Listen to a discussion of this article through our AI-generated podcast.
Beginning with our September 2024 issue, CAP TODAY now offers the option to listen to a discussion of our articles with voices reminiscent of a Siri or Garmin speaker. Since this AI-generated podcast uses technology that is still in its early stages of development, there may be some errors in pronunciation and the tone may sound too upbeat for the material covered.
Charna Albert
September 2024—In HPV primary screening, self-collected vaginal samples enable accurate clinical HPV testing, and neither extended ambient dry storage nor exposure to extreme temperatures influence HPV detection, say the authors of a study published in June (Qi M, et al. J Mol Diagn. 2024;26[6]:487–497).
The aim of their study was to validate the dry swab sample type for molecular HPV testing and provide a road map for other laboratories considering a dry swab collection protocol.
The Food and Drug Administration in May approved two HPV self-collection devices for use in clinic settings. A trial that enrolled this summer will evaluate self-collection device and HPV assay combinations for home collection.
In the study published in June, all samples were evaluated using the Roche Cobas 6800. The Cobas HPV test uses beta-globin DNA as an internal control to confirm the presence and amplification of human DNA. Lack of beta-globin DNA amplification in HPV-negative self-collected samples accurately identified participants who required recollection, the authors report.
“One of the things we saw is that beta-globin is an excellent internal control,” says study coauthor Dina N. Greene, PhD, D(ABCC), clinical associate professor, University of Washington. “And the internal control will fail before the HPV fails. Therefore, the possibility of getting a false-negative result is very low.”
The Cobas test is optimized such that beta-globin amplification is “just a little bit less stable than HPV,” Dr. Greene says. “So you may need to recollect in up to 10 percent of cases, because most people are going to be HPV negative.”
Dry swabs are the specimen of choice for collection that occurs outside the clinic because they simplify the kit design, user experience, and transport, but validating the dry swab sample type poses a difficulty. “I can’t use the same exact sample and put it under different conditions. That’s the biggest challenge and it’s also a limitation,” Dr. Greene says. “If I have a serum sample, I can aliquot it into however many [samples] I want” and put it under different conditions. “With a dry swab, you get one condition per swab.” Even if the same person collects two dry swabs in concert, “you’re going to have variability,” she says.
To analyze swab diagnostic variability, vaginal samples were obtained at the LetsGetChecked laboratory in California from 15 volunteers who sequentially self-collected five swabs each. “The early part of the paper is trying to show that if I collected five sequential swabs, what would the variability be between swab one, swab two, swab three, swab four, and swab five? So I can start, as an analytical chemist, to understand . . . what is the variability between collection” and what variability is due to preanalytical variables, Dr. Greene says.
Buffer was added to the swabs within 24 hours of collection. Cycle threshold values for beta-globin and/or HPV for the last and first swabs were used to calculate the change in cycle threshold between the sequential collections. The average coefficient of variation for the beta-globin internal control Ct per individual across the five collections was 2.7 percent (range, 0.4–6.9 percent). The average difference between beta-globin Ct values for the fifth swab collected and the first swab collected was - 0.1 (SD=1.5; range, -2.9 to 4.0). There was no statistically significant difference in beta-globin amplification between the first and last swabs.
For the subset of specimens testing positive for HPV (n=1 HPV16; n=2 HPV other), there was 100 percent qualitative concordance across the five collections. “These data,” Dr. Greene and her coauthors write, “suggest that there is minimal intrapatient diagnostic variability for self-collected vaginal swabs, and that for analytical studies, multiple swabs collected from the same participant will have minimal baseline differences.”
To compare provider-collected and home-based self-collected samples for HPV detection, paired self-collected vaginal/provider-collected cervical samples (n=144) were obtained at Kaiser Permanente Northern California from volunteers at their follow-up colposcopy visits and evaluated for HPV at Kaiser Permanente Northern California Regional Reference Laboratory. For all other studies the authors performed, there was no enrichment for known positive samples.
The self-collected dry vaginal swabs were suspended in buffer on receipt in the laboratory. The provider-collected cervical samples were placed in ThinPrep media immediately after collection. Amplification of HPV and the beta-globin internal control was equivalent between the two collections (P=0.454 for HPV and P=0.835 for beta-globin, paired two-sided t-test).
The results were highly concordant for HPV detection (total agreement, 90.3 percent; positive percentage agreement, 84.2 percent). For 127 participants (88 percent), test result interpretations were fully concordant for HPV detection between collection modalities. Among the discordant samples, cases in which the self-collected sample detected HPV and the clinician-collected sample was negative outnumbered cases in which the converse was true. Of the 144 participants, HPV was detected 85 times in participants using clinician-collected samples and 90 times in participants using self-collected samples.
“This could be a dilutional effect since the buffer volume differs by 10 times between the sample types,” Dr. Greene says, “or it could be the concentration of cellular shedding within the two anatomical locations. But ultimately we don’t know why the sensitivity is more robust in self-collected samples.”
Of the testing of samples from the colposcopy clinic, she says: “We wanted to enrich for positive HPV. Sometimes they’re not positive, and we didn’t have 100 percent positivity in there.” It was still necessary for those patients to get colposcopies, she says, because of the prior test showing they had been exposed to high-risk HPV.
To test if the number of days the self-collected vaginal swab remains dry influences DNA amplification, Dr. Greene and coauthors performed a subanalysis in which they evaluated relative amplification of beta-globin (n=115) and HPV (n=71) between provider- and self-collected samples as a function of the number of days the swab remained dry.
The average amount of time the swabs remained dry was 13.7 days. There was no relationship between the number of storage days and change in HPV Ct (R=0.0012; P=0.991), though there was a significant decrease in beta-globin amplification over time (R=0.3124; P=0.0005).
“I’m seeing the same results, the same delta Ct between provider and the self-collect, no matter how many days that swab sat,” Dr. Greene says.
In a similar study of ambient stability, 68 participants self-collected two vaginal swabs. The baseline sample was suspended in buffer immediately and evaluated for HPV within 24 hours of collection. The second swab remained dry for four to 41 days at an uncontrolled ambient temperature indicative of the indoor laboratory environment.
Eleven HPV-positive samples were present in the baseline (eight HPV other, one HPV16, and two HPV18). After storage at room temperature, 10 samples remained HPV positive, and the Ct didn’t differ significantly from baseline readings. For the one sample in which HPV other was not detected after storage, the baseline HPV other Ct was near the limit of detection (31.0) and the swab was stored for 28 days. There was no relationship between the number of storage days and change in HPV Ct. Dr. Greene and coauthors found that HPV detection is unlikely to be compromised when the dry swab is stored at ambient temperatures for at least 30 days and that beta-globin is a sensitive indicator of sample integrity.
They also assessed sample stability (n=78 participants collecting three swabs each) under controlled environmental conditions intended to simulate summer and winter temperature extremes. After exposure to the summer challenge, 13 samples were invalid. Comparing paired beta-globin Ct before and after the summer challenge was statistically significant. Beta-globin Ct was also influenced by the number of days the swab remained dry. Fourteen HPV infections were detected at baseline. After the summer challenge, 13 remained detectable, and an additional infection was detected that was not positive for the baseline sample.
A separate set of specimens was challenged with extreme winter temperatures. Comparing paired beta-globin Ct before and after winter thermal cycling was statistically significant. Beta-globin Ct was also influenced by the number of days the swab remained dry. Fourteen HPV infections were detected at baseline. After the winter challenge, 13 remained detectable; an additional two infections were detected that were not positive for the baseline sample. The authors concluded that peak summer and winter conditions can lead to degradation for the human beta-globin internal control, but the HPV DNA target is largely unaffected.
In a final challenge, they performed an orthogonal assessment of cellularity among invalid samples. A subset of self-collected vaginal swab samples were evaluated for the presence of RNase P, a common housekeeping gene, to determine whether there was a difference in cellular recovery between samples that did and did not amplify the beta-globin target. The purpose of the experiment, Dr. Greene explains, was to use RNase P to determine if the samples with invalid beta-globin were truly absent of any human DNA. If the HPV assay is designed such that beta-globin fails early, she says, the limit of quantitation on beta-globin is likely arbitrarily high. “So beta-globin is not there, until, let’s say, 500 copies. And everything below 500 copies is a black box in the instrument. I can’t see that.”
“This is not for me to do in production,” she notes. “I’m not going to take invalid samples and reflex them. I just wanted to see if it was truly a limit of quantitation thing, and that’s what we saw. There is DNA here. There’s not HPV DNA here. There’s probably beta-globin DNA here, but it’s at a concentration that’s lower than what the assay is telling me is positive.”
Some of the samples with invalid results had no RNase P content, which might reflect incorrect self-collection, she says. “The ones that were negative beta-globin, negative HPV had the biggest risk of not amplifying RNase P.” The invalid results could be owing to an inhibitor like soap or lubricant, she says. “But my intuition as a scientist says that’s not what’s going on here. There are just not enough cells to amplify.” There was no significant difference in RNase P Ct between samples that were HPV negative and HPV positive without beta-globin amplification. The authors concluded that lack of amplification for beta-globin from self-collected specimens was consistent with low cellularity.
This study can serve as a template to validate for self-collection any HPV assay with HPV primary screening clearance, Dr. Greene says. “I think labs can do it similarly to the way we did it.”
When it comes to preventing user error, Dr. Greene is skeptical that self-collection in the health care setting, as approved now, will be more effective than home collection.
“If we can have self-collection services with a customer service line people can call to get help if they need it, that is a more practical solution than thinking our lab assistants [in clinics] are going to be able to provide any type of assistance besides offering an immediate solution when a patient says, ‘I messed this up; can I have another one?’” she says.
Samples for fecal immunochemical testing are always collected in the home, she notes. “And that’s not FDA approved for home collection. All the clinical trials were done with home collection. But if you look at the stability and the analytical challenges, it would not pass.”
Then, too, she says, the public health benefits of collecting vaginal samples in the home outweigh the analytical differences. “Even if we knock off five percent sensitivity, we’re still going to detect way more than if we hadn’t offered the solution.” Clinic collection may be superior in sensitivity—“with FIT that is true,” she says—and that can be conveyed to patients if necessary. “We can tell people, ‘It might be better if you collect it in the clinic, but if you’re not going to come to the clinic, we’ll send you one at home. It might be a little worse, but if it’s positive, at least you’ll know.’”
Until recently, Dr. Greene was associate laboratory director at the LetsGetChecked laboratory in California. She left that position to consult as chief scientific officer for a virtual perimenopausal clinic, for which she’s building the laboratory services for perimenopause care. “I’m on the home collection train,” she says. “There are so many things we could be doing from home that we choose not to because we’re scared of breaking dogma.”
Home collection for HPV, she notes, does not equal direct-to-consumer testing. “The only way this can be done from home is in concert with a health care organization that is monitoring the frequency of screening,” she says. “You want to make sure people are getting screened at the right interval.” It’s unlike screening for sexually transmitted infections. “STI screening should be done at a certain limit depending on sexual activity, and HPV should be done at a certain frequency depending on your age. It is not behavior based.”
Remote collection for STI testing is another area of focus for Dr. Greene. The FDA in 2023 granted de novo marketing authorization to a commercial at-home collection testing option for detection of Chlamydia trachomatis and Neisseria gonorrhoeae (LetsGetChecked Simple 2), but the test is restricted to urine and vaginal swabs. About 70 percent of CT and NG infections may go undiagnosed if urogenital only testing is performed because the majority of rectal and throat infections are asymptomatic, Dr. Greene and coauthors report in a study published in July (Hockman BE, et al. J Clin Microbiol. 2024;62[7]:e00311–e00324). In that study, they performed a preanalytical validation of self-collected rectal and throat swabs and found that self-collection is an appropriate method of sample acquisition for detecting CT/NG infections. Self-collected samples were highly concordant to provider-collected samples over a large number of samples at each collection site challenged, they write, adding: “Percent positive and overall agreements were greater than 95% for all swab/organism combinations” relative to provider-collected samples.
“Oral and anal self-collection is a huge thing,” Dr. Greene says, but there are “hypothetical barriers” that some labs “have a hard time adjusting to.” Hence the hope that the study she and her colleagues did on HPV will serve as a validation road map.
Implementation hurdles are real, however. She and her HPV study coauthors cite a publication that reported laboratory readiness to accept self-collected specimens as an unexpected barrier in Australia’s implementation of primary HPV screening (Bavor C, et al. BMC Health Serv Res. 2023;23[1]:1073).
“It is operationally difficult because you’re getting one sample in one package, so automating that in any high-throughput way is difficult,” Dr. Greene says. “Not impossible. You have to standardize a lot of different things to make it happen—standardize a kit that can be opened in an automated way, and then some way of getting the tube out and getting it logged in,” as well as tracking when the sample was sent. “All samples have a bit of manual intervention usually, but minimizing that is difficult,” she says.
“We do all this, but it’s not optimized for home collection yet.”
Some will raise questions about follow-up testing, she notes. But when it comes to “investing in those who can’t make it to the doctor’s office,” the matter of follow-up testing shouldn’t concern the laboratory, she argues. It is the laboratory’s business to “make testing accessible,” regardless of what occurs after.
“HPV and cervical cancer screening could be done from home as seamlessly as fecal immunochemical testing is, if people would give it the opportunity to be done that way,” she says. “I always bring it back to women’s health. Women’s health is thought to be this delicate thing. But we end up marginalizing women while thinking we’re protecting them.”
Charna Albert is CAP TODAY associate contributing editor.