The need for universal screening for Lynch syndrome in elderly patients with newly diagnosed colorectal cancer (CRC) has been questioned, according to results from a retrospective cohort study.
In addition, discontinuing reflex CRC screening for Lynch syndrome in patients over age 80 years could be feasible because of very low efficiency.
“The universal strategy advocates screening all patients with newly diagnosed CRC for Lynch syndrome and has been shown to be the most sensitive method,” wrote Dan Li, MD, of Kaiser Permanente Northern California, Santa Clara, and colleagues. The findings were published in Annals of Internal Medicine.
The researchers studied 3,891 patients with newly diagnosed CRC who were screened for Lynch syndrome from 2011 to 2016. Data were collected from a population-based screening program at Kaiser Permanente Northern California.
“The system provides comprehensive medical care for more than 4 million members across 21 medical centers covering urban, suburban, and semirural areas,” Dr. Li and his colleagues wrote.
To compare universal and age-restricted screening, the team obtained surgical samples of all newly diagnosed CRC tumors and tested them for reflex mismatch repair protein expression using immunohistochemistry.
Subsequently, the age-restricted screening groups were divided into several age categories, ranging from age 50 to 85 years.
The diagnostic yield, defined as the “percentage of patients with pathogenic reflex mismatch repair gene variants among all patients with CRC screened with immunohistochemistry,” was measured and compared with the universal screening technique.
“We calculated the number of patients with CRC who needed to be screened in each age group to identify one case of Lynch syndrome by dividing the number of patients screened in each age group by the number of Lynch syndrome cases diagnosed in that group,” they explained.
After analysis, the researchers detected a total of 63 cases of Lynch syndrome (diagnostic yield, 1.62%) with universal screening, among which 5 (7.9%) were over age 70 years and 1 (1.6%) was over age 80 years.
When patients with CRC who were universally screened were used as the denominator, 58 cases (diagnostic yield, 1.49%) were detected in those with CRC diagnosed at or prior to age 70 years.
In addition, in patients diagnosed at or before age 75 and 80 years, 60 and 62 cases of Lynch syndrome (diagnostic yield, 1.54% and 1.59%) were detected, respectively.
“The incremental diagnostic yield decreased substantially after age 70 to 75 years,” they wrote.
With these findings, Dr. Li and his colleagues suggested that cessation of screening for Lynch syndrome post age 80 years may be acceptable, especially in resource-limited environments.
“Using age as the primary criterion is a simple method of selecting patients for Lynch syndrome screening in clinical practice,” they added.
In accordance with previous studies, a major reduction in Lynch syndrome incidence has been noted among elderly populations.
There remains a need for additional studies exploring the effects of diagnosing Lynch syndrome in elderly patients on family members.
The researchers acknowledged a key limitation of the study was that patients who did not finish germline analysis but were eligible for it were excluded from certain measurements. To reduce potential bias, the team conducted a sensitivity analysis, and the findings were negligible with respect to main results.
“Given the geographic variation in the reported prevalence of Lynch syndrome, the diagnostic efficiency of Lynch syndrome screening among elderly populations should be further investigated in other populations,” they concluded.
The study was funded by Kaiser Permanente Northern California. The authors reported financial affiliations with Bayer, Clinical Genomics, Covidien, Exact Sciences, Motus GI, Quorum, Universal DX, and the National Cancer Institute.
SOURCE: Li D et al. Ann Intern Med. 2019 Jun 11. doi: 10.7326/M18-3316.