Invader human papillomavirus (HPV) type 16 and 18 assays as adjuncts to HPV screening of cervical papanicolaou smears with atypical squamous cells of undetermined significance

Authors

  • Anna K. Wong MD,

    1. Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, California
    Search for more papers by this author
  • Raymond C. Chan PhD,

    1. Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, California
    Search for more papers by this author
  • W. Stephen Nichols MD,

    1. Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, California
    Search for more papers by this author
  • Shikha Bose MD

    Corresponding author
    1. Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, California
    • Department of Pathology and Laboratory Medicine, South Tower, Room 8732, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, CA 90048
    Search for more papers by this author
    • Fax: (310) 423-0122


Abstract

BACKGROUND:

High-risk (HR) human papillomavirus (HPV) testing is standard practice for triaging women who have Papanicolaou (Pap) smears with atypical squamous cells of undetermined significance (ASC-US), however, only 5% to 17% of these women have underlying cervical intraepithelial neoplasia 2 (CIN-2)/CIN-3. Recent reports have demonstrated that the presence of either HPV type 16 (HPV-16) or HPV-18 confers an elevated risk for CIN-2/CIN-3. The current study was designed to determine the prevalence of HPV-16 and HPV-18 in ASC-US Pap smears and to determine whether further typing would enhance the risk stratification of patients for CIN-2/CIN-3.

METHODS:

One hundred seventy-eight Pap smears with ASC-US were screened retrospectively for HR HPV by using the proprietary Invader screening assay followed by typing for HPV-16 and HPV-18 by using Invader type-specific probes on 100 of the samples. Clinical follow-up results were correlated with HPV types.

RESULTS:

Fifty-one percent of the ASC-US samples were positive for HR HPV, the majority of which (70%) harbored non-HPV-16/HPV-18 HR HPV types; 27% were associated with HPV-16, whereas only 3% contained HPV-18. The screening assay indicated that 46% of women who had Pap smears with ASC-US were in need of further HPV-16/HPV-18 typing. Testing for HPV-16 stratified women with ASC-US into 3 groups: 1) 14% of women were positive for HPV-16 and had a high risk (54%) of CIN-2/CIN-3 on follow-up biopsy, 2) 35% of women were positive for non-HPV-16 HPV types and had an intermediate risk (9%), and 3) 51% of women were negative for HPV and had a negligible risk for CIN-2/CIN-3.

CONCLUSIONS:

The combined application of a proprietary screening assay and a type-specific HPV-16 assay demonstrated global potential for the development of tailored management protocols for women who have Pap smears with ASC-US. Cancer 2009. © 2009 American Cancer Society.

The ability to screen for cervical neoplasia with the Papanicolaou (Pap) smear test has resulted in a dramatic reduction in the incidence of cervical carcinoma in the US. However, the Pap smear is plagued by low sensitivity and specificity.1-3 The human papillomavirus (HPV) has been implicated as a required factor in cervical carcinogenesis.4 Of the 40 HPV types known to infect the anogenital region, only 15 are oncogenic or ‘high risk’ (HR).5 Indeed, testing for HR HPV using pooled probes has become a powerful tool in the primary screening of cervical samples and in the triage of women who have Pap smears with atypical squamous cells of undetermined significance (ASC-US).6-8

Published studies recently have highlighted the importance of HPV type 16 (HPV-16) and HPV-18 in cervical neoplasia. These 2 virus types account for a major proportion of all cervical cancers (60% and 10%, respectively).5, 9, 10 Additional studies have demonstrated that women who are infected with HPV-16 and HPV-18 have the greatest risk for developing cervical intraepithelial neoplasia 3 (CIN-3) or cancer (≥CIN-3). Khan et al11 observed that the 10-year cumulative incidence rate of ≥CIN-3 was 17.2% for women who were positive for HPV-16 and 13.6% for women who were positive for HPV-18 compared with 3% for women who had HR HPV types that were not HPV-16 or HPV-18 (non-HPV-16/HPV-18 HR HPV). Similarly, in the ASC-US–Low-Grade Squamous Epithelial Lesion (LSIL) Triage Study (ALTS), Castle et al12 noted that women who were diagnosed with ASC-US and HPV-16 demonstrated a 32.5% 2-year cumulative absolute risk for ≥CIN-3 compared with a rate of 8.4% observed in women who were infected with non-HPV-16 HR HPV types. Subsequently, Wheeler et al13 individually typed these samples for 38 HPV types by using polymerase chain reaction (PCR) analysis and determined that HPV-16 had the highest cumulative risk of ≥CIN-2 and easily dominated the oncogenic risk of other HPV types. In a large population-based study of 5696 women ages 32 to 38 years, Naucler et al14 reported that positivity for HPV-16, HPV-31, and HPV-33 at baseline conveyed the highest oncogenic risk and were responsible for 33.1%, 18.3%, and 7.7% of ≥CIN-2 cases during a mean follow-up of 4.1 years.

Currently, women with ASC-US are segregated into 2 groups based on the absence or presence of HR HPV. Those who are positive for HR HPV are referred for colposcopy and cervical biopsy. However, only 5% to 17% of such women demonstrate underlying high-grade CIN (CIN-2/CIN-3).15 The Invader screening assay, version 2.0 (Inv2) (Third Wave Technologies Inc., Madison, Wis), is a new, sequence-specific signal-amplification assay for the detection of 14 HR HPV types in liquid-based cytology through 3 separate, pooled probe sets A5/A6 (HPV-51, HPV-56, and HPV-66), A7 (HPV-18, HPV-39, HPV-45, HPV-59, and HPV-68), and A9 (HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, and HPV-58). Moreover, by using the Invader technology platform, it is now possible to select specimens for specific HPV-16/HPV-18 typing. We and others recently reported that the Inv2 screening assay demonstrates high sensitivity and specificity (>95% to 97%) for the detection of HR HPV.16, 17 The current study was designed to determine the prevalence of HPV-16 and HPV-18 in women who had Pap smears with a cytologic diagnosis of ASC-US and to determine whether further typing of these women would assist in the process of risk stratification for colposcopy referrals.

MATERIALS AND METHODS

Residual samples from 178 consecutive cervical Pap smears with a cytologic diagnosis of ASC-US that were collected in BD SurePath liquid medium (BD, Franklin Lakes, NJ) between December 2006 and April 2008 and stored at 4°C were retrospectively retrieved. The women ranged in age from 18 years to 90 years (mean, 36.4 years; median, 33 years). DNA was extracted from these samples in the NucliSens EasyMAG platform, a second-generation system for automated isolation of nucleic acid from clinical samples based on silica extraction technology (BioMe´rieux, Durham, NC).

HR HPV Screening Assay

All specimens were screened for HR HPV using the Inv2 HPV screening assay by Third Wave Technologies, Inc. In these studies, the Inv2 screening assay was conducted in a semiautomated manner, using 3 probe sets for the detection of 14 HR HPV types as described previously.16 Each of the 3 probe sets contains a ‘pool’ of oligonucleotide probes for specific HR HPV types. When an HR HPV ‘target’ is present and probe positive, a unique DNA target-probe ‘overlap’ is recognizable by the enzyme Cleavase with the release of fluorescent signal. The assay features a built-in, internal positive control reaction for the human histone cluster 2 gene H2be to assure the presence of quality DNA in each sample and to avoid false-negative results.16

Inv2 HPV oligonucleotide probes for alpha/genital HPV species A5/A6 (HPV-51, HPV-56, and HPV-66), A7 (HPV-18, HPV-39, HPV-45, HPV-59, and HPV-68), and A9 (HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, and HPV-58) were added to individual wells.16 Similarly, for specifically for HPV-16 and HPV-18, individual probes were used. For the screening assay, fluorescent signal intensity for probe pools was expressed as FAM FOZ values (ie, the ratio of HPV-specific signals and no DNA target signals). A FAM FOZ ratio was then calculated from the FAM FOZ value of the highest and lowest FAM FOZ readings. Specimens with a FAM FOZ ratio ≥1.40 were considered positive for the probe pools.

HPV Type-specific Assay

Next, 100 samples, including all 75 samples that screened positive for probe sets A7 and/or A9 and 6 of the 16 samples that screened positive only for the probe set A5/A6, were tested for HPV-16 and HPV-18 using type-specific probes from Third Wave Technologies, Inc. The residual eluted materials from the nucleic acid extraction were tested in a manner similar to that used for the screening assay. This assay uses the proprietary Invader technology with isothermal signal amplification to detect HPV-16 and HPV-18. Briefly 2 10-μL DNA aliquots from each sample were added to separate wells of a 96-well reaction plate with 20 μL of liquid wax overlay, and the plate was incubated at 95°C for 5 minutes in a GeneAmp PCR System 9700 thermal cycler. Ten microliters of master mix containing 7 μL of FAM-labeled HPV-16 or HPV-18 oligonucleotide mixture, 1 μL of Redmond red (RED)-labeled HIST2H2BE oligonucleotide mixture, and 2 μL of Cleavase enzyme solution were added to 1 of the 2 aliquots for each sample. The plate was incubated at 63°C for an additional 4 hours. After incubation, the plate was cooled to room temperature and measured using a Tecan GENios microplate fluorometer with excitation/emission settings at 485 nanometers (nm)/535 nm for FAM and 560 nm/612 nm for RED. Data analysis was performed using a generic Microsoft Excel spreadsheet developed by Third Wave Technologies, Inc. with user-defined cutoffs. In this study, both HPV-16 and HPV-18 cutoffs were set at ≥1.50. The analytical sensitivity for the 2 probes was ≤625 copies per reaction for HPV-16 and ≤5000 copies per reaction for HPV-18. These sensitivities were determined using HPV-16 and HPV-18 DNA from cervical carcinoma cell lines, CaSki and HeLa from Advanced Biotechnologies Inc. (Columbia, Md).

During this interval, benchmarks for SurePath cytologic diagnoses at our institution were concordant with national levels (ASC-US rate, 5.2% [ours] vs 4.67% [national]; ASC-US/LSIL ratio, 1.8 [ours] vs 1.36 [national]).18

Validation of Type-specific Probes

Thirty-three samples that had adequate residual nucleic acid were subjected to the PCR-linear array HPV genotyping assay (PCR-LA) (Roche Molecular Systems, Inc., Branchburg, NJ). This assay employs amplification of target DNA by PCR and nucleic acid hybridization and has been designed to detect 37 genital HPV DNA genotypes, including 13 high-risk types (HPV-16, HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, and HPV-68). PCR-LA is a research use test only.19

Clinical Follow-up and Statistical Analysis

Clinical charts for all patients were reviewed for follow-up information. A 2-sided Fisher exact test was performed to assess for the significance of disease.20

RESULTS

HR HPV Screening of ASC-US Cytology Samples

All 178 ASC-US cervical cytology specimens were typed by using the Inv2 screening assay (Fig. 1). Ninety-one women (51%) had samples that were positive for HR HPV types. Forty of these women (44%) were aged <30 years, and 51 (56%) were aged ≥30 years. Of these, 73 women had samples that tested positive with 1 probe set (16 samples were positive for A5/A6, 18 samples were positive for A7, and 39 samples were positive for A9), whereas 18 samples tested positive with 2 probe sets (6 samples were positive for both A7 and A9, 2 samples were positive for both A5/A6 and A7, and 10 samples were positive for both A5/A6 and A9). The initial screening assay revealed that 55 samples most likely were HPV-16 (positive for the A9 probe). These comprised 31% of all ASC-US samples and 60% of ASC-US samples that were positive for HR HPV. Twenty-six samples were possibly infected with HPV-18 (positive for the A7 probe). These comprised 15% of all ASC-US samples and 29% of ASC-US samples that contained HR HPV DNA.

Figure 1.

Results of screening for high-risk (HR) human papillomavirus (HPV) in women who had Papanicolaou smears with atypical squamous cells of undetermined significance (ASC-US). The A5/A6 Invader technology type-specific probe detects HPV type 51 (HPV-51), HPV-56, and HPV-66; the A7 probe detects HPV-18, HPV-39, HPV-45, HPV-59, and HPV-68; and the A9 probe detects HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, and HPV-58.

Type-specific HPV-16 and HPV-18 Screening of ASC-US Cytology Samples

One hundred samples were tested for HPV-16 and HPV-18 (Tables 1 and 2). These included 19 samples of ASC-US that were negative for HR HPV and 81 samples of ASC-US that were positive for HR HPV according to the screening assay. All 19 samples that were negative for HR HPV were negative for HPV-16 and HPV-18. Of the 81 samples that were positive for HR HPV, there were 6 samples that were positive for the A5/A6 probe and negative for the A7 and A9 probes. These samples were negative for HPV-16 and HPV-18.

Table 1. The Prevalence of Human Papillomavirus Type 16 (HPV-16) and HPV-18 in Cytology Samples Characterized as Atypical Squamous Cells of Undetermined Significance
  HR Family Types 
HPV TypingNegativeA5/A6A7A9*A7 and A9*A5/A6 and A7A5/A6 and A9*Total
  • HR indicates high risk.

  • *

    Samples likely to contain HPV-16.

  • One sample tested positive for both HPV-16 and HPV-18.

No. tested19618396210100
HPV-160001630625
HPV-1800201003
Non-HPV-16/HPV-18 HR HPV types06162332454
Table 2. The Prevalence of Human Papillomavirus Type 16 (HPV-16) and HPV-18 in Atypical Squamous Cells of Undetermined Significance Cytology Samples That Were Positive for High-risk HPV
  No. of Samples (%)
HR HPVInv2 Probe*Total TestedHPV-16HPV-18Non-HPV-16/HPV-18 HR HPV Types
  • HR indicates high risk; Inv2, the Invader screening assay (version 2.0; Third Wave Technologies, Inc., Madison, Wis).

  • *

    The A5/A6 probe detects HPV-51, HPV-56, and HPV-66; the A7 probe detects HPV-18, HPV-39, HPV-45, HPV-59, and HPV-68; and the A9 probe detects HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, and HPV-58.

Possible HPV-16sA9+5525 (45)030 (55)
Possible HPV-18sA7+2603 (12)23 (88)
Non-HPV-16/HPV-18 HPV typesA5/6+6006 (100)

Twenty-six samples were positive for the A7 probe. These included 18 samples that were positive for the A7 probe only, 6 samples that were positive for the A7 and A9 probes, and 2 samples that were positive for A7 and A5/A6. Only 3 of these 26 samples (12%) were positive for HPV-18: Two of those samples initially had screened positive for A7 only, and 1 had screened positive for both A7 and A9 (the latter sample contained DNA of both HPV-16 and HPV-18). The remaining 15 samples (88%) presumably contained HPV-39, HPV-45, HPV-59, and/or HPV-68 (the other HR HPV types that are detected by the A7 probe set).

Fifty-five samples tested positive for the screening assay A9 probe set. This included 39 samples that screened positive only for A9, 6 samples that were positive for both A7 and A9, and 10 samples that were positive for both A5/A6 and A9. Of these samples, 25 (45%) were positive for HPV-16, including the 1 sample that had dual positivity for both HPV-16 and HPV-18, as described above. These included 16 of 39 samples that were positive for A9, 2 of 6 samples that were positive for both A7 and A9, and 6 of the 10 samples that were positive for both A5/A6 and A9. The remaining 30 samples (55%) presumably were HR HPV types HPV-31, HPV-33, HPV-35, HPV-52, and/or HPV-58 (the other HR HPV types that are detected by the A9 probe set).

Overall, only 25 of 91 HR HPV-positive ASC-US samples (27%) were positive for HPV-16, and 3 of 91 samples (3%) were positive for HPV-18. This included the 1 sample (1%) that was positive for both HPV-16 and HPV-18. The remaining 70% of HR HPV-positive ASC-US samples were non-HPV-16/HPV-18 HR HPV types.

Validation of the Type-specific HPV-16 and HPV-18 Probes

PCR-LA was performed on 33 samples to determine the specificity and sensitivity of the Invader HPV type-specific probes for HPV-16 and HPV-18. Multiple HPV types (range, 2-7 types) were detected in 23 samples by PCR-LA. The remaining 10 samples contained only 1 HIV type. Ten samples contained HPV-16 DNA, and 1 sample contained HPV-18 DNA. Details of the samples with discordant HPV genotype results are presented in Table 3. Of the 10 samples that were positive for HPV-16 by PCR-LA, 8 samples concordantly were positive for the Invader HPV-16 assay, 2 samples were discordant, including 1 sample that initially had tested positive with the A9 probe and 1 sample that was negative for all 3 of the screening assay probes. The type-specific HPV-16 assay also detected 2 samples with false-positive results that were negative for HPV-16 by PCR-LA. Both of these samples contained multiple HPV types, as demonstrated by PCR-LA, including HPV-31, which was present in both samples.

Table 3. Summary of Atypical Squamous Cells of Undetermined Significance Cases With Discordant Results From the Invader Screening and Type-specific Assays Versus Polymerase Chain Reaction-Linear Array Human Papillomavirus Genotyping
 Inv2    
Specimen No.Screen*HPV-16HPV-18PCR ResultsConcordance Inv2 HPV-16 and PCR-LA (Yes/No)Concordance Inv2 HPV-18 and PCR-LA (Yes/No)Cervical Biopsy Diagnosis
  • Inv2 indicates the Invader screening and type-specific assays (version 2.0; Third Wave Technologies, Inc., Madison, Wis); HPV-16, human papillomavirus type 16; PCR, polymerase chain reaction; PCR-LA, PCR-linear array HPV genotyping; NA, not available; w, weak, CIN, cervical intraepithelial neoplasia.

  • *

    The A9 probe detects HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, and HPV-58.

  • High-risk HPV types.

SP133NegativeNegativeNegative16, 39NoYesNA
SP55A9NegativeNegative16w, 52, 54, 58, 59w, 62, 84NoYesCIN-3
SP68A9NegativeNegative18, 31, 39w, 61wYesNoNegative
SP104A9PositiveNegative31, 42, 54, 61NoYesNegative
SP135A9PositiveNegative31, 84NoYesCIN-2

There was 1 sample in which HPV-18 was identified with PCR-LA but was not detected by the HPV-18 type-specific probe. The positive result of this sample with the A9 screening probe most likely detected HPV-31, which also was present (Table 3).

The type-specific HPV-16 assay gad a concordance rate of 88% (29 of 33 samples) with PCR-LA, a sensitivity of 80%, a specificity of 91%, a positive predictive value of 80% and a negative predictive value of 91%. The number of cases with HPV-18 was too small for statistical analysis.

Clinical Follow-up of Patients With ASC-US

Follow-up information was available in 70 of 178 women (39%) who had samples with ASC-US cytology (Table 4). The women ranged in age from 19 years to 68 years (median, 32 years; mean, 36 years). Twenty-four women were negative for HR HPV types, 13 women were positive for HPV-16, 1 woman was positive for HPV-18, and 32 women were positive for non-HPV-16/HPV-18 HR HPV types. Follow-up ranged from 3 weeks to 49 weeks (mean, 15 weeks, median, 21.5 weeks). Nine women (13%) had repeat Pap smears, and 61 women (87%) underwent colposcopy-guided cervical biopsy. No significant pathology was detected in 45 women. A follow-up Pap smear revealed LSIL in 1 woman. Colposcopic-guided cervical biopsy revealed HPV-induced changes and/or CIN-1 in 14 women and CIN-2/CIN-3 in 10 women. There were no women with invasive carcinoma.

Table 4. Follow-up of Women With Atypical Squamous Cells of Undetermined Significance Papanicolaou Smears
    No. of Women (%)
HPV StatusTotal No. of Women (%)No. With Follow-UpAverage Follow-up, wkNILMLSIL/CIN-1CIN-2/CIN-3
  1. HPV indicates human papillomavirus; NILM, negative for intraepithelial lesion; LSIL, low-grade squamous intraepithelial lesion; CIN, cervical intraepithelial neoplasia; HPV-16, HPV type 16; +, positive; HR HPV, high-risk HPV; −, negative.

HPV-16+24 (13)13155 (38)1 (8)7 (54)
HPV-18+2 (1)138010
HPV-16+ and HPV-18+1 (1)0
Non-HPV-16/HPV-18 HR HPV+64 (36)321619 (60)10 (31)3 (9)
HR HPV−87 (49)241521 (88)3 (13)0
Total178 (100)701545 (64)15 (21)10 (14)

Follow-up of the 24 women with ASC-US who were negative for HR HPV revealed that 21 women (88%) were without evidence of dysplasia, 3 women (13%) had LSIL/CIN-1, and no women had CIN-2/CIN-3. Of the 32 women who had non-HPV-16/HPV-18 HR HPV types, 19 women (60%) were negative for dysplasia, 10 women (31%) had LSIL/CIN-1, and 3 women (9%) had CIN-2/CIN-. Five of the 13 women (38%) with HPV-16-positive ASC-US had no dysplasia, 1 woman (8%) had LSIL/CIN-1, and 7 women (54%) had CIN-2/CIN-3. Significantly more CIN-2/CIN-3 was detected among women with HPV-16-positive ASC-US on follow-up biopsies compared with the other 2 groups of women (ie, women who were positive for non-HPV-16/HPV-18 HR HPV types [P = .003] and women who were negative for HR HPV [P < .001]). Alternatively, of the 10 cases of ASC-US with underlying CIN-2/CIN-3 in this series, 7 (70%) occurred in women with HPV-16-positive ASC-US. Follow-up of the 1 woman with HPV-18-positive ASC-US revealed CIN-1 on cervical biopsy.

Follow-up revealed a similar age distribution in the different lesions detected (Table 5). Although 7 of the 32 women (22%) aged <30 years had CIN-2/CIN-3 on follow-up biopsy, compared with 3 of 38 women (8%) aged ≥30 years, statistical significance was not achieved (P = .168). Four of the 7 women aged <30 years were positive for HPV-16, whereas the remaining 3 women had non-HPV-16/HPV-18 HR HPV types. All 3 women aged ≥30 years were positive for HPV-16 (Table 6).

Table 5. Results of Follow-up in Women With Atypical Squamous Cells of Undetermined Significance Papanicolaou Smears Distributed by Age
 No. of Women (%)
Follow-up Results, n=70Age <30 Years, n=32Age ≥30 Years, 38
  • NILM indicates negative for intraepithelial lesion; LSIL, low-grade squamous intraepithelial lesion; CIN, cervical intraepithelial neoplasia.

  • *

    The difference was not significant (P=.168).

NILM, n=4517 (53)28 (74)
LSIL/CIN-1, n=158 (25)7 (18)
CIN-2/CIN-3, n=107 (22)*3 (8)*
Table 6. Human Papillomavirus Type 16-positive Women With Follow-up Distributed by Age
Biopsy ResultsAge <30 YearsAge ≥30 Years
  1. NILM indicates negative for intraepithelial lesion; CIN, cervical intraepithelial neoplasia.

NILM/CIN-151
CIN-2/CIN-343

Four of the 5 women who had discordant type-specific HPV-16/HPV-18 assay results and PCR-LA results underwent cervical biopsies (Table 3), which revealed CIN-2 in 1 woman, CIN-3 in 1 woman, and no evidence of dysplasia in 2 women. Of the 2 women who had ≥CIN-2, 1 would have been identified as positive by the Invader HPV-16 probe.

To summarize, 91 of 178 of Pap smears with ASC-US (51%) in our series were positive for HR HPV types, and the majority (64 of 91 Pap smears; 70%) harbored non-HPV-16/HPV-18 HR HPV types, 25 of 91 Pap smears (27%) were associated with HPV-16, and only 3 of 91 Pap smears (3%) contained HPV-18. The screening assay identified 31% (A9 probe positive) and 15% (A7 probe positive) of women with ASC-US who were in need of further HPV-16 and HPV-18 typing. Forty-five percent of the former group tested positive for HPV-16, whereas 12% of the latter group tested positive for HPV-18. Testing for HPV-16 identified a subgroup of women with HR HPV-positive ASC-US who had a significantly higher incidence of CIN-2/CIN-3 on follow-up biopsy. A statistically significant difference was not noted in the frequency of detecting CIN-2/CIN-3 lesions in women aged <30 years compared with women aged ≥30 years.

DISCUSSION

HR HPV DNA testing in conjunction with Pap smears has become standard practice in the US for cervical cancer screening.6, 7 Reflex HR HPV DNA testing using pooled probes is a cost-effective management strategy for the triage of women with ASC-US.8 It is well documented that this strategy yields a high clinical sensitivity in predicting ≥CIN-2 in women with ASC-US Pap smears. With the high sensitivity and negative predictive value (NPV) of HR HPV DNA testing, women who have negative results largely are assured that they have a minor risk of having or developing ≥CIN-2. However, the high sensitivity and NPV of this test are achieved at the expense of relatively lower specificity and lower positive predictive values. Consequently, unnecessary colposcopy and biopsy may be experienced in the setting of insignificant or transient HR HPV infection. Conversely, in women who have a significantly greater risk of developing ≥CIN-2, such as those with persistent HPV-16 or HPV-18 infections, a positive HR HPV DNA result unaccompanied by viral genotyping is poorly informative for individual case management.21 Currently, little information is available describing the utility of HPV typing in patient care. The complexity of HPV types, the widely ranging oncogenic potential, and the frequent presence of multiple HR HPV infections in precancerous lesions make it difficult to determine which HR HPV type should be given priority in the prediction of ≥CIN-2 or progressive lesions.5, 12 Our study clearly demonstrates that women who have Pap smears that contain ASC-US and are positive for HPV-16 carry an enhanced risk for CIN-2/CIN-3 (54%) on follow-up biopsy compared with women who are positive for non-HPV-16 HR HPV types (9%) and women who are negative for HR HPV (0%). These findings are reminiscent of the ALTS trial.12 In addition, in a recent study that evaluated the presence of 8 selected HR HPV virus types in mildly abnormal Pap smears, Guo et al22 noted the highest association of ≥CIN-2 with HPV-16 and HPV-18 in women who had Pap smears with ASC-US.

The incidence of HPV-16 in the current series of ASC-US Pap smears seems small (14% of all ASC-US Pap smears; 27% of HR HPV-positive ASC-US Pap smears), although the numbers are similar to previous reports. In the ALTS trial,10 an incidence of 14.9% was reported for HPV-16, and Evans et al15 reported an incidence of 17.5% in ASC-US specimens. The majority of our ASC-US samples (73%; 64 of 91 samples) were non-HPV-16 HR HPV types, a heterogeneous group that carried a significantly lower (9%), but not necessarily negligible, risk for ≥CIN-2. In the ALTS trial,10 a risk of 8.4% was reported for ≥CIN-2 in non-16 HR HPV-positive patients. Guo et al,22 studying 8 of the most common oncogenic HPV types (HPV-16, HPV-18, HPV-31, HPV-33, HPV-35, HPV-45, HPV-52, and HPV-58) responsible for >80% of ≥CIN-2, reported 2 cases of ≥CIN-2 in patients who were not infected by those 8 HR HPV types. Altogether, it is reasonable to assume that the risk of ASC-US Pap smears with HPV-16 is considerable and that specimens without HPV-16 carry a risk that should not be ignored.

The incidence of HPV-18 was extremely low (3%) in our series, and the single women who had histologic follow-up available was diagnosed with CIN-1. Similar studies addressing HPV-18 are scarce. In a survey of 1100 Pap smears with cytologic diagnoses based on Bethesda 2001 guidelines, Evans et al15 described small numbers of non-HPV-16 HR HPV types in the different cytologic categories, such that a statistical association between HPV type and cytologic diagnosis could not be obtained. This indirectly suggests a low incidence of HPV-18 in ASC-US Pap smears as opposed to a false-negative result by the Inv2 HPV-18 type-specific assay, which is less sensitive analytically than the Inv2 HPV-16 assay. In addition, this is supported further by our previous experience of a low HPV-18 incidence in patients with ASC-US who were evaluated by PCR.16 Further studies will be needed to address the significance of various values for analytical sensitivity, because the presence of very low HPV DNA levels detectable by highly sensitive assays may not correlate clinically with the development or presence of disease.

Because the majority (73%) of our known HR HPV-positive women was negative for HPV-16 and at low risk for ≥CIN-2, the triage of these women to immediate colposcopy may represent excessive intervention for a relatively low-risk process. This issue is in need of clarification. More informative HR HPV testing of patients who are at risk for cervical cancer may distinguish between those who need immediate colposcopy and individuals who require less ‘aggressive’ management.

When it is used as an initial test, the informative Inv2 screening assay provides HR HPV type results that identify individual women with ASC-US who need confirmatory HPV-16 testing. The Inv2 screening assay is sensitive and specific.16, 17 Positivity for the A9 probe allows direction of approximately 31% of specimens with HR HPV ASC-US to a second test, specifically for HPV-16. HPV-16 type-specific testing permits a 3-tiered classification of ASC-US patients based on HPV status: 1) women with HPV-16 who, thus, are at high risk for ≥CIN-2 on follow-up biopsy; 2) women with non-HPV-16 HR HPV types, who have a corresponding low risk for ≥CIN-2; and, 3) women who are negative for HR HPV, who have a negligible risk for ≥CIN-2. On the basis of this risk stratification, women with ASC-US Pap smears who are negative for HR HPV should return for routine Pap smear screening at 12 months (current standard practice) (Fig. 2A). Individuals who are positive for HPV-16 should be directed to immediate colposcopy and biopsy (Fig. 2B). The remaining women with non-HPV-16 HR HPV types (those with ASC-US who are positive for the A5/A6, A7, or A9 probes positive but negative for HPV-16) may be followed reasonably by heightened surveillance at a 6- to 12-month intervals with repeat Pap smears and HPV testing. Currently, research is lacking in the area pertaining to HPV persistence and cervical carcinogenesis. However, a repeat cytologic examination at 6- to 12-month intervals in patients who have a Pap smear with ASC-US is recommended by the American College of Gynecology and by Wright et al.6, 23 It is reasonable and desirable, then, that women with HR HPV (but not HPV-16) should undergo repeat Pap smear and HPV testing at 6- to 12-month intervals. If abnormalities persist on follow-up cytology or HPV testing, then we believe that these women should undergo colposcopy. In the future, it will be important to approach the issue of non-HPV-16 HR HPV ‘persistence,’ frequency, and duration of repeat cytology and the risk of CIN in these women who are infected with non-HPV-16 HR HPV strains. To date, it is not clear whether age should factor into the consideration of management in such patients. In the ALTS trial,12 a slight increase was reported (1.6%) in the 2-year absolute risk of ≥CIN-3 in patients who were infected with non-HPV-16 HR HPV types and who were ≥30 years compared with women aged <30 years. In the current study, the incidence of ≥CIN-2 in the 2 age groups was not statistically significant. We suggest that additional, prospective studies would be relevant and are needed to determine the most appropriate follow-up for this group of patients.

Figure 2.

(A) The current management algorithm for women who have Papanicolaou (Pap) smears demonstrating atypical squamous cells of undetermined significance (ASC-US) (see Wright et al, 200224). HPV indicates human papillomavirus; HR HPV, high-risk HPV; (−), negative; (+), positive. (B) The proposed management algorithm for women diagnosed with ASC-US Pap smears through the combined use of the Invader screening assay and HPV-16 type-specific assay.

In the current study, the false-negative detection for HPV-16 with the Invader type-specific assay in samples with low DNA copy numbers raises the potential for either 1) adjustment of the sensitivity of the assay or 2) the need for repeat sampling for patients. With regard to the latter, it is possible that the accuracy of cervical testing for HPV overall simply requires a small number of repeat assays because of otherwise normal, clinical variation(s). It is not clear whether the latter issue could be resolved by greater ‘caution’ in obtaining clinical specimens.

We conclude that the Inv2 screening assay and the type-specific Invader HPV-16 assay, together, have the potential for mass application in women who have Pap smears with ASC-US. Currently, the benefit of rapid triage for the patients at greatest risk (HPV-16-positive) seems clear. However, additional studies will be necessary to determine the most useful follow-up process for patients who are positive for non-HPV-16 HR HPV types with either persistent or clearing infection.

Conflict of Interest Disclosures

Invader screening probes and type-specific HPV-16/HPV-18 probes were provided free of cost for this study.

Ancillary