OUP user menu

Follow-up Outcomes in a Large Cohort of Patients With Human Papillomavirus–Negative ASC-H Cervical Screening Test Results

David Cohen MD, R. Marshall Austin MD, PhD, Christopher Gilbert MD, Richard Freij MD, Chengquan Zhao MD
DOI: http://dx.doi.org/10.1309/AJCPYK60BZRNNAHQ 517-523 First published online: 1 October 2012


Limited follow-up data are available on patients with cervical cytology results of atypical squamous cells, cannot exclude a high-grade intraepithelial lesion (ASC-H), who test negative for high-risk human papillomavirus (hrHPV). Between June 2005 and December 2010, 885 patients were identified with ThinPrep results of Hybrid Capture 2 (HC2) hrHPV-negative cervical ASC-H liquid-based cytology and follow-up histopathology or cytology results extending to September 2011. Of the 885 patients with available follow-up results, 549 (62.0%) had at least 1 histopathologic result during the entire follow-up period, whereas 336 (38.0%) had only cytologic follow-up documented. In an average follow-up period of 29 months, 14 (1.6%) of 885 patients with HPV-negative ASC-H results showed evidence of high-grade cervical intraepithelial neoplasia (CIN2/3). No cases of invasive cervical cancer were diagnosed. Four of 14 patients with HPV-negative ASC-H results with follow-up diagnoses of CIN2/3 had a history of earlier CIN2/3 diagnoses before HPV-negative ASC-H results. Follow-up of patients with HPV-negative ASC-H results using methods specified in this study yielded low rates of detectible CIN2/3 and no diagnoses of cervical cancer. Triage of study patients with HPV-negative ASC-H results to routine HPV and cytology cotesting at 1 year was a safe follow-up option.

Key Words
  • Cervical cytology
  • Atypical squamous cells, cannot exclude HSIL
  • ASC-H
  • Human papillomavirus
  • Follow-up results

Atypical squamous cells, cannot exclude high-grade intraepithelial lesion (ASC-H), is an uncommon cytology interpretation with an estimated incidence of 0.27% to 0.56% among the approximately 65 million Papanicolaou (Pap) tests performed annually in the United States.1,2 A relatively new classification, ASC-H is a subset of abnormal cytomorphologic changes formally introduced in the United States in the 2001 Bethesda System (TBS2001).3 Early data on the association of similar cytologic changes with histopathologic high-grade cervical dysplasia came from several small studies reported before TBS2001.4,5 In the influential atypical squamous cells of undetermined significance (ASCUS) low-grade squamous intraepithelial lesions (LSIL) Triage Study (ALTS), 110 women with retrospective panel interpretations of ASC-H were reported to have an 84% rate of high-risk (hr) HPV-positive results and a 58.7% rate for underlying histopathologic cervical intraepithelial neoplasia (CIN) 2 or worse (CIN2+) findings.6 This retrospective analysis of ALTS data, involving a highly selected young patient population and an internal consensus pathology review panel, served largely as the basis for the 2006 American Society for Colposcopy and Cervical Pathology (ASCCP) Guidelines suggesting universal colposcopic referral for women with a Pap interpretation of ASC-H without routine HPV testing.7 Since 2001, however, the methods used in the United States for routine cervical screening have changed significantly. Liquid-based cytology (LBC), reflex hrHPV testing after ASC-US Pap results, computer- assisted screening, and adjunctive hrHPV cotesting in women aged 30 years and older have all become much more common. Substantial additional data have accumulated on both the prevalence of hrHPV-positive and hrHPV-negative results in women with ASC-H Pap test results in different age groups and the likelihood of underlying histopathologic CIN2+ lesions.8,9 Although most studies have concluded that a negative hrHPV test result accompanying ASC-H offers high negative predictive value and a reasonable basis for considering triage to routine periodic screening, a few studies have disagreed.10,11 Women with Pap interpretations of ASC-H and concurrent hrHPV negative test results remain an incompletely studied population with little published follow-up beyond short-term colposcopic findings. Data confirming a sufficiently high negative predictive value for hrHPV-negative ASC-H results would support adjusting initial recommendations for colposcopic referral of all women with ASC-H Pap results. This issue is clearly important in balancing the benefits and risks of colposcopic examination, because colposcopy is an invasive procedure with associated costs and risks that produce a high level of anxiety in many women.12 In this study, we report follow-up outcomes for the largest number of women studied to date with hrHPV-negative ASC-H test results.

Materials and Methods

A retrospective study was designed and initiated after obtaining approval from the institutional review board at the University of Pittsburgh Medical Center (UPMC), Pittsburgh, PA. A computer-based search was carried out on our CoPath laboratory information system to retrieve LBC Pap tests with ASC-H interpretations and with concomitant hrHPV testing over a 67-month span, between June 2005 and December 2010. All specimens were processed in the pathology laboratory of Magee-Womens Hospital (MWH) of UPMC and reported using current TBS2001 terminology.3 Thin-Prep Pap tests (TPPT; Hologic, Marlborough, MA) were prepared according to the manufacturer's specifications from Preserv-Cyt samples using an automated processor (ThinPrep 3000). Staining of slides was done on a Sakura Tissue Tek Automated Slide Stainer (Sakura Finetek USA, Torrance, CA) according to a Food and Drug Administration (FDA)–approved manufacturer's protocol. Location-guided computer-assisted screening of TPPT slides was accomplished using the Thin- Prep Imaging System.13 The MWH cytopathology laboratory is a large, subspecialized academic hospital laboratory that usually reports more than 100,000 Pap tests per year from a large, integrated hospital health system that serves a metropolitan area with a significantly older population than the national average.14 UPMC is a large, integrated private health system in which Pap tests are collected by a highly diverse group of clinical providers that includes gynecologists, family physicians, internists, nurse practitioners, physician assistants, and house staff trainees. Our laboratory reporting profile has been documented in numerous recent publications.15-28 hrHPV DNA testing was ordered by clinicians according to several ordering options as follows: reflex testing triggered by ASC finding on Pap test; cotesting with Pap tests in women aged 30 years and older; and cotesting regardless of either age or Pap test results. hrHPV DNA detection in TPPT Preserv-Cyt vial fluid was performed using the FDA-approved Hybrid Capture 2 (HC2) assay method (Qiagen, Gaithersburg, MD), which tests for high-risk and intermediate-risk HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68. The results of hrHPV DNA testing were either positive or negative.

At MWH, all histopathologic results of CIN2/3 are routinely confirmed and cosigned by a second pathologist.29 Immunohistochemical stains for p16 and Ki-67 are used electively and liberally by staff pathologists to increase the reliability of a CIN2/3 diagnosis.30 Histopathologic follow-up included endocervical curettage, cervical biopsy, cervical conization using a loop electrosurgical excision procedure, or cold-knife conization. In this report, CIN terminology is used to refer solely to histopathology results, whereas high-grade squamous intraepithelial lesion (HSIL) and low-grade squamous intraepithelial lesion (LSIL) terminology is used to refer solely to Pap test results. Time elapsed from an ASC-H test result until colposcopic examination, cervical biopsy, and follow-up procedures were abstracted from record reviews extending through September 2011. For patients undergoing 2 or more procedures during a single follow-up period, only the most abnormal histopathologic result was recorded. Time elapsed from ASC-H test results until a further cytologic examination was also recorded. Any follow-up hrHPV testing, either cotesting or stand-alone HPV test requests ordered by a clinician, was recorded. The slides of cases with CIN2/3 or HSIL results were reviewed and confirmed. The results of history (4 years) for cases with CIN 2/3 or HSIL follow-up findings were checked and documented.


During the 67-month study period from June 1, 2005, to December 31, 2010, a total of 604,035 Pap tests were reported. During this time, ASC-H was interpreted in 3,155 cases (0.52%). Of these 3,155 samples, 2,305 (73.1%) had an hrHPV cotest result, including 1,251 hrHPV-positive (54.3%) results and 1,054 hrHPV-negative (45.7%) results. Thirty-nine patients showed more than 1 HPV-negative ASC-H test result. The follow-up study included 1,015 individual patients with HPV-negative ASC-H test results. The mean age of this study group was 37.4 years, and the median age was 36 years with a range from 17 to 89 years. Of these 1,015 patients, 130 (12.8%) had no documented follow-up results in our database. During a follow-up period extending to September 2011, 885 (87.2%) of the 1,015 patients had at least 1 follow-up Pap test or histopathologic result in our CoPath files. The mean followup period was 29 months (range, 1-73 months). Using the date of the first documented Pap test or histopathologic procedure after the initial HPV-negative ASC-H result, the average time to initial follow-up was 6 months (range, 1-61 months). Of the 885 patients with follow-up results, 549 (62.0%) had at least 1 histopathologic result during the follow-up period, whereas 336 (38.0%) received only cytologic follow-up. The follow-up results are shown in Table 1. A first follow-up evaluation consisting of histopathologic examination was conducted for 518 patients (58.5%), with an average follow-up interval of 2.3 months (1-61 months).

View this table:
Table 1

The follow-up outcomes with histopathology (CIN) and cytology (SIL) are shown separately in Table 2. CIN1/LSIL or worse lesions (CIN1/LSIL+) were documented in 215 (24.3%) of 885 patients for whom follow-up data were available. High-grade intraepithelial lesions (CIN2/3/HSIL) were documented in 14 patients (1.6%). Of the 14 cases with CIN2/3/HSIL findings, one was diagnosed with cytologic examination only (HSIL) with no histopathologic follow-up results available in our laboratory information system. No cases of glandular neoplasia or invasive carcinoma were diagnosed during the follow-up period. When women with HPV-negative ASC-H results were divided into 2 age groups (<50 years and ≥50 years), the incidence of CIN2/3/HSIL was not found to be significantly different (P = .976). No woman older than 60 years with HPV-negative ASC-H results had follow-up findings of CIN2/3/HSIL.

View this table:
Table 2

During the follow-up period, 452 women had at least 1 follow-up HPV test and 172 women had 2 or more follow-up HPV results, with a mean of 1.5 tests (range 1-6). The HPV follow-up test results are shown in Table 3. Forty-three (9.5%) of 452 patients had at least 1 HPV-positive result during the follow-up period; 33 (7.3%) of these 452 patients had an initial HPV-positive follow-up result Table 4. Of women with follow-up HPV testing, 10 were HPV negative for a period before converting to HPV positive. The mean intervals to first follow-up HPV result or first follow-up positive HPV result were 16 months (range, 1-64 months) and 22 months (range, 1-62 months), respectively.

View this table:
Table 3
View this table:
Table 4

Table 5 details the 14 patients who had follow-up CIN 2/3/HSIL findings after HPV-negative ASC-H results. The average interval from HPV-negative ASC-H to follow-up CIN2/3/HSIL findings was 13 months (1-42 months). Of these 14 patients, 4 had previous CIN 2/3 diagnoses by 15 to 38 months (mean, 22.5 months) and 3 of them had a loop electrosurgical excision procedure before HPV-negative ASC-H results; 3 had CIN1/LSIL findings within the prior 4 years. Three patients had earlier HPV tests 11 to 13 months before HPV-negative ASC-H results; 1 was HPV-positive and 2 were HPV-negative. Of the 14 patients with CIN 2/3/HSIL findings after HPV-negative ASC-H results, 7 had additional follow-up HPV testing; of these 7, 4 (57%) tested HPV-negative and 3 tested HPV-positive (43%) during the later follow-up period.

View this table:
Table 5


To the best of our knowledge, ours is the largest follow-up study of patients with hrHPV-negative ASC-H cervical screening test results. The ASC-H category is a relatively new classification, first seen in TBS2001 and now accounting for an estimated 0.27% to 0.56% of reported United States Pap test findings.1 This uncommon equivocal interpretation is thought to identify a population with significantly higher rates of histopathologic CIN2/3+ than detectable in patients with ASC-US results.31 Over the past decade, however, HPV testing has significantly changed the way clinicians treat patients with equivocal ASC findings. When the HC2 HPV test was first approved by the FDA for ASCUS reflex testing, the category ASCUS included cytologic changes that were later separated into the new category of ASC-H. Therefore, many patients with what we now may interpret as ASC-H underwent ASCUS reflex HPV testing and triage in the past.5 Also, because the distinction between ASC-H and ASC-US is not highly reproducible,32,33 patients with cytologic changes potentially interpreted as ASC-H by one observer might simply be classified as ASC-US by another observer and therefore receive HPV testing and triage. The decision by the 2006 ASCCP consensus conference to not recommend routine HPV testing in patients with ASC-H Pap test results largely stemmed from the high hrHPV-positive rate (84%) reported by ALTS investigators (who rendered ASC-H interpretations only after consensus panel reviews) rather than from data-driven concerns about the negative predictive value of HPV test results in this subset of patients. In contrast, most studies in which ASC-H interpretations were not based on consensus panel reviews reported that patients with ASC-H cytology had HPV-positive rates around 50% to 60%,8,9 similar to the HPV-positive rate reported with ASCUS in ALTS,34 a rate that was judged sufficiently low to accommodate reflex HPV triage testing. The 2006 ASCCP follow-up guidelines already recommend HPV testing alone at 12 months as an acceptable follow-up option when colposcopic referral of a patient with an ASC-H result does not identify CIN2/3.7

HPV testing is now routinely ordered for many patients with ASC-H Pap test results, either as a routine cotest in women aged 30 years and older or as an elective reflex test after ASC-US or ASC-H findings.8-11 The purpose of our study was to document the histopathologic and cytologic follow-up findings for a large group of patients with HPV-negative ASC-H results. The negative predictive value for hrHPV in most studies, based on short-term colposcopic and histopathologic follow-up examinations, has been in the range of 95% to 100%.9 Efforts to provide a database for informed consensus on the subject have been limited by the small number of patients with HPV-negative ASC-H findings reported in most studies and the short follow-up periods. Our study documents 1,015 women with HPV-negative ASC-H results, of whom 885 (87.2%) had follow-up and 549 had follow-up histopathologic examination. Over a 5-year period and a mean follow-up period of 29 months, only 14 (1.6%) of 885 patients developed evidence of high-grade intraepithelial neoplasia. No cases of invasive cervical cancer were diagnosed. For comparison, a summary of previously published reports documented that 355 (38.8%) of 915 patients with hrHPV-positive ASC-H results had histopathologic CIN2/3+ diagnoses established during short-term diagnostic follow-up.8

In a large study by Kaiser Permanente in Portland, OR, conventional Pap smears and HC2 HPV cotesting were routinely used in almost 1 million women aged 30 years and older. The study reported 26 CIN2, 14 CIN3, and 4 invasive cervical cancer diagnoses in 44 (10.6%) of 414 patients with HPV-negative ASC-H results who were followed up for at least 5 years. This study, however, did not record abnormal cytologic or HPV test results in the interval preceding CIN2 and more severe lesion (CIN2+) diagnoses after initial HPVnegative ASC-H findings. Because CIN2+ may arise within months of a new hrHPV infection,35,36 the follow-up findings of the Kaiser Permanente study may reflect in part the effect of interval HPV infections during the long (5-6 year) follow-up period. Also, the Kaiser Permanante study was conducted in a large nonacademic healthcare setting before widespread availability and routine use of P16 immunohistochemical staining to refine CIN2+ diagnoses.30 A recent report from the Addressing the Need for Advanced Diagnostics (ATHENA) trial, for example, documented that more than half of CIN2+ diagnoses rendered by pathologist investigators blinded to HPV-negative results on the COBAS HPV test (Roche Molecular Diagnostics, Pleasanton, CA) turned out to be histopathologic misclassifications, when adjudicated with P16 immunohistochemistry.37

Compared with other studies, our study has a number of unique features. Incorporating 885 women with HPV-negative ASC-H results with follow-up over a 5-year period, this retrospective cohort is the largest studied to date. In comparison, the ALTS ASC-H report had 110 patients, only 10 of whom were older than 35 years, and whose Pap test results were retrospectively reclassified by panel review as ASC-H.6 With the exception of an earlier large study from our own institution,8 all previous studies other than the Kaiser Permanente study have included fewer than 50 women with HPV-negative ASC-H results.9 In our cohort, a mean follow-up period of 29 months allowed examination for a period well beyond the initial colposcopy to see whether a detectible CIN2/3 lesion might develop during the vigilant surveillance received by many of these women. With a relatively small loss-to-follow-up (10.1%), during a mean surveillance period of almost 2.5 years, CIN2/3/HSIL results were documented in 1.6% of cases, with no diagnoses of invasive cervical cancer. In our opinion, these follow-up findings support the option (already used by many US clinicians) of following women with HPV-negative ASC-H results and with no previous CIN diagnosis with repeat Pap and HPV cotesting in 1 year, similar to the approach recommended for women with HPV-negative ASC-US results in the 2006 ASCCP guidelines.7

The average age of our significantly older study population (37.4 years) included a large percentage of women from an age group in which the background rate of transient HPV infections is significantly reduced. In an earlier study on ASC-H from our institution,8 around 40% of patients aged 30 years and older with ASC-H LBC results tested HPV positive. The negative predictive value for CIN2/3 in HPV-negative ASC-H patients was 98.1% in women younger than 40 years and 100% in women aged 40 years and older.8 Forty-three women in our cohort had an additional follow-up HPV test result that was HPV positive after an initial HPV-negative ASC-H result, with an average time to conversion of 22 months. Of these women, 9 tested HPV negative more than once before testing HPV positive. Persistent carcinogenic HPV infections are understood to play a central role in cervical carcinogenesis38; however, quantitative and/or qualitative changes may occur during some persistent hrHPV infections and on occasion be sufficient to cause patients with persistent hrHPV infection to test HPV negative.39 The factors causing these fluctuating positive and negative results in persistently infected women are still not well understood.40 Variation in host immunologic responsiveness may be one important factor.41 False-negative hrHPV results are relatively uncommon, occurring in 7% to 10% of patients when FDA-approved HC2 test methods are used shortly before histopathologic diagnoses of invasive cervical cancer.28,42 However, false-negative hrHPV test results increase significantly when testing occurs a few years before cervical cancer diagnoses.43-45 It is reassuring that no cases of invasive carcinoma were detected in this study during followup of patients with HPV-negative ASC-H results.

At MWH, histopathologic sampling with endocervical curettage alone is generally used after a preceding colposcopic examination result is negative. Some CIN2+ lesions may be small in patients with ASC-H, and such lesions are more likely to evade detection with colposcopy.46-51 Accordingly, this follow-up study will be extended to assess the impact of further lengthening the follow-up interval. This study also found a higher rate of CIN1 on histopathology than some previous reports,52 probably reflecting the challenges inherent in consistently interpreting equivocal CIN1 changes.53 Finally, a retrospective cohort study is inherently limited because of its inability to follow up women who may be lost to further management or who may choose a different provider for cervical cancer screening.

Of 14 women with ASC-H who developed CIN2+ lesions, 4 had a history of CIN2/3 within the past 4 years. This observation is consistent with the long-term increased cervical cancer risk known to be associated with a diagnosis of carcinoma in situ54 and underscores the limitations of examining women at an initial entry point of ASC-H without risk stratification based on previous history. Modeling studies from our laboratory database have documented that previous history of CIN2/3, HSIL, and/or multiple positive hrHPV test results identifies a group of patients at significantly increased risk for histopathologic CIN2+ diagnoses compared with other patients with current HPV-negative ASC-US screening test results.15 We believe that additional study is needed on the risk stratification effect of patient history factors distinct and separate from current screening test results.

In summary, this largest study to date on patients with HPV-negative ASC-H results documented a very low (1.6%) incidence of CIN2/3 or HSIL findings and no cases of invasive cervical cancer. Because a single HPV test may be falsely negative, repeat Pap and HPV testing after 1 year appears to be a safe option for follow-up of these patients. Increased rates of histopathologic CIN2+ (10.6%) reported in 1 study with over 5-year average follow-up after HPV-negative ASC-H conventional smear results seems to argue against more extended screening intervals.11 HPV testing–based risk stratification of high-risk patients such as those with ASC-H cytology results, should, in particular, rely on hrHPV test methods thoroughly validated through externally scrutinized randomized controlled clinical trials.55-58


View Abstract