Atypical squamous cells: The case for HPV testing
New data, terminology, and guidelines are in, and human papillomavirus testing is emerging as the most efficient and cost-effective triage option.
- Management by immediate colposcopy, repeat cytology, or HPV testing is acceptable for ASC-US, but testing for HPV is preferred when the Pap test is liquid-based.
- The sensitivity of HPV triage for high-grade CIN is essentially equivalent to colposcopy, and reduces the need for colposcopy by half.
- HPV testing is a good option for follow-up after treatment with cryosurgery, loop electrosurgical excision procedure, laser, or cold-knife conization.
What’s the best management strategy for the roughly 2 to 3 million women each yearUpdate on Cervical Disease, for commentary by Thomas C. Wright, Jr, MD, Department of Pathology, College of Physicians and Surgeons of Columbia University.The difference is even more pronounced when the cumulative 2-year detection rate for CIN 2,3 is added in for women referred for HPV-positive ASCUS but not found to have CIN 2,3 at initial colposcopy. That rate rises from 20.1% at initial colposcopy to 26.9% at 2 years.8 Although many experts consider even HPV-positive ASCUS of minimal risk, few would consider a risk of high-grade disease exceeding 1 in 4 to be minimal. In fact, 39% of the total CIN 2,3 cases reported from a routine screening population were detected following triage of ASCUS, and fully 69% were from all equivocal and low-grade Pap diagnoses.9
Risk of cervical intraepithelial neoplasia grade 2 or greater at initial colposcopy
TOTAL RISK FOR ALL ASCUS
Hybrid capture 1 (expanded first–generation test)
Hybrid capture 2
Hybrid capture 2
ALTS = ASCUS/LSIL Triage Study; ASCUS = atypical squamous cells of undetermined significance; HPV = human papillomavirus
Bethesda 3 redefines ASCUS
The third Bethesda System workshop took place in May 2001 with the aim of evaluating and updating earlier terminology.10 It began by eliminating the words “of undetermined significance” from the overall ASCUS category, which is now called simply “atypical squamous cells,” or ASC. Most subcategories of the former ASCUS were eliminated as well. (Note: Within this article, the acronyms ASCUS and ASC-US are both used to describe atypical squamous cells of undetermined significance. The latter acronym reflects usage and guidelines developed after the third Bethesda workshop.)
Now the ASC classification is broken down into 2 distinct groups:
Atypical squamous cells–undetermined significance, or ASC-US. This new subcategory includes cells previously termed “favor reactive” but not relegated by the pathologist to normal, as well as cells previously in the “unqualified” and “favor HPV” or “favor low-grade squamous intraepithelial lesion (LSIL)” subcategories.
Atypical squamous cells–cannot rule out high-grade squamous intraepithelial lesions, or ASC-H. This category includes atypical cells difficult to distinguish from high-grade cells but not definitive for that classification. Women with such Pap tests are at greater risk for high-risk HPV and histologic CIN 2,3 (TABLE 2).
Evidence-based guidelines reflect Bethesda 3 changes. By the time of Bethesda 3, extensive new data on the management of abnormal cytology was available, including but not limited to data from ALTS, making it possible to create evidence-based guidelines on management of abnormal cervical cytology and CIN. These guidelines were developed in 2001 at a consensus conference hosted by the American Society for Colposcopy and Cervical Pathology (ASCCP),11 with input from 29 professional organizations, federal agencies, and national and international health organizations.
The entire set recommendations for all types of abnormal Pap tests were published in the April 24, 2002 issue of the Journal of the American Medical Association, and management recommendations for histologically proven CIN were published in the July 2003 American Journal of Obstetrics and Gynecology and the July 2003 Journal of Lower Genital Tract Disease. The management algorithms for both cytology and histology can be downloaded from ASCCP.org.
Comparison of risk for high-risk HPV and CIN grade 2,3, by Pap results
CIN 2 OR GREATER
Data from Sherman et al29
ASC-US = atypical squamous cells–undetermined significance;
ASC-H = Atypical squamous cells–cannot rule out high-grade squamous intraepithelial lesion; CIN = cervical intraepithelial neoplasia; HPV = human papillomavirus; HSIL = high-grade squamous intraepithelial lesion
All 3 triage options safe, effective
An evidence-based review found all 3 options safe and effective.11 Therefore, management by immediate colposcopy, repeat cytology, or HPV testing is acceptable for ASCUS, but testing for HPV is preferred when the Pap test is liquid-based (FIGURE 1).
Liquid-based cytology (ThinPrep; Cytyc, Boxborough, Mass and SurePath, Raleigh-Durham, NC) has several advantages. For example, residual cells in the fluid can be tested for HPV, eliminating a return visit.
Immediate colposcopy: Low predictive value, high anxiety and expense. Proponents of immediate colposcopy for all women with ASC-US argue that this would theoretically detect all CIN 2,3 and cancer. However, the positive predictive value of this approach will always be extremely low due to the low rate (6.4% to 11.9%) of CIN 2,3 in women with ASCUS.4,6,7 What’s more, the cost and anxiety generated by immediate colposcopy are high.12
2 repeat cytologies: Sensitivity, cost issues. This approach requires at least 2 repeat, optimized (liquid-based) Pap tests to equal the sensitivity of a single HPV test. This, compounded with the high rate of repeat abnormal cytology requiring colposcopic evaluation, means repeat cytology is unlikely to be cost-competitive with HPV testing.4,13
- Cervical cytology as a triage option. Cytology has been a good screening test, but its comparatively low sensitivity (51% to 83%) and poor reproducibility reduces its value as a triage test.13-17 For example, in ALTS, of 1,473 repeat Paps originally read as ASCUS by good clinical pathologists, only 633 were reread as ASCUS when 2-of-3 agreement was obtained in a blinded review by an expert panel of pathologists.16 In other words, 840 (57%) were reread as something other than ASCUS. Most were downgraded to normal.
- The sensitivity of the HPV test in detecting CIN 2,3 was 92.4%. This rate was matched only by 2 repeat Pap tests, provided the threshold for referral to colposcopy was ASCUS or greater.17 At this threshold, 95% of the CIN 2,3 was detected with repeat Pap testing, but only after an average of 8 to 12 months. This contrasts with the immediate reassurance provided by the initial HPV test.
- ALTS did not evaluate repeat conventional Pap smears. Nor do the guidelines differentiate between conventional and liquid-based methods in the number of follow-up Pap tests required for reassurance, despite consensus that the sensitivity of liquid-based cytology is better than that of the conventional “dry slide.”
- Any woman with a repeat Pap result of ASC-US or greater should be referred to colposcopy. Referral at a threshold of LSIL or greater would result in far fewer colposcopies, but has not been shown to be sufficiently sensitive for CIN 2,3.17
HPV testing identifies clear risk. Any objective test that initially indicates which women with ASC-US are at risk for CIN 2,3 and which are not—either now or in the future—should confer a major advantage.
HPV-positive women are clearly at risk, justifying the anxiety and cost of colposcopic referral, while HPV-negative women may be reassured (FIGURE 2). Also, ALTS data showed HPV triage is essentially equivalent to immediate colposcopy in sensitivity for high-grade CIN, while halving colposcopic referrals.17,18
Because low-risk HPV types do not cause CIN 3 or cancer, the HPV test should document only high-risk types.11 The only HPV test approved by the US Food and Drug Administration (Hybrid Capture 2, Digene, Gaithersburg, Md) includes both low- and high-risk HPV panels. For cost savings, the laboratory can be asked to use only the high-risk panel. All positive high-risk HPV cases should be referred to colposcopy.
FIGURE 1 3 triage options for management of ASC-US
FIGURE 2 Management of atypical squamous cells–cannot exclude high–grade squamous intraepithelial lesions (ASC-H)
Some high-grade lesions are still overlooked
The guidelines state that women who undergo immediate colposcopy with negative results or who have a negative initial HPV test should undergo a follow-up Pap test in 12 months. Note that the guidelines do not state that these women can return directly to routine screening. The reason: In some settings, “routine” screening is at 2- or 3-year intervals, and some risk still exists—albeit minimal—for missed CIN 2,3.
For example, 1 of 83 cases of CIN 2,3 were missed by HPV testing in the study by Manos et al.7 In ALTS, that number was 1 in 90,4 and in the study by Cox et al6 it was 1 in 136. Further, colposcopy did not initially detect 25% of the cumulative high-grade lesions detected over 2 years of follow-up in ALTS.17
In contrast, the recommendation for women with 2 repeat normal Pap tests is to return to “routine screening.” This inexplicably departs from the 12-month repeat Pap testing urged for women with negative results on the other 2 triage options, despite a similar risk of missed high-grade disease.
- In my opinion, all 3 scenarios should be managed by repeat Pap testing in 12 months.
Reducing referrals to colposcopy
If all women returned as directed for repeat cytology, more of them would be referred to colposcopy by repeat abnormal Pap tests at the ASC-US threshold than by testing positive for high-risk HPV types. In ALTS, 53% tested positive for high-risk HPV and were referred to colposcopy, compared with 67% who had an abnormal Pap test on the first or second repeat (these women also had 1 or 2 more office visits prior to referral to colposcopy.).
No difference for conventional smears. All the advantages of HPV testing in the triage of women with ASC-US persist when the initial referral Pap test is a conventional smear. The only exception is that HPV testing would require the patient to return for a repeat office visit. An alternative would be co-collecting an HPV-test sample at the time of the primary screening Pap test.
One major health-maintenance organization collects a separate sample from all women when the routine conventional Pap test is obtained using a standard Hybrid Capture 2 HPV test kit. The HPV-testing samples are then held until the results of the Pap smear are reported. For women reported to have ASCUS, the samples are sent to the lab for HPV testing; the remaining samples (approximately 95% in most practices) are discarded as medical waste. The cost of each discarded kit is approximately $1. Modeling has found this approach to be cost-effective.19
Many clinicians are concerned that women referred for the evaluation of HPV-positive ASC-US and found not to have CIN or other manifestations of HPV at colposcopy have a “false-positive” HPV test. However, although there are occasional HPV tests that misclassify a low-risk HPV type as high-risk, actual false-positive tests are very rare.
The 2-year ALTS longitudinal data provide the best information on what to expect when a woman with HPV-positive ASC-US or LSIL is found at colposcopy to have no CIN or to have only CIN 1 that is subsequently managed expectantly.8
The cumulative risk of CIN 2,3 over the 2 years was nearly equivalent for women referred initially for LSIL (27.6%) and for women referred for HPV-positive ASCUS (26.7%), further verifying that management should be similar. Two thirds of the CIN 2,3 was detected at initial colposcopy, and the remaining one third during the postcolposcopy 2-year follow-up.
The risk for subsequent detection of high-grade CIN was nearly identical for all women initially found not to have CIN 2,3 regardless of whether CIN 1 was detected at initial colposcopy, whether the colposcopy was initially completely normal, or whether there were changes that were biopsied and found not to have CIN (risk for CIN 2,3 was 13%, 11.3%, and 11.7% respectively).
Hence, all women referred for evaluation of HPV-positive ASC-US or LSIL and not treated for CIN 2,3 require similar diligent follow-up.
A single HPV test at 12 months detected 92% of all CIN 2,3 found over the 24-month follow-up; 55% tested HPV-positive and were referred to colposcopy.20 Repeat liquid-based cytology at 6 and 12 months referred to colposcopy 63% of women (using a threshold of a repeat Pap test of ASCUS or greater). Cumulative sensitivity of 2 repeat cytologies for CIN 2,3 was slightly less (88%). Combining a repeat Pap test with an HPV test did not increase sensitivity, but did significantly increase referral to colposcopy.