Feature Story

Pap and HPV DNA testing: Is there room for both?

December 2002
Karen Southwick

One of the first studies to compare DNA testing with Pap testing as a primary screen for cervical abnormalities in a large U.S. population reveals that while both have their deficiencies, together they have strong negative predictive value.

The study, "Evaluation of human papillomavirus testing in primary screening for cervical abnormalities," appeared in the Oct. 9 issue of JAMA (1749–1757). Seven authors from the University of Washington, Seattle, examined the cases of 4,075 women who attended Planned Parenthood clinics in the state of Washington between December 1997 and October 2000. The women were a mean age of 25 and had a lifetime median of six sex partners.

The women were screened simultaneously using thin-layer Pap and HPV DNA testing by a PCR-based method and by liquid-based RNA-DNA hybridization capture with signal amplification. Women who were positive for high-risk HPV or had Pap test findings of ASCUS (atypical squamous cells of undetermined significance) or greater were referred for colposcopy and biopsy. In addition, some women with negative screening test results were randomly referred for colposcopy.

The estimated prevalence of CIN 3 or higher was 3.2 percent. Compared with referral for colposcopy of all women with ASCUS or higher, signal amplification testing of women with ASCUS or higher was about as sensitive (61.3 percent versus 60.3 percent) and significantly more specific (82.4 percent versus 88.9 percent, respectively). All tests were more specific and less sensitive in women older than 30.

In other findings, the prevalence of HPV increased in women aged 18 to 24 and then declined sharply. The prevalence of low-grade squamous intraepithelial lesion, or LSIL, declined steadily with increasing age, while the prevalence of ASCUS averaged about 10 percent across all age groups.

While the earlier ALTS trial concluded that triage with HPV after an ASCUS Pap result made sense, the JAMA study considers whether HPV can be used as a primary screen. "The strengths and weaknesses of cytology are well documented," says Nancy B. Kiviat, MD, one of the authors of the study and professor of pathology at the University of Washington. Dr. Kiviat is also director of pathology at Harborview Medical Center, an affiliated teaching hospital.

Previous studies of HPV as a primary screen examined populations in China (Belinson) and Costa Rica (Schiffman). "This article confirms some of the findings with a U.S. population and adds a word of caution that HPV testing is better in certain groups than in others," says Dina Mody, MD, chair of the CAP Cytopathology Committee and director of the cytology laboratory at Methodist Hospital, Houston.

The JAMA study’s complex conclusion is that testing for HPV has higher sensitivity but lower specificity than thin-layer Pap screening, so it might make sense to substitute HPV in some settings where screening intervals are long or haphazard. The sensitivity of thin-layer Pap for women with CIN 3 or higher was only 61.3 percent, compared with 88.2 percent for HPV testing by PCR and 90.8 percent by signal amplification.

"What we saw is that HPV screening is sensitive but lacks specificity," Dr. Kiviat says. Using HPV as a primary screen, especially in women under 30 in whom prevalence of the virus is high, would result in too many referrals for invasive diagnostic tests. However, "this study shows that HPV should play a role in cervical cancer control," she says.

Coauthor Constance Mao, MD, an obstetrician-gynecologist who runs the colposcopy clinic at Harborview Medical Center, describes the Planned Parenthood group used in the research as a young, high-risk population with "high rates of HPV infection and cervical pathology." That meant high rates of referral for colposcopy, although many of the women failed to show up for the additional test. Those who did tended to have more severe Pap smear findings and positive HPV results.

R. Marshall Austin, MD, PhD, medical director and director of cytopathology and gynecologic pathology services for Coastal Pathology Laboratories, Charleston, SC, and a member of the CAP Cytopathology Committee, asked why the JAMA study’s results revealed lower sensitivity on ASCUS or worse with the liquid Pap and the hybrid capture tool than did the earlier overseas studies. "I thought that was a significant question mark for the study," he says. "If a study’s findings are significantly different from other reports, you have to ask why."

Coauthor Laura Koutsky, PhD, professor of epidemiology at the University of Washington, told CAPTODAYthat the ALTS trial showed a similar percentage of CIN 3 among women with ASCUS at screening. Thus, "our cytology and histology readings appear to be similar," she says. The populations in China and Costa Rica were older than the Planned Parenthood cohort, she adds, and had probably not been routinely screened in the past. "I believe that our study was picking up small foci," Dr. Koutsky says. With the overseas study, "when you have an unscreened population who are older you’ve got larger lesions that are more likely to be picked up on a Pap."

Role for HPV screening
Experts agree that HPV testing will be used more and more in cervical cancer screening. But several issues remain: first, defining guidelines for when HPV can be used as a primary screen; second, educating clinicians and payers and obtaining reimbursement; and third, dealing with women who are negative on their Pap test but positive on HPV, which is often the case with younger women. Says Dr. Mao, "The downside with HPV testing as a primary screen is you would be doing a lot more colposcopy [as a followup] in women who turn out not to have disease." Some experts have suggested doing followup HPV tests at six or 12 months to see if the virus clears and referring for colposcopy and biopsy if it doesn’t.

Another part of the JAMA study, not yet published, focuses on cost-effectiveness. So far, Dr. Mao says, "it appears that primary HPV screening is currently significantly more expensive than Pap-only screening when comparing dollars per year of life saved. That is mainly due to the poor specificity of the test." HPV testing may look more favorable in the future if it becomes more specific and if its cost declines with volume.

"HPV as a triage for ASCUS Pap results is becoming accepted," Dr. Mao says, but she predicts it will take many years, at least in this country, before HPV is used as a primary screen. At Harborview Medical Center, "we’re slowly getting the word out to do HPV as a triage if the Pap shows ASCUS," she says, adding that clinicians generally appreciate pathologists’ input on the new guidelines. "I encourage all of my trainees to review the slides with the pathologists to really understand a Pap smear, especially when results do not fit the clinical picture."

In Dr. Mody’s lab at Methodist Hospital, the forms have been changed so clinicians can check off HPV reflex testing, which some do and some don’t. The charity care population is routinely screened for HPV because the hospital is willing to pay for it. However, "we are not doing HPV as a primary screen," Dr. Mody says. "The prevalence is very high in women under 30, and many women have a transient infection."

Even so, the recommendation for women over and under 30 is now the same: With an ASCUS Pap result, get high-risk HPV typing done. If that’s positive, go to colposcopy. Or you can repeat the Pap test every six months.

"One of the things you could fine-tune is when to go straight to HPV as a primary screen," says Dr. Mody, "but we are not there yet. We may or may not go that way." What may force such a move is the national shortage of cytotechnologists. "Most of the doctors are sending out their Paps to two large national labs," she says, and that capacity is becoming strained.

The shortage, she adds, is even more acute in Third World countries, where HPV testing could be valuable in areas with a lack of trained cytotechnologists. "Pap smears are very labor-intensive. With HPV, you can collect the material and ship it to one central lab. It’s not like the Pap smear, where you have a human looking for the needle in the haystack," Dr. Mody says.

Double-negative
While either Pap or HPV alone could yield false-negatives, their negative predictive value when combined is nearly 100 percent, experts agree. In the JAMA study, of 2,631 women who were Pap- and HPV-negative, 202 were randomly referred for colposcopy and none of those women had CIN 3 or higher.

"Neither test is perfect," says Diane D. Davey, MD, advisor to the CAP Cytopathology Committee and professor of pathology and laboratory medicine and lab director of the cytopathology and bone marrow labs at the University of Kentucky Chandler Medical Center, Lexington. "But your chance of having a high-grade lesion if neither is positive is close to zero," she says.

Even though HPV testing is the consensus guideline for triage, all clinicians and payers are not yet convinced. And it will be difficult to get beyond that reluctance to a primary screen. "For us to offer HPV as a screen doesn’t make sense yet," Dr. Davey says. "We’re just getting payers to pay for HPV as a triage for ASCUS." If a woman wants an HPV test as a primary screen, she’ll probably have to pay for it herself.

With a negative Pap test and no history of HPV, "you’d have difficulty getting payment from a third party," she adds. "We really do need better guidelines because people may be ordering HPV too much or not enough."

The JAMA study is a reminder that the nearly universal perception that the Pap test is perfect "is erroneous," Dr. Austin cautions. "The message that we can’t catch everything was never delivered since the advent of screening in the 1950s."

For that reason, he favors routinely using cytology and HPV together. "All of my reports to my doctors say that on the negative Paps, if you want to approach 100 percent sensitivity, the only way to accomplish that is to add HPV as a co-test," Dr. Austin says. He has few takers. "It’s a brand-new concept, and they would have to pay for it out of pocket."

HPV testing may also help detect glandular precancerous lesions that are remote from the cervix, Dr. Austin adds. For reasons that may be related to exposure to HPV at an earlier age, endocervical adenocarcinoma has increased from five percent of surgical cancer in the 1950s to more than 33 percent now. Although the data are "very preliminary," they indicate that women with glandular lesions could be HPV-positive and Pap-negative. "Then you go reread the Pap," he says. In addition, "if the virus persists over time, that’s a red flag."

Although a double test obviously costs more, Dr. Austin points out that the screening interval could eventually be increased to up to three years for women who are negative on Pap and HPV, which should offset the additional testing costs. He also believes that annual screening can be stopped in double-negative women of a certain age, perhaps 65, who have had no incidence of disease.

Additional tests are needed, but advances are often impeded by cost. Payers prefer to rely on existing data. "There’s a great deal of economic pressure driving these decisions," he says. Decisions have often relied on extrapolating data from older studies "that have not taken into account such things as the rising rates of adenocarcinoma."

Some of the needed research would look at the results when the screening interval is increased on the front end for double-negative women. "We need to do long-term followups," Dr. Austin says, "to see if these women develop cancers at any appreciable rate."

Karen Southwick is a writer in San Francisco.