NEW YORK (GenomeWeb) – A study of a primary cervical cancer screening setting in Scotland has shown that patient-collected vaginal samples have comparable results to clinician-collected cervical samples when tested using Roche's cobas 4800 HPV assay.
Adoption of self-collected samples for cervical cancer screening via human papillomavirus testing must overcome hurdles of regulation and acceptance by clinicians. Comprehensive studies could influence the latter, chipping away at biases favoring clinician-collected samples.
Specifically, a study published last month in BMJ Open considered 5,318 women receiving routine cervical cancer screening in a part of southern Scotland between April 2013 and July 2014. It evaluated sensitivity and specificity in detecting pathological lesions above cervical intraepithelial neoplasia grade 2, or CIN2+, comparing patient-collected vaginal and urine samples to clinician-collected swabs of the cervix.
Sensitivity for detecting CIN2+ using the Roche test was about 98 percent with cervical samples and 95 percent for patient-collected vaginal samples, but only 63 percent for "random void" urine samples. Specificities were about 87, 85, and 90 percent, respectively.
On a subset of about 500 samples retested at two different labs, the researchers also showed nearly 100 percent intralaboratory agreement in HPV positivity for vaginal and urine samples, and similarly high interlaboratory agreement using the Roche test.
"The exciting thing about the study is it's real-world — the researchers now have a set of data with clinical outcomes that says [the Roche] test can work this way," Sean Boyle, Roche's director of medical affairs, told GenomeWeb in an interview.
Molecular testing as a standalone cervical cancer screen remains somewhat controversial. Molecular diagnostic tests for HPV are frequently used in conjunction with a Pap smear in a "co-testing" protocol for cervical cancer screening. Some argue that co-testing is superior, or more cost-effective if paired with a specific diagnostic test.
Likewise, testing clinician-collected cervical samples is the standard of care, but there is a growing body of evidence suggesting patient-collected vaginal samples are also feasible.
The Roche HPV test is the only HPV assay that is approved by the US Food and Drug Administration for use alone in primary screening. It was also recently adopted for that purpose as the official primary screening test in the Netherlands.
Technology enabling women to collect their own samples for molecular testing could impact women in many different settings and contexts. In low-resource areas, clinics might save money by using self-collected samples. In higher-resource areas self-collection may appeal to women who avoid traditional testing, and could even be performed on sample kits sent in the mail.
Roche has a self-collection claim on its Chlamydia trachomatis and Neisseria gonorrhoeae test which was part of the original FDA clearance in 2012. The Scottish investigators used self-sampling collection swabs from Roche that were validated for CT/NG self-vaginal sampling, according to the study.
The use of the test in this way would be considered "off-label" at this time, Boyle noted.
"These kits are approved for self-collection in CT/NG — when [the first author] reached out to us and said she was interested in doing a self-sampling study we thought why not try this kit since it is already out there," Boyle said. Roche provided both the self-sampling kits and the HPV assays for the study.
Boyle noted that this "off-label" use was also likely the case with a recently reported study using self-collection and an HPV test from Cepheid.
"Now the onus is on Roche to take that back and do all the appropriate development studies and share that with regulatory bodies so that we can get an official indication," Boyle said. He added that it is unclear at this time whether the company will pursue the self-collection claim in the US.
There is more consensus around self-collection for women in low-resource settings or women who avoid traditional screening than for replacing clinician collection in established screening programs.
Boyle also said unscreened women might benefit the most from this test. "If you have a good screening program in place already, you want to be careful that self-sampling doesn't erode that screening program," he added.
A "test-and-treat" strategy, which favors molecular testing over pathology, might also be appropriate for targeting unscreened women in developing countries. "In a developing country they may only have one shot to see these women, so when you screen them and if they are identified as being at elevated risk you might just want to treat them right away."
Boyle noted that studies of self-sampling, using the Roche test and others, are becoming more frequent, and are more frequently showing patient-collected samples yield similar, if slightly less sensitive results compared to clinician-collected samples. But if women who are not otherwise getting screened could be captured with a slightly less sensitive approach, the benefits might outweigh the risks, Boyle said.
But in the overall gynecological testing space, the emphasis on self-collection remains to be determined.
"There are so many possible solutions now that the data are coming out, and a solution for a developed country is not the same as a solution for a low-income country, and a country that has a good screening program in place may not need the same solution as a country that doesn't have a good screening program," Boyle said.
Clinician acceptance may ultimately be most influenced by studies like the Scottish one. "The data is coming and will continue to come, and the best way to convince physicians is with high-quality prospective studies," he added.