Developers of a genetic-susceptibility test for ovarian cancer might have jumped the gun, according to an OCRF-funded consortium that found no association between a KRAS mutation and risk of the cancer.
Analysis of genetic information on 9,500 ovarian cancer patients and 12,000 members of control groups yielded an odds ratio of 1.02. Separate analyses limited to familial ovarian cancer and BRCA1 mutation carriers produced similar, nonsignificant results.
The investigators also found no association between KRAS mutation and overall survival or progression-free survival, as reported here at the Society of Gynecologic Oncology meeting.
“These data exclude the possibility of an association between the KRAS single nucleotide polymorphism (SNP) and a clinically significant risk of sporadic or familial ovarian cancer,” said Andrew Berchuck, MD
, of Duke University. “Use of this single nucleotide polymorphism for ovarian cancer risk prediction does not appear warranted.”
By means of genome-wide association studies, Berchuck and colleagues at the Ovarian Cancer Association Consortium (OCAC)
have identified other SNPs that do affect ovarian cancer risk. The group’s goal is to combine genetic and epidemiologic risk factors to provide guidance for prevention and early detection. OCAC’s work is made possible by a grant from the Ovarian Cancer Research Fund.
Interest in KRAS testing for ovarian cancer susceptibility quickly evolved from a study published in July 2010 that showed an association between the KRAS SNP rs61764370 and both sporadic and familial ovarian cancer (Cancer Res 2010; 70: 6509-6515).
The study showed that 27% of patients with sporadic ovarian cancer had the KRAS SNP compared with 15% of a control group. With respect to familial ovarian cancer, the SNP occurred in 22% of BRCA1 mutation carriers, 8% of BRCA2 carriers, and 61% of patients with a family history of ovarian cancer.
Also in July 2010, MiraDx of New Haven, Conn., introduced the PreOvar test for the KRAS SNP to aid ovarian cancer genetic risk assessment.
Citing the limited data from a single study, the SGO issued a position statement in September 2010 recommending against use of the test in clinical practice until validation studies had been performed.
In an effort to validate the published findings, the OCAC initiated a study to examine the association of the KRAS SNP with ovarian cancer. The consortium represents investigators in 30 clinical studies involving 20,000 ovarian cancer patients and 20,000 participants in control groups.
The OCAC offered the means to examine the KRAS SNP in multiple study populations with stringent genotyping quality-assurance procedures, said Berchuck. In addition to examining the overall association of the SNP with ovarian cancer risk, the study’s objectives included assessment of the association in women who have a family history of ovarian cancer or who are BRCA1 carriers and to determine whether the SNP is associated with progression-free survival or overall survival.
The study included 8,669 cases and 10,012 controls from studies participating in the OCAC, as well as 683 cases of epithelial ovarian cancer and 2,044 carriers of BRCA1 mutations from studies participating in the Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA).
Berchuck and colleagues also examined the relationship between the KRAS SNP and prognosis in subsets of studies that had data on progression-free survival and all-cause mortality.
In the overall analysis, investigators found no evidence of an association between the KRAS SNP and risk of ovarian cancer, as the 95% confidence intervals crossed the 1.0 threshold (0.95 to 1.10).
Invited discussant Johnathan M. Lancaster, MD, PhD, of the Moffitt Cancer Center in Tampa, Fla., credited the investigators for use of a much larger and more diverse database compared with the study that provided impetus for the genetic test. Moreover, he noted, the genotyping results of 13 studies came from a single laboratory, ensuring consistency.
Lancaster asked Berchuck to speculate on the basis for the differences between the data reported last year and the results in the study reported at SGO. In response, Berchuck cited two potential factors: genotyping errors and an inherent propensity for false-positive results in the genetic association studies employed by investigators in the study published in 2010.
Berchuck A, et al “Association of KRAS SNP rs61764370 with risk of invasive epithelial ovarian cancer: Implications for clinical testing” SGO 2011; Abstract LBA 3.