FDA Panel: Roche’s DNA Test Can Replace Pap Smear
COLLEGE PARK, Md. — An FDA advisory committee voted unanimously Wednesday to recommend that the Pap smear be replaced with a human papillomavirus (HPV) test as the first-line standard of care for cancer screening.
The FDA’s Medical Devices Advisory Committee Microbiology Panel agreed by a vote of 13-0 in each of three successive votes that the cobas viral DNA test for HPV — made by Roche Molecular Systems — was safe and effective for cervical cancer screening, and that the benefits of the tests outweighed the risks.
The cobas test currently has approval as a follow-up assessment for women 21 and older who have abnormal Pap tests, and as a co-test with the Pap smear to screen for the high-risk p16 and p18 HPV strains in women 30 to 65. The test comprises genotyping for HPV16 and 18 and pooled assessment of 12 additional high-risk HPV strains.
According to the proposal submitted by Roche, women 25 and older who test positive for HPV16 or 18 would proceed directly to colposcopy for further assessment.
Patients who test negative for HPV16 or 18 but positive for the other high-risk strains would have a Pap test to determine the need for colposcopy. Women who have a completely negative test would be followed at their physician’s discretion.
Panelists did express some concerns about dropping the age at which women should have the test from 30 to 25. The ATHENA study used as the basis for the application found that about 11% of women ages 25 to 29 tested positive for HPV16 or 18 with the cobas test, compared with 7.28% among women 25 to 29 who had cytology alone as their first-line screening. Panel member Paula Hillard, MD, of Stanford University in California, said that would mean more patients in that age group “who will be anxious about potentially having cancer.”
In addition, Hillard expressed concern about off-label use. “I’m concerned that all those women potentially with other high-risk positivity won’t go to Paps next but go [straight] to colposcopy. That’s not what’s proposed here, but what control does FDA have once it’s out there?”
Panelist Kenneth Noller, MD, of the American Board of Obstetrics and Gynecology, in Dallas, agreed that real-world use could differ from the protocol proposed by Roche. “I’ve been watching how people practice; if you’re high-risk HPV positive you’re going to get colposcopy,” he said. “That doesn’t necessarily mean it’s bad — it’s what you do with the colposcopy.”
Noller added that although he was “somewhat biased against dropping the age to 25 before I came here … I find the data presented today somewhat compelling to drop it to 25.”
Panel member Kimberly Hanson, MD, MHS, of the University of Utah and ARUP Laboratories, both in Salt Lake City, agreed. “Now we have the opportunity to identify women earlier, and to me that’s compelling,” she said. “Although colposcopy is invasive and can be anxiety-provoking, it’s really very safe, so I think I’m leaning toward earlier screening.”
The committee members were not concerned about losing other diagnostic benefits of the Pap smear — such as finding trichomonas and abnormal endometrial cells — that would be lost if Pap smears were no longer the first-line screening method. “If this is how you’re diagnosing trichomonas and [other diseases], there are much more effective ways of doing it,” said panel chair Angela Caliendo, MD, PhD, of Rhode Island Hospital in Providence.
The presentations by FDA staff and Roche relied on long-term follow-up data from ATHENA, a 47,208-patient cohort study that was conducted in the U.S.
According to the summary submitted by FDA staff members, “The data show that the proposed primary screening indication for the cobas HPV test detects more women with disease and requires fewer women without disease to go to colposcopy than cytology alone.”
The staff summary included the following performance characteristics of the cobas test versus cytology in the prospective cohort study:
- Sensitivity for ≥cervical intraepithelial neoplasia (CIN3) -58.26% versus 42.63%
- Positive predictive value -12.25% versus 6.47%
- Negative predictive value -0.42% versus 0.59%
- False-positive rate -4.09% versus 6.04%
Benefit-risk analyses favored the HPV DNA test whether expressed in terms of number of cases of high-grade cervical disease per 10,000 women screened or per 100 colposcopy procedures.
The FDA is not bound to follow its advisory committees’ recommendations, but does so in most cases.