NEW YORK – An Italian team led by investigators at Humanitas University has come up with a test for detecting early-stage high-grade serous ovarian cancer (HGSOC) that relies on DNA sequencing of cervical smear samples collected during routine Papanicolaou (Pap) testing.
"The study provides the basis for a new approach to the early detection of high-grade serous ovarian cancer," Maurizio D'Incalci, a researcher affiliated with Humanitas University and Humanitas Research Hospital, said in an email.
In a study published in Science Translational Medicine on Wednesday, D'Incalci and colleagues from Italy retrospectively analyzed 250 archival Pap smear samples from 113 women who went on to develop HGSOC and 77 women who did not, using low-coverage whole-genome sequencing to discover copy number alterations associated with ovarian cancer.
Together, the retrospective data provided in the paper demonstrated that "early detection of high-grade serous ovarian cancer is potentially achievable by examining the genomic instability of DNA extracted — [and] assessed by shallow DNA sequencing — from archival Pap test smears taken years before the diagnosis," D'Incalci explained.
The approach follows from prior research pointing to the presence of TP53 mutations in DNA from Pap tests that originated from women with ovarian tumors, he explained. Because pathogenic TP53 mutations can also turn up in normal tissue samples, though, the current analyses focused on a broader collection of copy number changes linked to HGSOC in an effort to come up with potential diagnostic markers for early-stage ovarian cancer.
After demonstrating that a subset of 62 individuals from the pre-HGSOC discovery group consistently had higher copy number profile abnormality (CPA) scores in Pap test DNA compared to their cancer-free counterparts, for example, the team used CPA scoring to assess genomic instability in 80 HGSOC samples from the Cancer Genome Atlas (TCGA) project.
From there, the investigators set CPA score thresholds for negative, uncertain, and positive early ovarian cancer tests using Pap DNA profiles for 65 pre-HGSOC individuals and 25 unaffected women before applying the test to retrospective samples from other members of the pre-HGSOC group.
In an assessment based on regions flagged in the TCGA analysis, the test had 76.7 percent accuracy and nearly 67.7 percent specificity at a set sensitivity of 100 percent. The accuracy jumped to 81 percent with a sensitivity of 75 percent and 96 percent specificity when the researchers based the CPA scores on genome-wide somatic copy number alteration patterns rather than defined regions.
"This proof-of-principle study indicates that the early diagnosis of HGSOC is feasible through the analysis of genomic alterations in DNA from endocervical smears," the authors wrote.
When they tracked CPA scores over time in individuals who were ultimately diagnosed with HGSOC, meanwhile, the investigators found that their assay uncovered signs of the disease as far out as nine years before a patient's diagnosis.
All told, more than three-quarters of women diagnosed with HGSOC showed signs of aneuploidy in their Pap test DNA prior to diagnosis, while 9.2 percent more had results classified as uncertain after low-coverage genome sequencing. In samples collected between two and five years before diagnosis, they picked up aneuploidy and uncertain results in 71 percent and 14 percent of samples from pre-HGSOC participants, respectively — results backed up by CPA scores found in TCGA ovarian cancer samples.
Even so, the team cautioned that larger, prospective studies are needed to further test the approach and its effects on HGSOC outcomes.
"To demonstrate that the test benefits the survival of patients with HGSOC, a prospective study has to involve necessarily a very large population of women and a long follow-up time with longitudinal sampling," the authors wrote, noting that "the relative simplicity and cost-effectiveness of the procedure render serial analyses from the same patient, such as every [six] months or every year, highly feasible."