NEW YORK (GenomeWeb) – An international consortium of researchers has developed an ultrasound-based method for predicting the risk of malignancy in adnexal masses.
Detailed in a paper published this week in the American Journal of Obstetrics and Gynecology, the test looks at 10 ultrasound features of adnexal masses to determine whether they are likely cancerous or benign.
Though not itself a molecular diagnostic, the test has relevance to proteomic diagnostics like Vermilion's OVA1 and Fujirebio's ROMA that are, likewise, intended to determine whether an adnexal mass is malignant or benign.
But while the method, called the Simple Rules risk test, is a potential competitor to these proteomic diagnostics, Dirk Timmerman, a researcher at KU Leuven and one of the developers of the test, told GenomeWeb that he hopes ultimately to combine the ultrasound-based approach with molecular tests like OVA1 and ROMA.
"We are very interested in combining biochemical tests with ultrasound, because I think the future is in the combination of the two methodologies," he said. "There are always tumors that are very difficult [to assess] with ultrasound, even if you are very experienced. And then, also, you will always have cases that are very close to the cut-off used to asses malignancy. So a second-stage [molecular] test would be very important if it were reliable."
Developed by the International Ovarian Tumor Analysis group, the Simple Rules test considers a total of 10 ultrasound features, five associated with benign tumors and five associated with malignant.
The first version of the test (referred to as original Simple Rules), which was published in 2008 based on data from more than 1,000 patients, was a basic set of guidelines aimed at enabling clinicians to evaluate a mass without the need for computers or algorithms.
This initial test was limited, though, by the fact that results were inconclusive in cases where both benign and malignant features were present or where no features were present at all, Timmerman said.
"If one or more benign features are present and no malignant features, the tumor is classified as benign," he said. "And if one or more malignant features are present without any benign features, then it is classified as malignant. But if no features are present, you cannot classify it, and if conflicting features are present you cannot classify it either."
This, Timmerman noted, meant that ovarian masses in around 20 percent to 25 percent of patients could not be classified.
The new version of the test, on the other hand, uses logistic regression analysis to estimate the risk of malignancy based on the 10 ultrasound features, allowing clinicians to classify all tumors while also providing the individual percent risk of malignancy. The test was developed using data from 5,020 patients enrolled between 2002 and 2012, with malignancy status ultimately determined by histopathology.
Depending on the risk of malignancy cut-off used, the test was able to distinguish benign from malignant masses with sensitivity and specificity ranging from, respectively, 99.7 percent and 33.7 percent (1 percent risk cutoff) to 89.0 percent and 84.7 percent (30 percent risk cut-off).
These findings, Timmerman said, indicate that the method could be effective in managing patients with adnexal masses, allowing clinicians to determine the likelihood that their mass is malignant, in which case they would be referred to a gynecological oncologist for surgery, or benign, in which case the surgery could be done by their general gynecologist. Studies have shown this patient triage to positively influence patient survival.
This is the same intended use for Vermillion's OVA1 test, and Timmerman and his co-authors noted that while a comparison of the Simple Rules test and OVA1 on the same patient cohort has not been done, OVA1 "seems to have lower specificity at similar sensitivity," which leads to higher numbers of false positives.
Indeed, specificity has been an issue for OVA1 since its launch. Vermillion has aimed to address this problem with its next-generation test, Overa, which recent data released by the company shows has improved on OVA1's specificity and positive predictive value by 28 percent and 29 percent, respectively. Overa received a CE mark last year and is currently under US Food and Drug Administration review.
In general, Timmerman said, while there was much excitement within the field about the potential of proteomic tests, this excitement has faded somewhat as such tests have failed thus far to deliver notable improvements over existing methods.
"Some years ago there was a lot of enthusiasm, but in reality you see that the test is very specific or very sensitive, but a good combination of the two does not exist at the moment," he said.
He noted, though, that he was hopeful that adding molecular information to methods like ultrasound could ultimately improve testing.
"We hope that by combining different sources of information it will lead to a better test, but this is still to be explored; the research is ongoing," Timmerman said.
In fact, Vermillion itself explored such a combination of methods in a 2014 study published in AJOG,in which researchers looked at 1,024 women who had ovarian imaging, OVA1 analysis, and surgery for an adnexal mass. They found that combining ultrasound and OVA1 significantly improved specificity compared to either test done separately, though at the cost of lowered sensitivity.
And, last year, the company said that it was, through its wholly owned subsidiary Aspira Labs, in the process of building new laboratory-developed tests combining its Overa test with patient ultrasound results.
Vermillion did not respond to several requests for comment.
Timmerman said that he and his colleagues have in the past tried to incorporate OVA1 in some of their work, but that the test is not currently available in European clinical practice and that its high cost presents a challenge in terms of obtaining the test for their research.
The original Simple Rules test is currently widely used throughout Europe and is incorporated in the guidelines of the UK's Royal College of Obstetricians and Gynaecologists, Timmerman said. But, he noted, it is relatively unknown in the US. He said he hopes the AJOG publication on the improved Simple Rules risk test, along with a series of lectures he is giving this year at US institutions, will help spread its adoption.
"In the US the original Simple Rules test is almost non-existent at this moment, but I hope that this paper in an American journal will change this," he said.