Large Database Study Finds 11 Genes Associated with Increased Ovarian Cancer Risk

Allison W. Kurian, MD, MSc

The largest-ever genetic analysis of ovarian cancer patients yielded a short list of 11 suspect germline genetic mutations that may be associated with an increased risk for developing ovarian cancer, according to an oral presentation at the 2016 ASCO Annual Meeting in Chicago.

Among the list of likely culprits wasATM, which has previously been implicated in a heightened risk for breast cancer. “But this is the first timeATMhas been associated with an increased risk for ovarian cancer,” said Allison Kurian, MD, MSc, associate professor of Medicine and of Health Research and Policy at Stanford University School of Medicine, and director of the Stanford Women’s Cancer Genetics Clinic, who gave the presentation.¹

While the Human Genome Project was completed in 2003, clinical laboratories began compiling databases as early as 1995, when the relevance ofBRCA1/2was discovered, Kurian said. The relevance of additional mutations in cancer risk assessments are still being studied.

Researchers of the sequencing study also discovered a significant association to increased risk of ovarian cancer inBRIP1,MLH1,MSH2,MSH6,NBN,STK11,RAD51C, andRAD51D, in addition toBRCA1andBRCA2mutations.

However, the reliability of the findings varied greatly, sometimes being statistically significant, while at other times the prognostic value of the gene changed depending on the case-control method, according to the commentator.

In the study, Kurian and colleagues reviewed data from 95,561 women in the clinical laboratory database of Myriad Genetics. Initially, the researchers looked for a link in 25 genes, between 2013 and 2015. Mutations were detected in 6,775 patients (7%). Of these, 44% (3,007 patients) had mutations inBRCA1/2, while 56% (3,768 patients) were in other genes. Variants of unknown significance were found in 33% (31,788) of the patient pool. For the patients with ovarian cancer (n = 701), 14% had mutations, of which 63.5% were inBRCA1/2.

Kurian and colleagues calculated the association of each gene with an increased risk for developing ovarian cancer by multivariable logistic regression. Researchers also conducted a case-control analysis addressing the same question.

While the team initially queried an assay of 25 genes, only 11 were found to be significantly associated. The germline mutations ruled out includedAPC,BARD1,BMPR1A,CDH1,CDK4,CHEK2,MYH,P14ARF,P16,PALB2,PMS2,PTEN,SMAD4, andTP53.

Confidence in Findings Varies from One Gene to Another

“One of the most interesting finds wasATM, which causes a very rare syndrome called ataxia telangiectasia in children who inherit 2 bad copies. But what’s not so rare is people who carry a copy of the mutation. They may not get the rare childhood disease, but studies show that they might have an increased risk of getting breast cancer,” said Kurian, who noted that up to 1% of breast cancer patients may carry theATMmutation.

Notwithstanding the unusually large database, the statistical confidence of investigators in the reliability of the findings varied from one gene to the next. With regard toSTK11, for example, estimates of the odds ratio measuring the increased risk for ovarian cancer faced by women with the gene was 41.9.

“That has to be taken with a grain of salt, because there were a relatively small number of women with the gene. Their increased risk may be between 6 and 315 times that of those who don’t have the mutation,” Kurian said.

With regard to another mutation,MLH1, the finding was similarly in question. “When the matched case-control method was used, MLH1 flips from one side to the other,” said Linda Mileshkin, MBBS, associate professor at Peter MacCallum Cancer Centre in Melbourne, who served as a commentator on the presentation at ASCO.

At other times, the data seemed to suggest an unusually high degree of confidence in the findings. For those with theBRIP1mutation, for example, the study found an estimated increased risk for ovarian cancer of 2.62 times that of baseline. The true risk ranged from 1.7 to 3.8 times higher, with aPvalue of a decimal point with 181 zeroes behind it, Kurian said.

“This exciting next-generation sequencing research has value in the clinic. The key is that we need more studies in different populations, with a range of ovarian cancer risks. We need studies where we can test family members, and to do studies in more diverse populations,” said Kurian, who presented 3 such genomic studies at ASCO 2016.²

Kurian A, Hughes E, Handorf E, et al. Association of ovarian cancer (OC) risk with mutations detected by multiple-gene germline sequencing in 95,561 women.J Clin Oncol34, 2016 (suppl; abstr 5510).

See, for example, Kurian A, Idos G, Culver J, et al. Safety of multiplex gene testing for inherited cancer risk: Interim analysis of a clinical trial.J Clin Oncol34, 2016 (suppl; abstr 1503).