NEW YORK (360Dx) – Researchers at Uppsala University and the University of Gothenburg's Sahlgrenska Academy in Sweden are targeting early detection and screening for ovarian cancers by developing a multiplexed test that uses plasma protein biomarkers.
The platform, if proven useful in future clinical trials, could eventually be made available as a test that provides information needed for clinicians to distinguish healthy women from those who have ovarian cancer, and to more precisely select women for associated surgeries, Ulf Gyllensten, a researcher at Uppsala University and one of the test's developers, said in an interview.
The multiplex protein test has 11 plasma biomarkers that create mutually exclusive combinations of signatures, Gyllensten said. The group described the development and early clinical validation of the assay in a paper published recently in the journal Communications Biology. The panel consists of the protein biomarkers MUCIN-16, SPINT1, TACSTD2, CLEC6A, ICOSLG, MSMB, PROK1, CDH3, WFDC2, KRT19, and FR-alpha.
According to Gyllensten, the test has potential to enable detection of ovarian cancers before women begin to experience symptoms. If successfully validated, clinicians would use the new tool in parallel with ultrasound to identify with greater certainty women that need to be referred for surgery and differentiate them from women who can safely wait and undergo subsequent ultrasound monitoring or become healthy without surgery.
The assay combines new biomarkers with MUCIN-16, also known as cancer antigen 125 (CA-125), which is the most important single biomarker for epithelial ovarian cancer management, but "not considered sufficient" when it is used by itself to discriminate benign from malignant masses, Gyllensten said.
Ovarian cancer is commonly caught late, and the overall five-year survival rate is between 30 percent and 40 percent. Only about 15 percent of women are currently diagnosed early with a localized tumor, classified as Stage I, a point when the five-year survival rate is 92 percent, according to the National Cancer Institute.
"The reported work in Nature Communications reflects a novel take on an old diagnostic conundrum — how to improve upon CA-125," Cesar Castor, director of the cancer program at Massachusetts General Hospital Center for Systems Biology, said in an interview. "The researchers undertook significant a priori considerations and modeling to inform a diagnostic signature with good performance across validation sets," said Castro, who is not involved in the development of the test.
He and his colleagues are developing a multiplexed nanoplasmonic assay that has been tested on ovarian cancer patients and is being further developed for early detection of pancreatic cancer.
Castro said that the next step for the Swedish researchers will likely entail independent testing by outside investigators to ensure "any unaccounted variable did not drive the performance." Future performance testing of the Swedish team's assay could include women with high inherited risk of ovarian cancers, including women with BRCA mutations but without a pelvic mass, he said.
Castro added that, in general, a "high unmet need remains in the biomarker space to synergize with existing early detection imaging strategies," such as the use of ultrasound. Meanwhile, cure rates for ovarian cancers haven't improved due to an absence of robust early detection strategies, he said.
In the current study, the researchers in Sweden used high-throughput proteomics to identify multiple mutually exclusive biomarker signatures differentiating benign conditions from ovarian cancers at different stages, grades, and all histological subtypes. In a discovery cohort and two replication cohorts, they selected signatures from 593 characterized plasma proteins using a proximity extension assay (PEA).
To validate the performance of a proof-of-concept model that emerged from the discovery work, they developed a custom PEA assay that they used to characterize protein abundance levels in clinical samples, and tested its capability to distinguish benign from malignant tumors ranging in progression from Stages I to IV.
The 11-biomarker panel showed a positive predictive value of 92 percent, sensitivity of 85 percent, and specificity of 93 percent for detection of ovarian cancers ranging from Stages I through IV, Gyllensten said. Further, for women diagnosed with more severe cancers at Stages III and IV, the sensitivity of the panel increased to 91 percent; its specificity increased to 95 percent; and its positive predictive value increased to 94 percent, he added.
The discovery cohort consisted of 90 patients diagnosed with benign tumors and 79 patients with ovarian cancers ranging from Stages I through IV. A first replication cohort consisted of 71 patients diagnosed with benign tumors and 100 patients with ovarian cancers ranging from Stages I through IV. A second replication cohort consisted of 77 patients with ovarian cancers ranging from Stages I through IV, and a third replication cohort consisted of 106 patients with benign conditions, 28 with borderline diagnosis, and 93 with ovarian cancers ranging from Stages I through IV.
Gyllensten said that he and his colleagues are continuing to develop the biomarker panel and are validating it in a 700-patient cohort, with the objective of eventually making it commercially available for screening. The group is conducting the study in collaboration with an industrial partner, Uppsala-based Olink Proteomics, involving women presenting to clinics with pelvic pain who clinicians suspect could have ovarian cancer. The investigators expect that they will complete the study and publish its results later this year.
The proximity extension assay provides high throughput detection of protein biomarkers, and operates in a similar fashion to a sandwich-based ELISA, except with a more specific readout, Martin Lundberg, Olink's chief technology officer, said in an interview.
In operation, two different antibodies target and bind to a single protein of interest. When the two antibodies are in proximity to the protein, they generate a DNA signal read by qPCR. "We couple our antibodies with DNA barcodes, and when the two different antibodies are in proximity, one DNA barcode forms an extension of the other," Lundberg said.
In that way, the assay detects the abundance of DNA amplicons and associated proteins in a sample. Associating target proteins with DNA enables signal amplification and high levels of sensitivity and specificity from a small sample, Lundberg said, adding that the assay isn't affected by signal cross-reactivity or high background noise that can reduce performance of other antibody-based tests.
If the broader validation trial is successful Olink will consider doing further development and commercially launching the assay, Lundberg said. Olink could seek CE marking, as well as clearance from the US Food and Drug Administration that would allow the company to sell the test in a kit format in the US, he said.
Timelines for launching such a test are unclear, however, because Olink will await the results of the 700-patient clinical trial before deciding what to do next.
Overall, there is a collective motivation in ovarian cancer testing "to enhance the pre-analytical and pre-competitive space to ensure success for all stakeholders," Castro said. In general, ongoing work involves developing and analyzing a number of diagnostic methods, including those that employ "secreted proteins as in this work [and] other forward-thinking biomarkers such as exosomes or circulating tumor DNA," he said.
Among the tests already available for ovarian cancer diagnosis is the OVA1 test from Vermillion, which recently announced that the test will receive preferred coverage from Blue Cross Blue Shield of Texas and Blue Cross Blue Shield of Arizona as an in-network covered test.
A number of investigators also have reported progress in developing in vitro diagnostics for ovarian cancer detection, including a team from the University of Pittsburgh who recently reported that they are developing a combined protein assay that measures MIF, OPN, and IL-8 autoantibodies alongside CE125 to improve the ability to detect early-stage cancer.
Investigators are also looking at the applicability for detection of ovarian cancer of a liquid biopsy-based test being developed by HeiScreen.
Meantime, researchers at Griffith University and the University of Adelaide reported that serum from patients with all stages of ovarian cancer had elevated levels of N-glycolylneuraminic acid (Neu5Gc).