yet however if current screening methods should be extended below the age of 40 years in any group due to the drastically lower prevalence of disease in this age group. As such, the risk of harm from screening could outweigh the benefits.

Screening participation in younger people

Getting adequate participation in any screening programme will likely be an issue that is particularly relevant to young people. This is exemplified a large Danish study, which found that younger patients were least likely to participate in colorectal cancer screening.⁴⁹ This is consistent with a systematic review assessing factors associated with colorectal cancer screening participation,⁵⁰ and is likely to be even more pronounced in those under the age of 50, where convincing busy, healthy young patients to partake in a screening programme may be difficult, particularly one which involves collection of a stool sample. There is also the potential for the ‘healthy user effect’ where those respond to the request to be screened are less likely to have underlying disease, consuming resources that may be better spent in the work up of older higher risk or symptomatic patients.⁵¹ As such, advertising support for screening programmes will need to be carefully targeted to different age groups.

Biomarkers

The current specificity and sensitivity with of FIT testing means large numbers of patients require investigation for each cancer found.³¹ As such, the measurement of biomarkers that have the potential to improve the accuracy of the FIT currently used to screen for evidence of haemoglobin in stool samples test.^{52, 53} This approach would have the potential to help target limited colonoscopy resources to those patients who are most at risk of having underlying neoplasia, and is behind the international search for biological markers of underlying malignancy before symptoms arise that include markers of inflammation, altered metabolism and/ or changes in an individual's gut micriobiota.⁵³ While these markers are currently not ready to be used for large scale clinical practise, this is a promising area of future research. Moreover, as technology develops, novel biomarkers may even replace the current FIT and this has the potential to increase participation in a screening programme if the test involves the collection of breath or urine as opposed to a stool sample.

Familial risk

Higher risk patients should be the first to be targeted and this means screening those with significant family history of CRC. Family history of CRC in a first degree relative (FDR) is a risk factor for the development of CRC, with the risk being greater in those where the FDR was diagnosed at a younger age.⁵⁴ Guidelines recommend those with a history of a FDR diagnosed under the age of 55 should start screening 10 years prior to the age at which their relative was diagnosed.⁵⁵ A retrospective study however of 2473 EOCRC cases found one in four people diagnosed with EOCRC met criteria for early screening centred on family history.⁵⁶ Of these, 98.4% would have been diagnosed earlier or prevented altogether if screening with colonoscopy had been undertaken.⁵⁷ This suggests that improving accurate assessment of familial risk in patients and appropriate screening has the potential to help reduce the impact of rising EOCRC prevalence. It is worth noting that there are some disparities in both access to primary care,^{14, 15, 58} and the accuracy of family history with certain demographic factors such as ethnicity, socioeconomic status and education level being associated with inaccurate reporting or unknown family history.⁵⁹ Therefore improvements in diagnosis in patients with high familial risk has the potential to worsen inequity in these groups. In the future, we may see advances in polygenic testing to a point where testing of individuals for genetic variations which confer CRC risk overtakes the reliability of family history to stratify those at risk of CRC for screening but further research here is needed.⁶⁰ However, it is also likely that risk associated with a subset of familial EOCRC may reflect intra-familial lifestyle factors and this should also be considered.

Conclusion

Early onset colorectal cancer is a growing problem in many countries around the world and this is unlikely to change until younger patients are included in colorectal cancer screening programmes. While reducing delays to diagnosis in young symptomatic patients will improve outcomes for some, a disproportionately high percentage of patients already have stage 4 disease. Accordingly, diagnosis of cancers prior to symptoms with bowel cancer screening from age 40 would improve outcomes for the majority of EOCRC patients, while remaining cost effective. For this approach to be successful, it will be vital to ensure that this equates to equitable access and outcomes in minority and indigenous populations. The emerging use of novel biomarkers that improve adherence to any screening programme also needs to be considered, as does the accuracy of the screening test(s) in order to reduce the numbers of those requiring investigation to only those at highest risk of having underlying malignancy.

Resources will remain a significant issue, and health systems need to balance increasing roll out of screening without compromising the workup of higher risk symptomatic and older patients. This however should not stop goals from being set. Health systems need to prioritize allocating resources to increase their capacity to extend colorectal cancer screening. The costs of which will be largely balanced by savings as a result of CRC prevention and earlier CRC diagnosis. The incidence of EOCRC is increasing, and screening from the age of 40 will significantly improve outcomes in the majority of patients affected.

Author contributions

Oliver Waddell: Conceptualization; data curation; investigation; methodology; writing – original draft. Jacqueline Keenan: Conceptualization; project administration; writing – original draft. Frank Frizelle: Conceptualization; supervision; writing – original draft.

Conflict of interest

Professor Frank Frizelle is a medical advisor for Bowel Cancer New Zealand. Otherwise no conflict of interests to declare.