What did the age analysis show?

The team classified each patient's cancer as early-, average- or late-onset based on the typical age distribution for that specific cancer type, rather than a blanket cutoff at 50. Early-onset meant developing the cancer more than one standard deviation below the average age for that tumor type; late-onset meant more than one standard deviation above it.

Under this classification, the prevalence of pathogenic variants was:

• 18.4% among early-onset patients
• 15.6% among average-onset patients
• 12.3% among late-onset patients

High-penetrance variants — changes highly likely to produce the cancer risk they are associated with — occurred in 9.1% of early-onset patients, 5.5% of average-onset patients and 2.6% of late-onset patients. Moderate-penetrance variants appeared in 3.3%, 3.4% and 2.5% of the three groups, respectively.

Taken together, high- and moderate-penetrance variants, which carry potential implications for treatment strategies, turned up in about 9.7% of all patients.

The gap between early- and late-onset groups is real but far smaller than the age-50 cutoff assumes. That, Stadler said, is exactly the problem.

What would universal testing change?

The study's co-corresponding author, Dr. Luis A. Diaz Jr., head of Solid Tumor Oncology at Memorial Sloan Kettering, said the findings should shift clinical practice.

"Hereditary cancer can occur at any age, and we should move toward a new standard in which genetic testing is considered for every patient diagnosed with cancer, regardless of age," Diaz said. "Our manuscript shows that relying on age at diagnosis misses a substantial proportion of patients with germline pathogenic variants."

Beyond guiding the patient's own treatment, universal testing would expand cascade testing — screening a patient's relatives once a variant is found, so relatives can learn whether they carry the same risk.

"Cascade testing remains one of the most underused opportunities in cancer prevention," Stadler said. "Once a hereditary cancer predisposition variant is identified in one patient, testing relatives becomes highly targeted and informative, yet only a minority of eligible relatives currently undergo testing."

What are the study's limits?

The findings come from a single major tertiary cancer center, which may not reflect the broader patient population. Certain cancer types were over- or underrepresented in the cohort. The study population was also predominantly of European and Ashkenazi Jewish ancestry, so the results may not generalize to other groups. Whether universal germline testing should become standard care will require confirmation in more diverse, multi-center settings.