Genetic Testing Guidelines for Hereditary Cancer Miss Many At Risk

The National Comprehensive Cancer Network genetic testing criteria for hereditary cancer miss almost one in five at-risk for breast and ovarian cancer, and more than one in three at-risk for Lynch Syndrome, according to a study published in the July issue of the Journal of Molecular Diagnostics.

“These data suggest that personal history alone is a poor indicator of pathogenic variant carrier status,” write the authors led by Cynthia Neben, from Color Genomics in Burlingame, Calif. “More individuals are at genetic risk for hereditary cancer than are identified by current testing guidelines and/or use of single-gene or single-site testing.”

Color Genomics reports the results from 23,179 individuals (83.1 percent women; 52.1 percent Caucasian) who were referred, independent of current testing guidelines, for the Color Hereditary Cancer Test, a 30-gene next-generation sequencing panel testing that examines hereditary cancer risk for breast, ovarian, uterine/endometrial, colorectal, melanoma, pancreatic, prostate, and stomach cancer) between May 2016 and September 2017. Every variant was reviewed by at least two variant scientists and all pathogenic and likely pathogenic variants were confirmed — single-nucleotide variants and insertions and deletions were confirmed with Sanger sequencing, while structural variants were confirmed by variant-specific polymerase chain reaction, array comparative genomic hybridization, or multiplex ligation-dependent probe amplification.

Overall, testing identified 2,811 pathogenic variants in 2,698 individuals for a pathogenic frequency of 11.6 percent or 9.1 percent when excluding common low-penetrance alleles. Irrespective of the presence of pathogenic variants, the frequency of individuals having a variant of uncertain significance in was 19.0 percent.

The researchers report that 3,845 individuals had a personal history of breast cancer (16.6 percent), 341 had ovarian cancer (1.5 percent), 438 had colorectal cancer (1.9 percent), and 1476 had another hereditary cancer associated with genes on the panel (6.4 percent). A personal history of stomach, colorectal, or ovarian cancer correlated with the highest pathogenic frequencies at 23.4 percent, 23.1 percent, 17.3 percent respectively, or 19.1 percent, 17.1 percent, or 17.3 percent when excluding common, low-penetrance alleles. Among the 42.4 percent of the cohort not reporting a personal history of cancer, the pathogenic frequency was 10.2 percent (7.1 percent, excluding common, low-penetrance alleles.

A majority of the 2,811 pathogenic variants identified (61.5 percent) were in CHEK2, BRCA2, MUTYH, and BRCA1, followed by 9.5 percent in APC and 6.2 percent in ATM. After excluding common low-penetrance alleles, of the 2,698 positive results, 846 were from BRCA1 and BRCA2, 189 from a Lynch syndrome gene (MLH1, MSH2, MSH6, and PMS2), 147 low-penetrance allele (a monoallelic MUTYH pathogenic variant or APC p.I1307K), and 1,016 other genes.

Single-nucleotide variants accounted for 58.9 percent of all pathogenic variants, while insertions and deletions accounted for 35.4 percent and structural variants 5.7 percent. Just over half of all insertions and deletions were found in BRCA2 and BRCA1 (51.7 percent). In addition, nearly one-third (32.1 percent) of structural variants were also in BRCA1.

Overall, more than one-third of those tested and provided a complete family history (38.7 percent) would not have met guideline-driven testing criteria. In this ineligible population, the pathogenic frequency was 8.2 percent (6.2 percent when excluding, common low-penetrance alleles). However, more than one in five individuals with pathogenic variants in genes with well-established genetic testing recommendations did not meet corresponding National Comprehensive Cancer Network testing criteria (21.7 percent). Specifically, of the 749 individuals who had a pathogenic variant in BRCA1, BRCA2, TP53, or PTEN, 138 (18.4 percent) would not have met criteria for genetic testing for breast and ovarian cancer, while 144 individuals who had a pathogenic variant in MLH1, MSH2, PMS2, or MSH6, 52 (36.1 percent) would not have met criteria for genetic testing for Lynch syndrome.

“As the cost of sequencing continues to decrease and genetic testing becomes more accessible to a broader population, we will gain a better understanding of the genes associated with elevated risk for hereditary cancer,” the authors write. “With this additional knowledge, it will be imperative to reevaluate the genes included on multi-gene panels. The genes on this panel were selected on the basis of studies in high-risk cohorts and, thus, the population-based prevalence and penetrance of many of these genes are unknown.”

Takeaway: Use of personal cancer history and National Comprehensive Cancer Network testing criteria both miss individuals that carry pathogenic variants raising hereditary cancer risk.