Image-guided ThinPrep Papanicolaou tests and cotesting with high-risk human papillomavirus in women aged 30 years and older in a low-risk private practice population

Authors

  • Yimin Ge MD,

    Corresponding author
    1. Department of Pathology, Methodist Hospital, Houston, Texas
    2. Department of Pathology, Weill Medical College of Cornell University, New York, New York
    • Department of Pathology, The Methodist Hospital, 6565 Fannin Street, Houston, TX 77030
    Search for more papers by this author
    • Fax: (713) 793-1603

  • Debora Smith CT,

    1. Department of Pathology, Methodist Hospital, Houston, Texas
    Search for more papers by this author
  • Mary R. Schwartz MD,

    1. Department of Pathology, Methodist Hospital, Houston, Texas
    2. Department of Pathology, Baylor College of Medicine, Houston, Texas
    Search for more papers by this author
  • Dina R. Mody MD

    1. Department of Pathology, Methodist Hospital, Houston, Texas
    2. Department of Pathology, Weill Medical College of Cornell University, New York, New York
    Search for more papers by this author

Abstract

BACKGROUND:

Screening for cervical cancer precursors has evolved considerably with the introduction of new technologies to improve the early detection of disease. The objective of this study was to analyze the accuracy and effectiveness of combined screening with cytology and high-risk human papillomavirus (HR-HPV) testing in a low-risk population of women aged ≥30 years.

METHODS:

Consecutive unselected samples from a group of 1871 women aged ≥30 years were screened with image-guided ThinPrep tests and HR-HPV tests during a 6-month period. Histologic follow-up was reviewed among women with positive HR-HPV tests.

RESULTS:

A total of 85 (4.5%) women had positive HR-HPV tests. In 48 HR-HPV–positive women with follow-up biopsies, 41 (85%) were found to have histologic abnormalities. Thirty-three (1.9%) women with cytologically normal Papanicolaou (Pap) tests harbored HR-HPV, and a cervical intraepithelial neoplasia (CIN) 2+ lesion was detected in 1 (16%) of 6 women with histologic follow-up. Conversely, 2 (28%) of 7 women with high-grade intraepithelial lesion on cytology tested negative for HR-HPV during the same period. A case of serous carcinoma with atypical glandular cells on cytology was also negative for HR-HPV, as expected.

CONCLUSIONS:

In this low-risk population of women aged ≥30 years, histology-confirmed CIN2+ lesions were identified in women with negative cytology and positive HR-HPV tests, as well as in those with positive cytology and negative HR-HPV tests. Because both cytology and HPV testing alone missed significant lesions, cotesting with Pap and HR-HPV in women aged ≥30 years appears to be a reasonable option in a low-risk population. (Cancer Cytopathol) 2009. © 2009 American Cancer Society.

Persistent cervical infection with high-risk human papillomavirus (HR-HPV) is a causal agent of invasive cervical cancer and has been linked to 86% to 94% of cases in a recent international meta-analysis.1 Building on the tremendous success of conventional smear cytology, liquid-based cervical cytology has improved the ability to reliably detect significant cervical lesions,2-5 which is further enhanced by the ThinPrep Imaging System6-8 approved by the US Food and Drug Administration. Despite of the success of cytology as a cervical cancer screening method, it has had limitations, especially its relatively low sensitivity. These limitations have led to considerable interest in using combining cervical cytology with HR-HPV testing, which has demonstrated high sensitivity for detecting high-grade squamous dysplasia (cervical intraepithelial neoplasia [CIN] 2+). The American Cancer Society (ACS)9 and the American College of Obstetricians and Gynecologists (ACOG)10 have recently recommended cotesting with cytology and HR-HPV as routine cervical screening for women aged ≥30 years. Others have advocated alternative screening strategies using HR-HPV testing as a primary screening method to maximize the benefit of its high sensitivity and reproducibility.11, 12 Several primary screening strategies with different combinations have been proposed. The debate has mainly been centered on issues regarding positive and negative predictive values as well as cost-effectiveness of the proposed primary screening strategies. We studied the data from a low-risk private practice population of women aged ≥30 years who were screened with combined ThinPrep Imaging System and HR-HPV DNA cotesting. The performance of each testing method in detecting significant lesions was analyzed along with histologic findings in follow-up biopsies.

MATERIALS AND METHODS

The cytology laboratory of the Methodist Hospital, Houston, Texas processes approximately 21,000 Papanicolaou (Pap) tests annually, >99% of which are liquid-based preparations. Our patients from private practices in obstetrics and gynecology primarily represent a normal screening population with a low risk of cervical cancer. Some of the physicians routinely order HR-HPV and liquid-based Pap cotesting for patients aged ≥30 years in their practice. Patients from dysplasia and charity clinics were excluded from this study. The study was conducted with approval of the institutional review board of the Methodist Research Institute (IRB# 00000653).

From July 1, 2007 through December 31, 2007, consecutive cases of women aged ≥30 years who presented for cervical screening at private practices in obstetrics and gynecology were evaluated with combined liquid-based cytology and HR-HPV DNA cotesting. All patients included in the study had no previous history of an abnormal Pap test. The Pap tests were prepared and interpreted at the Methodist Hospital using a ThinPrep 2000 Cytyc Processor assisted with an imaging system (Cytyc Corporation, Marlborough, Mass). The HR-HPV DNA tests were performed by ARUP Laboratory (Salt Lake City, Utah) using Hybrid Capture 2 assay (Digene, Gaithersburg, Md).

All negative Pap tests with positive HR-HPV were manually rescreened before sign-out. Among women with positive HR-HPV tests or positive Pap tests/negative HR-HPV tests, our surgical pathology database was searched for histologic follow-up. All available cervical biopsies in these patients were reviewed by 2 pathologists independently.

RESULTS

A total of 1871 woman aged ≥30 years was screened by combined image-guided ThinPrep Pap test and HR-HPV DNA test. Among these women, 51 (2.7%) tested positive for both HR-HPV and cytology, 33 (1.8%) tested positive for HR-HPV with negative cytology, 61 (3.3%) tested positive for cytology with negative HR-HPV, and 1700 (90.0%) tested negative for both.

A total of 85 (4.5%) women had positive HR-HPV tests (Table 1). A high prevalence of HR-HPV infection was observed in women with cytologic abnormalities (Table 1) (Fig. 1), which included high-grade squamous intraepithelial lesion (HSIL; 71%), low-grade squamous intraepithelial lesion (LSIL; 63%), atypical squamous cells/cannot exclude HSIL (ASC-H; 50%), atypical squamous cells of undetermined significance (ASCUS; 29%), and atypical glandular cells (AGC; 25%). Follow-up cervical biopsies demonstrated histologic abnormalities in 41 (85%) of 48 women with positive HR-HPV tests (Table 1) (Fig. 2).

Table 1. Correlation Between Cytology, HR-HPV Testing, and Histology
CytologyHR-HPV TestHistologic Follow-Up in HR-HPV+ Patients
Diagnostic CategoryCases%HPV+%*CasesNPCHPV EffectCIN1CIN2+
  • HR-HPV indicates high-risk human papillomavirus; +, positive; NPC, no pathologic changes; CIN1, type 1 cervical intraepithelial neoplasia; NILM, negative for intraepithelial lesion or malignancy; ASCUS, atypical squamous cells of undetermined significance; LSIL, low-grade squamous intraepithelial lesion; HSIL, high-grade squamous intraepithelial lesion; ASC-H, atypical squamous cells/cannot exclude HSIL; AGC, atypical glandular cells.

  • *

    Percentage of diagnostic category.

  • CIN2+ included CIN2, CIN3, adenocarcinoma is situ, and invasive carcinoma.

NILM173392.6331.964 (66.7%)1 (16.7%)01 (16.7%)
ASCUS703.72029152 (13.3%)8 (53.3%)2 (13.3%)3 (20%)
LSIL351.92263191 (5.3%)8 (42.1%)6 (31.6%)4 (21.1%)
HSIL70.4571501 (20%)1 (20%)3 (60%)
ASC-H60.335020002 (100%)
AGC40.212510001 (100%)
Malignancy000000000
Unsatisfactory150.816.700000
Total1871100854.5487 (14.6%)18 (37.5%)9 (18.8%)14 (29.2%)
Figure 1.

Prevalence of high-risk human papillomavirus (HR-HPV) infection is shown for each category of cytologic diagnosis. + indicates positive; −, negative; NILM, negative for intraepithelial lesion or malignancy; ASCUS, atypical squamous cells of undetermined significance; LSIL, low-grade squamous intraepithelial lesion; HSIL, high-grade squamous intraepithelial lesion; ASC-H, atypical squamous cells/cannot exclude HSIL; AGC, atypical glandular cells; UNSAT, unsatisfactory.

Figure 2.

(A) A high-risk human papillomavirus-positive patient had a normal Papanicolaou test. (B) Moderate squamous dysplasia (type 2 cervical intraepithelial neoplasia) was detected in a subsequent cervical loop electrosurgical excision procedure.

As many as 1.9% of women with cytologically normal Pap tests harbored HR-HPV, which constitutes 39% of total HR-HPV–positive cases in this low-risk population (Fig. 3). High-grade squamous dysplasia (CIN2+) and HPV effect were detected in 2 of 6 women with follow-up biopsies (Table 1) (Fig. 2).

Figure 3.

A breakdown of high-risk human papillomavirus-positive cases by cytologic diagnosis is shown. NILM indicates negative for intraepithelial lesion or malignancy; LSIL, low-grade squamous intraepithelial lesion; ASCUS, atypical squamous cells of undetermined significance; UNSAT, unsatisfactory; AGC, atypical glandular cells; ASC-H, atypical squamous cells/cannot exclude high-grade squamous intraepithelial lesion; HSIL, high-grade squamous intraepithelial lesion.

In 19 women with LSIL on Pap tests, 18 (95%) had histologic abnormalities on histologic follow-up, which included 4 (21%) with CIN2+ lesions. Among 15 women with atypical squamous cells of undetermined significance on Pap tests, 13 (86%) had histologic abnormalities, with 3 (13%) CIN2+ lesions noted on histologic follow-up. Both women with atypical squamous cells/cannot exclude HSIL on Pap tests had CIN2+ lesions (Table 1).

There were 7 cases of HSIL detected on Pap tests, with 5 of them positive for HR-HPV. The remaining 2 cases tested negative or equivocal for HR-HPV. On review, these 2 cases were unequivocal HSILs of normal cellularity and many dysplastic cells. In both of these patients, CIN3 was detected in subsequent biopsies (Figs. 4A and 4B). A case of endometrial papillary serous carcinoma with a cytologic diagnosis of AGC was also found to be HR-HPV negative (Figs. 4C and 4D).

Figure 4.

Negative high-risk human papillomavirus tests were noted in patients found to have (A) high-grade squamous intraepithelial lesion and (C) atypical glandular cells on Papanicolaou tests. Histologic follow-up revealed (B) severe squamous dysplasia (cervical intraepithelial neoplasia 3) and (D) serous carcinoma, respectively.

DISCUSSION

By using the ThinPrep Imaging System combined with HR-HPV DNA testing, we detected HR-HPV infection in 4.5% of this low-risk population aged ≥30 years. Notably, 1.9% of women who had cytologically normal Pap tests were found to harbor HR-HPV. Although the rate is much lower than that reported in several large groups of demographically diverse women (range, 2.4‒9%),13-15 a significant cervical lesion (CIN2+) was detected in 1 (16%) of 6 women with histologic follow-up. Conversely, negative or equivocal HR-HPV tests were identified in 2 (28%) of 7 women found to have HSIL on Pap tests and CIN2+ on follow-up biopsies. The HR-HPV test was also negative in a woman who had atypical glandular cells on Pap test and a papillary serous carcinoma on subsequent biopsy. Both cytology and HR-HPV testing alone missed a significant portion of high-grade lesions in this small sample of low-risk women. Therefore, cotesting with cytology and HR-HPV appears to be a reasonable option in a low-risk population of women aged ≥30 years.

Cytology-based cervical cancer prevention programs have proven to be highly effective and are considered to be 1 of the most successful approaches to cancer prevention to date. Despite overall success, cervical cancer still continues to be a cause of cancer morbidity and mortality in women in the United States. This has been attributed to limitations inherent in cytology-based screening and, more importantly, the lack of screening in some populations. The relatively low sensitivity of cytology-based screening has been considered to be 1 of the prominent limitations, and was reported to range from 30% to 87% based mainly on experiences with the conventional Pap smear.5, 16, 17 The introduction of liquid-based cytology increased the detection rate for LISL, but it was reportedly difficult to further improve the sensitivity for detecting CIN2+ cervical lesions.18-20 However, a comparative trial of 2047 women with biopsy verification demonstrated that the sensitivity for the detection of CIN2+ lesions was similar for the ThinPrep Pap test and HPV direct test (94% vs 95%).21 Recently, the ThinPrep Imaging System has been shown to considerably improve the sensitivity of the Pap test.6-8, 22, 23 Even with advanced technologies and rescreening negative Pap tests in HR-HPV–positive cases, the results of the current study demonstrate that cytology alone missed significant cervical lesions (CIN2+) in this low-risk population. In addition, histologic follow-up in women with positive HR-HPV indicated that high-grade cervical lesions (CIN2+) were underdiagnosed on cytology in 13% of women with ASCUS and in 21% of women with LSIL (Table 1). The difficulty in further improving the sensitivity of cytology-based screening necessitates seeking more sensitive screening tests that can be used either alone or in combination with cytology.

On the basis of our understanding of the pathogenetic role of HPV in cervical cancer, sensitive molecular tests for detecting HR-HPV DNA have become widely available. In recent large European trials,24-26 HR-HPV DNA testing was found to have significant higher sensitivity (>97%) in comparison with cytology, but lower specificity. However, the false-negative rate of HPV testing for CIN2+ lesions has been reported to range from 7% to as high as 18.5%.1, 27, 28 In the current study, negative or equivocal HR-HPV tests were identified in 2 (28%) of 7 women with HSIL on Pap tests, which was confirmed to be CIN2+ cervical lesions on follow-up histology. This finding raises a critical question regarding the reliability of HR-HPV testing for detecting high-grade cervical lesions in low-risk women aged ≥30 years. Despite the small study population, the unacceptably low sensitivity and negative predictive value in the current study strongly argue against the appeal11, 12, 29 of using HR-HPV testing as a primary screening test for cervical cancer. Furthermore, glandular preinvasive or malignant lesions commonly detected by cytology may not be related to HPV infection, as was also observed in this study. Although glandular lesions are much less common than cervical squamous lesions, they are significant lesions that have a dramatic impact on the quality of patient outcome if they remain undetected. We believe that the ability to identify significant glandular lesions, especially those of endometrial origin, should be considered when formulating screening strategies, and that cytology is still the most effective screening test for these lesions at the present time.

The current study data support the recent recommendations by the ACS9 and ACOG10 to incorporate HR-HPV testing into cytologic evaluation for the primary screening of women aged ≥30 years in expectation of improving the sensitivity of the Pap test and the negative predictive value of a double-negative test (Pap−/HPV−). On the basis of the findings presented herein, we believe that cytology and HR-HPV testing are complementary, and the cotesting strategy will greatly increase the sensitivity for detecting squamous intraepithelial lesions, while retaining the ability to identify other significant lesions that are unrelated to HPV infection. Negative Pap and HR-HPV tests were found in >90% of women in our low-risk population. Recent studies demonstrated that women aged ≥30 years who had both negative Pap and negative HR-HPV tests were at extremely low risk for developing high-grade squamous dysplasia (CIN2+) in the next 3 to 5 years.30 After analyzing 735 cases of LSIL in women aged ≥30 years during the past 2 years at our institution, Thrall et al found that women with positive HR-HPV have a significantly higher likelihood of harboring CIN2+ lesions than those who test negative for HR-HPV (unpublished data). Using a combination of testing methods to focus screening and clinical attention to cases at risk might permit the identification of those women with LSIL who are likely to develop disease progression at an earlier and curable stage of disease and lengthen screening intervals for most women in the primary screening pool. This strategy is even more appealing for a low-risk population with a larger fraction of women being negative for both cytology and HR-HPV tests.

The most prominent issue related to combined cytologic and HPV testing is how to manage women who are positive for HR-HPV and negative on Pap test. On the basis of recent studies, those women with positive HR-HPV and negative cytology are at low risk for having CIN2+ lesions.24, 31 On the basis of this low risk, the 2006 American Society for Colposcopy and Cervical Pathology consensus guidelines recommended conservative follow-up with repeat cytology and HPV testing at 12 months for cytology-negative, HR-HPV–positive women.32 In the current study, we detected 1 (16%) histology-confirmed CIN3 lesion among 6 Pap-negative, HR-HPV–positive women with follow-up biopsies. Although a definite conclusion cannot be drawn from this small study population, further investigation is needed to search for better management approaches for women with a positive HR-HPV test and negative cytology.

Data from the current study should be interpreted with caution because of the small study population. Although a significant proportion of high-grade lesions were missed by either cytology or HR-HPV testing alone, there was not enough statistical power to determine the sensitivity and specificity of the 2 tests in this low-risk population of women aged ≥30 years. In addition, valuable colposcopic data were unavailable in the study, and verification bias may also be present because of the absence of histology follow-up in patients with negative HR-HPV tests. Further studies with large cohorts will provide valuable information regarding issues that could have a significant impact on assessment of screening strategies.

Conflict of Interest Disclosures

The authors made no disclosures.

Ancillary