Germline Testing in Colorectal Cancer: Is Age-Based Testing Too Restrictive?

A colorectal cancer diagnosis at 62: Is hereditary testing still worthwhile?

Consider a 62-year-old patient newly diagnosed with colorectal cancer (CRC). There is no obvious hereditary cancer syndrome, and the patient does not meet the conventional definition of early-onset CRC. Should germline multigene panel testing still be part of the discussion?

Age has long influenced decisions about hereditary cancer evaluation. Patients diagnosed before 50 years are routinely considered for genetic assessment under guidance discussed in the current literature, while testing in older patients is often more selective.

This approach reflects the higher probability of inherited cancer predisposition among younger patients. However, it raises an important concern: does restricting testing primarily by age leave clinically meaningful hereditary findings unidentified?

A systematic review and meta-analysis published in The Lancet Gastroenterology & Hepatology on 3 September 2026 provides new evidence to inform this question.

The UPTOWN study, led by Alessandro Mannucci and colleagues, found pathogenic or likely pathogenic germline variants in 15.2% of patients with CRC in its primary analysis.

More importantly, the diagnostic yield remained 13.0% among patients diagnosed at age 50 or older.

These results challenge the assumption that hereditary testing has limited value beyond early-onset disease. However, determining how many patients should undergo testing requires consideration of variant penetrance, clinical actionability, testing costs and the limitations of the underlying evidence.

UPTOWN examined patients without conventional hereditary-risk selection

The investigators conducted a systematic review and random-effects meta-analysis to estimate the diagnostic yield of germline multigene panel testing in patients with CRC.

Eligible studies offered testing to consecutive patients using next-generation sequencing panels containing at least five genes.

Importantly, patients were not selected for testing on the basis of family history, mismatch repair status or other conventional hereditary-risk criteria.

This distinction is central to the study.

Many previously published hereditary cancer studies enrolled patients already considered at increased genetic risk. Such populations can overestimate the diagnostic yield expected when testing is offered more broadly.

UPTOWN attempted to address that limitation by examining comparatively unselected CRC cohorts.

The investigators searched PubMed, Embase, Scopus and the Cochrane Central Register of Controlled Trials, with searches updated through 2 August 2025.

Of 2,707 records identified, 21 studies met the eligibility criteria.

The principal pooled analysis included 12 studies comprising 6,925 patients. Additional age-stratified analyses incorporated other eligible cohorts, including studies focusing specifically on early-onset CRC.

The primary outcome was the proportion of patients with at least one pathogenic or likely pathogenic germline variant in a cancer predisposition gene.

The investigators also examined whether diagnostic yield varied according to age at diagnosis, family history, the number of genes included in testing panels and publication year.

By separating variants according to penetrance and cancer association, the study also explored an important practical issue: detecting a genetic variant is not necessarily equivalent to identifying a finding with established management implications.

One in seven patients had a pathogenic or likely pathogenic variant—but what does that mean?

Across the primary analysis, the pooled prevalence of any pathogenic or likely pathogenic germline variant was 15.2% (95% confidence interval [CI] 12.8–18.1).

This is a substantial finding in patients not specifically selected because of a hereditary cancer phenotype.

However, the 15.2% figure should not be interpreted as the proportion of patients with definitively established hereditary CRC syndromes.

The investigators distinguished different categories of genetic findings.

The pooled analysis revealed important differences in the types of germline variants identified.

Overall, 15.2% of patients carried at least one pathogenic or likely pathogenic germline variant. However, the prevalence of high-penetrance variants was 7.9%, indicating that not all detected variants carried the same level of inherited cancer risk.

More specifically, 5.4% of patients had high-penetrance variants associated with colorectal cancer predisposition.

These findings highlight an important distinction: although approximately one in seven patients had a clinically relevant germline finding, a smaller proportion carried high-penetrance variants directly associated with hereditary colorectal cancer susceptibility.

High-penetrance variants generally represent inherited alterations associated with a greater probability of cancer development than low-penetrance variants.

The distinction matters because the clinical implications of a positive result depend on the gene, the variant, its associated cancer risks and the strength of evidence supporting management recommendations.

Approximately 1.7% of patients carried low-penetrance findings, including APC p.I1307K or a single MUTYH pathogenic variant.

The authors noted that these findings do not have universally agreed management implications.

When these low-penetrance variants were excluded, the overall diagnostic yield was 13.0%.

Therefore, the study provides at least two clinically relevant messages.

First, potentially important germline findings are not uncommon in broadly tested CRC populations.

Second, the diagnostic yield becomes more informative when variants are separated by their clinical significance rather than grouped under a single positive-testing category.

Hereditary findings remain common after the age of 50

The most consequential analysis concerned age at CRC diagnosis.

As expected, the diagnostic yield was higher among patients diagnosed before age 50.

However, substantial yields persisted in older patients.

The analysis demonstrated that pathogenic or likely pathogenic germline variants were identified in both younger and older patients with colorectal cancer, although the prevalence was higher among those diagnosed before age 50.

Among patients diagnosed before age 50, 17.7% carried a pathogenic or likely pathogenic germline variant, compared with 13.0% of patients diagnosed at age 50 or older.

The difference was more pronounced for high-penetrance variants, which were identified in 14.2% of younger patients versus 6.5% of older patients.

When restricted specifically to high-penetrance variants associated with colorectal cancer predisposition, the prevalence was 12.2% among patients diagnosed before age 50, compared with 3.8% among those diagnosed at age 50 or older.

These findings demonstrate that although younger patients have a higher diagnostic yield, clinically important hereditary cancer predisposition variants remain detectable in patients diagnosed after age 50. Age alone may therefore be an insufficient criterion for determining eligibility for germline testing.

The early-onset analysis included 7,444 patients from 16 studies. The analysis of patients diagnosed at age 50 or older included 5,243 patients from 11 studies.

The older group had a pooled prevalence of 13.0% (95% CI 10.8–15.7) for any pathogenic or likely pathogenic variant.

Even after excluding low-penetrance variants, the yield remained 10.5%.

High-penetrance variants were identified in 6.5% of patients diagnosed at 50 or older, while high-penetrance variants in genes predominantly associated with CRC occurred in 3.8%.

These findings demonstrate that hereditary genetic findings are not confined to early-onset CRC.

Nevertheless, the differences between the age groups should not be simplified into an argument that age is clinically irrelevant.

Age remained associated with diagnostic yield, particularly for high-penetrance CRC-associated variants.

Rather, the findings suggest that age alone may be insufficient to define who could benefit from genetic evaluation.

The distinction is important when considering broader access to testing, particularly in patients who do not meet conventional hereditary-risk criteria.

The proposed 67-year benchmark is a model estimate, not a new testing cutoff

One of the most interesting findings came from the meta-regression analysis.

The investigators found that the diagnostic yield of high-penetrance variants progressively declined as the age at CRC diagnosis increased.

However, the predicted yield remained above 5% until approximately 67 years of age.

The authors noted that a 5% pre-test probability has been used as a threshold for recommending genetic evaluation.

At first glance, this finding might appear to support routine germline testing up to age 67.

Such an interpretation would be premature.

The 67-year estimate was derived from a meta-regression examining the relationship between the median age of study cohorts and their observed testing yields.

It was not established through prospective testing of individual patients at different ages.

Consequently, the estimate describes a between-study association rather than an individualized prediction of a patient's probability of carrying a pathogenic variant.

The reported uncertainty was also substantial. The confidence interval around the estimated age at which the yield crossed 5% extended from approximately 59 to 75 years.

The study authors specifically cautioned against interpreting 67 years as a rigid threshold.

This is an important methodological distinction for clinicians.

A statistically estimated benchmark can help frame future research or policy discussions, but it is not automatically suitable for determining whether an individual patient should receive testing.

The findings support reconsidering restrictive age-based strategies. They do not establish a validated new age cutoff.

Testing more genes does not necessarily improve the clinically meaningful yield

An additional question concerns the composition of germline testing panels.

If broader access to testing is considered, should clinicians also favour increasingly large gene panels?

UPTOWN found that panel size did not significantly influence overall diagnostic yield or the detection of high-penetrance variants across the full analysis.

Among patients diagnosed at age 50 or older, larger panels were associated with increased detection of variants of any penetrance, but not with significantly greater detection of high-penetrance variants.

This distinction deserves attention.

Increasing the number of genes tested may identify additional inherited findings. However, those findings may not all have the same implications for CRC risk or clinical management.

The authors' interpretation was that expanding the population offered testing could be more valuable than continually increasing panel size, provided clinically important high-penetrance genes are included.

The study also identified different age-related patterns across gene categories.

High-penetrance variants in genes predominantly associated with CRC became less common with increasing age.

In contrast, high-penetrance variants in genes traditionally associated with other cancers showed a more stable pattern across age groups.

These non-CRC-associated genes contributed approximately 1.5% to overall diagnostic yield, with findings largely involving BRCA1, BRCA2 and CHEK2.

The investigators cautioned that some of these findings could reflect background variant prevalence rather than a direct causal contribution to CRC.

This underscores why variant interpretation must remain gene-specific.

A pathogenic germline variant can have important implications for hereditary cancer assessment without necessarily being established as the cause of the patient's colorectal malignancy.

A positive finding may matter beyond the patient with colorectal cancer

The clinical relevance of germline testing extends beyond identifying inherited susceptibility in the person receiving cancer treatment.

A pathogenic variant may create an opportunity for cascade testing, in which relatives are offered targeted genetic evaluation after a familial pathogenic variant has been identified.

This is an important consideration when estimating the potential value of broader germline testing.

A positive result in one patient may identify previously unrecognized hereditary risk within a family.

Depending on the gene and the applicable established guidance, that information may support individualized genetic counselling, risk assessment and preventive planning.

However, UPTOWN did not directly measure the real-world effectiveness of cascade testing.

The analysis did not determine how many relatives underwent testing, how frequently preventive interventions followed, or whether broader testing reduced subsequent cancer incidence or mortality.

It also did not establish that all detected variants would justify the same clinical action.

The family-level implications therefore provide a clinically reasonable rationale for further evaluation, rather than direct evidence of demonstrated downstream benefit in the included cohorts.

For gastroenterologists, the findings reinforce the importance of recognizing when genetic expertise may be relevant during CRC assessment.

For clinical genetics services, they raise important implementation questions about counselling capacity, access, interpretation and subsequent family testing.

Family history remains informative, but the study questions relying on it alone

The investigators examined whether the prevalence of CRC in first-degree relatives influenced germline testing yield.

A higher prevalence of first-degree family history was associated with increased detection of pathogenic variants overall and in genes predominantly linked to CRC.

However, it was not associated with increased detection of high-penetrance variants in genes traditionally associated with other cancers.

These findings support the continuing relevance of family history.

At the same time, they highlight a limitation of testing strategies that depend primarily on whether a recognized hereditary-risk pattern is present.

UPTOWN specifically included patients offered testing without selection based on family history or conventional hereditary criteria, yet substantial genetic findings were still identified.

This suggests that relying exclusively on recognizable hereditary patterns could miss some pathogenic germline variants.

However, the study did not directly compare a universal testing pathway against a specific guideline-based selection strategy in a randomized clinical setting.

It therefore cannot quantify the incremental clinical benefit, cost or resource requirements of replacing a selective strategy with universal testing.

The appropriate role of family history remains one component of a broader risk assessment rather than a question resolved by this meta-analysis.

High diagnostic yield does not automatically justify universal testing

Despite the strength of the pooled findings, several limitations constrain their immediate application.

Substantial heterogeneity

Between-study heterogeneity was high.

For the primary outcome, the I² statistic was 84.1%. For high-penetrance variants, it was 88.4%.

These values indicate considerable variation between the included studies beyond what would be expected from random sampling alone.

Differences in gene panels, population characteristics and study settings may have contributed to this variation.

Consequently, the pooled estimates should not be assumed to apply identically across healthcare systems or patient populations.

Age-related representation

Many included cohorts came from tertiary or academic settings and had median ages below the population median reported for CRC in the United States.

Patients older than 80 years, community-based populations and lower-income healthcare settings were relatively underrepresented.

This limits the precision with which the results can be generalized to those populations.

Variant classification

The meta-analysis relied on the variant classifications reported by the original studies.

Differences in classification methods and changes in genetic knowledge over time may influence the interpretation of individual findings.

Some earlier studies also did not include genes more recently recognized as cancer predisposition genes.

Outcomes not assessed

The primary endpoint was diagnostic yield, not improved patient survival, reduced cancer incidence, cost-effectiveness or uptake of familial cascade testing.

The study also did not assess patient preferences or psychological consequences of testing.

These omissions are especially relevant when considering a population-wide testing policy.

A test may identify a clinically relevant inherited finding without demonstrating that universal implementation is the most effective or efficient approach.

What should clinicians take from the evidence now?

UPTOWN provides a substantial quantitative foundation for reconsidering how germline testing eligibility is determined in CRC.

The persistence of pathogenic germline variants among patients diagnosed at age 50 or older suggests that age-restricted strategies warrant further evaluation.

However, the results should not be presented as a new professional society guideline or an established mandate for universal genetic testing.

For practising gastroenterologists and GI oncologists, the immediate value lies in recognizing that absence of early-onset CRC does not exclude meaningful inherited findings.

Assessment should continue to incorporate relevant family history, personal cancer history, tumour characteristics and established genetic risk models, consistent with applicable clinical guidance.

The study also emphasizes the importance of distinguishing any positive pathogenic finding from variants with a well-established, high-penetrance cancer association.

Before a broader testing strategy can be adopted confidently across different healthcare settings, prospective evaluation is needed.

Future research should examine the additional yield achieved beyond existing selection criteria, the proportion of findings that lead to appropriate management changes, uptake of cascade testing and the long-term effect on patient and family outcomes.

Formal economic analyses are also needed to establish whether expanded testing provides acceptable value within different healthcare systems.

Research involving older patients and populations outside specialist academic centres will be particularly important.

Clinical Takeaway: Age should inform genetic assessment, not define its entire value

The UPTOWN meta-analysis found pathogenic or likely pathogenic germline variants in 15.2% of broadly tested patients with CRC, with a substantial 13.0% yield persisting in those diagnosed at age 50 or older.

The evidence challenges an overly restrictive reliance on age while confirming that diagnostic yield declines with increasing age.

However, the proposed 67-year benchmark is a model-derived estimate, not a validated clinical cutoff, and the study does not demonstrate that universal testing improves patient outcomes.

GastroAGI perspective: These findings support broader consideration of germline testing in CRC and provide a strong rationale for reassessing age-dependent eligibility criteria. The next step is to determine which expanded testing strategies deliver clinically actionable findings, meaningful family benefits and appropriate use of healthcare resources.

The question is no longer simply whether pathogenic germline variants occur in older patients with CRC. It is how reliably, efficiently and equitably those findings can be translated into better care.

Primary Reference

Mannucci A, Puzzono M, Marzocca G, et al. Germline multigene panel testing for colorectal cancer: a systematic review and meta-analysis. The Lancet Gastroenterology & Hepatology. Published online 3 September 2026.

DOI: https://doi.org/10.1016/S2468-1253(26)00187-1

Supporting approved source

Barbor M. Meta-analysis finds germline variants in 15% of patients with colorectal cancer. GI & Hepatology News. 13 October 2026.

https://news.gastro.org/issues/2026/october-2026/metaanalysis-finds-germline-variants-in-15-of-patients-with-colorectal-cancer/