Research Methods for Health Sciences 2

amaay2

hpv_.pdf

Home >Applied Sciences homework help >Research Methods for Health Sciences 2

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 473 DOI: 10.1309/AJCPHVD7MIP3FYVV

Human Papillomavirus Oncogenic mRNA Testing for Cervical Cancer Screening

Baseline and Longitudinal Results From the CLEAR Study

Jennifer L. Reid, PhD,1 Thomas C. Wright Jr, MD,2 Mark H. Stoler, MD,3 Jack Cuzick, PhD,4 Philip E. Castle, PhD,5 Janel Dockter,1 Damon Getman, PhD,1 and Cristina Giachetti, PhD1

From 1Hologic, Incorporated, San Diego, CA; 2Department of Pathology, Columbia University School of Medicine, New York, NY; 3University of Virginia Health System, Charlottesville; 4Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, UK; and 5Global Cancer Initiative, Chestertown, MD.

Key Words: Aptima HPV; Adjunct screening; Cervical cancer; E6/E7 mRNA; Cervical intraepithelial neoplasia Am J Clin Pathol September 2015;144:473-483

DOI: 10.1309/AJCPHVD7MIP3FYVV

ABSTRACT

Objectives: This study determined the longitudinal clinical performance of a high-risk human papillomavirus (HR-HPV) E6/E7 RNA assay (Aptima HPV [AHPV]; Hologic, San Diego, CA) compared with an HR-HPV DNA assay (Hybrid Capture 2 [HC2]; Qiagen, Gaithersburg, MD) as an adjunctive method for cervical cancer screening.

Methods: Women 30 years or older with a negative result for intraepithelial lesions or malignancy cytology (n = 10,860) positive by AHPV and/or HC2 assays and randomly selected women negative by both assays were referred to colposcopy at baseline. Women without baseline cervical intraepithelial neoplasia (CIN) grade 2 or higher (CIN2+) continued into the 3-year follow-up.

Results: The specificity of AHPV for CIN2 or lower was significantly greater at 96.3% compared with HC2 specificity of 94.8% (P < .001). Estimated sensitivities and risks for detection of CIN2+ were similar between the two assays. After 3 years of follow-up, women negative by either human papillomavirus test had a very low risk of CIN2+ (<0.3%) compared with CIN2+ risk in women with positive AHPV results (6.3%) or positive HC2 results (5.1%).

Conclusions: These results support the use of AHPV as a safe and effective adjunctive cervical cancer screening method.

Cervical cancer is one of the most frequent can- cers in women worldwide, accounting for approximately 530,000 new cases and 275,000 deaths annually.1 Countries with well-organized screening programs using conventional Papanicolaou (Pap) stain cytology have experienced sub- stantially reduced mortality from the disease in the past 5 decades.2-4 Despite this advance, the relatively low sensitiv- ity and reproducibility of both conventional Pap smear and liquid-based cytology screening methods have prompted investigation into identifying adjunctive methods with Pap cytology for improving detection of cervical neoplasia.5-9

Infection with 14 high-risk human papillomavirus (HR- HPV) genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) is associated with almost all cases of cervical precancer, defined as cervical intraepithelial neoplasia (CIN) grade 2 (CIN2), grade 3 (CIN3), and cancer.10 Addition of HR-HPV nucleic acid testing to a cervical cytology screen- ing regimen offers higher sensitivity and negative predictive value (NPV) for detection of cervical precancer and cancer compared with cytology alone, especially in older women.11-15 For this reason, HR-HPV nucleic acid testing is recommended as an adjunctive test to cytology to assess the presence of HR-HPV types in women 30 years of age or older.16 In this context, HR-HPV testing guides patient management by iden- tifying women at elevated risk for CIN2 or higher (CIN2+) but, importantly, also reassures women who are negative for HR-HPV of their extremely low cancer risk.17-19

First-generation HR-HPV molecular tests used for adjunctive cervical cancer screening function by detecting viral genomic DNA in cellular samples from the uterine cer- vix. However, because the presence of HR-HPV in the female genital tract is common and often transient in nature,20,21

Reid et al / HPV Oncogenic mRNA Detection for Cervical Cancer Screening

474 Am J Clin Pathol 2015;144:473-483 DOI: 10.1309/AJCPHVD7MIP3FYVV

and most cervical HPV infections resolve without becoming cancerous,22,23 HR-HPV DNA-based test methods yield only moderate specificity for detection of high-grade cervical dis- ease.12,24 This leads to unnecessary follow-up and referral of patients to colposcopy, increasing the physical and emotional burdens on patients and elevating health care costs.

A test approved by the US Food and Drug Adminis- tration (FDA) for detection of HR-HPV E6/E7 messenger RNA (mRNA) (Aptima HPV [AHPV]; Hologic, San Diego, CA) has shown higher specificity with similar sensitivity for detection of CIN2+ compared with HPV DNA-based tests in patients referred for colposcopy due to an abnormal Pap smear result as well as in a screening setting.25-30 Expres- sion of mRNA from viral E6 and E7 oncogenes is highly associated with the development of CIN,31,32 and extensive investigation into the role of E6 and E7 oncoproteins in the human papillomavirus (HPV) life cycle has revealed that the expression of the corresponding oncogenes is necessary and sufficient for cell immortalization, neoplastic transforma- tion, and the development of invasive cancer.33-35

To confirm and extend the previous evidence on the clinical utility of HR-HPV oncogenic mRNA testing in a US population-based setting, the clinical performance of AHPV was evaluated as an adjunctive method for cervical cancer screening in women aged 30 years or older with negative for intraepithelial lesions or malignancy (NILM) cytology results from routine Pap testing in a pivotal, prospective, multicenter US clinical study including 3 years of follow-up (the Clinical Evaluation of Aptima mRNA [CLEAR] study). We report herein the results from this study.

Materials and Methods

Study Design, Conduct, and Participants The CLEAR study consisted of two parts: the ASC-

US (Atypical Squamous Cells of Unknown Significance) Study30 and the Adjunct Study described here ❚Figure 1❚. Women 30 years and older undergoing routine Pap testing who had a NILM cytology result were eligible to participate in the Adjunct Study and were recruited from 19 US family planning and obstetric/gynecologic clinics (private and aca- demic), family practice medical groups, and clinical research centers encompassing a wide geographic area representative of the US population. Informed consent was obtained prior to enrollment of participants. The study protocol was approved by institutional review boards at the participating centers, and the study was conducted in accordance with applicable regulatory requirements and good clinical practices.

Women were excluded from the study if they were pregnant, were vaccinated against HPV, had a history of

cervical disease (cancer or precancerous) or an abnormal Pap test result in the previous 12 months, or had a history of illness that could interfere with the study or create an unac- ceptable risk to the participant. Demographic information and relevant medical information (cervical cancer history, prior HPV diagnosis, and any abnormal cytology history) were collected from each participant. The study employed a baseline evaluation and a 3-year follow-up period with annual cytology visits for longitudinal disease ascertain- ment. Participants completed and exited the study once they had a CIN2+ diagnosis.

Cytology (Referral Pap) At the baseline evaluation and each annual visit there-

after, a cervical specimen was collected with a broom-like device (Papette; Wallach Surgical Devices, Orange, CT) or an endocervical brush and spatula (Cytobrush Plus GT and Pap Perfect Plastic Spatula; Medscand, Trumbull, CT) and placed into a ThinPrep Pap Test (Hologic) vial contain- ing PreservCyt Solution (“referral Pap” specimen). Pap specimens were processed locally using the ThinPrep 2000 System (Hologic) and evaluated for cytologic abnormalities. Cytology results were classified using the 2001 Bethesda System for reporting cervical cytology.36

HPV Testing Baseline PreservCyt specimens (1-mL aliquot) were

tested with the AHPV (Hologic) on both the automated Tigris DTS System and Panther System. Results from the two systems were similar; Panther System results are pre- sented here. AHPV is a target amplification assay that uses transcription-mediated amplification to detect the E6/E7 oncogene mRNA of 14 HR-HPV genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68). Three clinical laboratories each tested approximately one-third of all sam- ples with AHPV. Most of the PreservCyt specimens were also tested at one laboratory for HR-HPV DNA using the Hybrid Capture 2 assay (HC2; Qiagen, Gaithersburg, MD), an FDA-approved test that detects HR-HPV DNA of 13 of the 14 HR-HPV types detected by the AHPV and is known to cross-react with the 14th type.37 Testing and results inter- pretation of both HR-HPV tests were done according to the manufacturer’s instructions.38,39 Technicians performing HC2 and AHPV assays were masked to the other HPV test results and the participants’ clinical status and colposcopic/ histology results.

Disease Ascertainment At baseline, women who tested positive in either the

AHPV or the HC2 assay (HPV-positive women) and, to adjust for verification bias, approximately 6% of women who tested negative in both HPV assays (HPV-negative

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 475 DOI: 10.1309/AJCPHVD7MIP3FYVV

women) were randomly selected and referred to colposcopy. Most colposcopy visits (>60%) were completed within 16 weeks from the baseline visit (median, 14 weeks; interquar- tile range, 8 weeks).

Colposcopists were masked to HPV results and collect- ed cervical punch biopsy specimens from each visible lesion (“directed” biopsy) and an endocervical curettage (ECC) biopsy specimen. The biopsy specimens were processed

according to the normal site procedures to produce H&E- stained slides. After local pathologist review, slides were reviewed by two central panel pathologists (T.C.W. and M.H.S.) and classified using the three-tiered CIN terminol- ogy.40 Slides with discordant central panel diagnoses were reviewed by a third central pathologist to reach a consensus diagnosis (two out of three agreement). If agreement was not achieved, the three central panel pathologists reviewed

❚Figure 1❚ Clinical evaluation of Aptima mRNA study participant disposition. aReasons for withdrawal: did not meet eligibility criteria (70); Pap volume insufficient for AHPV testing (117); specimen expired or unsuitable for testing (190); specimen lost (58); noncompliant site (320); other reasons (26). bReasons for withdrawal: collection site did not participate in follow-up (243); subject terminated participation (37); participant had hysterectomy (22); participant not eligible (17); participant treated prior to CIN2+ diagnosis (8); other reasons (4). AHPV, Aptima HPV (Hologic, San Diego, CA); ASC-US, atypical squamous cells of unknown significance; CIN2+, cervical intraepithelial neoplasia grade 2 or higher; HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD); HPV, human papillomavirus; NILM, negative for intraepithelial lesions or malignancy; Pap, Papanicolaou test.

Enrollment (n = 13,495)

≥30 years (n = 12,869)

<30 years (n = 626)

Non–ASC-US (n = 6)

ASC-US (n = 620)

ASC-US (n = 731)

NILM cytology

(n = 11,641)

Adjunct study evaluable subjects

(n = 10,860)

Eligible for follow-up

(n = 10,509)

Year 1 Pap (n = 7,247) (9 CIN2+)

Completed (20 CIN2+)

Withdrawnb (n = 331)

Year 2 Pap (n = 6,517) (7 CIN2+)

Year 3 Pap (n = 6,339) (11 CIN2+)

Baseline coloscopy HPV+: 525 HPV–: 339

Withdrawna (n = 781)

HPV–, AHPV–, and HC2–

(n = 10,015) Randomly selected for coloscopy referral (n = 556)

HPV+, AHPV+, and/or HC2+

(n = 845) All referred

to coloscopy

Eligible for ASC-US study

(n = 1,351)

>ASC-US (270) unsatisfactory/

missing cytology (227)

(n = 497)

Reid et al / HPV Oncogenic mRNA Detection for Cervical Cancer Screening

476 Am J Clin Pathol 2015;144:473-483 DOI: 10.1309/AJCPHVD7MIP3FYVV

the slides in conference to reach consensus. A participant’s cervical disease status represents the highest grade consen- sus histology result from the colposcopy biopsy specimen. Review pathologists were masked to all other pathologists’ diagnoses, the participants’ clinical status, enrollment status (ASC-US Study or Adjunct Study), and HPV test results.

During follow-up, women with ASC-US or more severe cytology results were referred to colposcopy. Colposcopists collected the same types of biopsy specimens, which were processed and submitted to the same central pathology review as done at baseline to obtain the consensus histology result for that visit. Women with ASC-US or more severe cytology results who did not have a colposcopy were con- sidered to have indeterminate disease status. Participants with NILM cytology at a follow-up visit were not referred to colposcopy and were considered to have a normal cervix.

The final disease status after 3-year follow-up was determined for each participant who completed the study. To complete the study, a woman must have either (1) a consen- sus result of CIN2 or worse or (2) at least one cytology visit during the first or second year of follow-up and one cytology visit during the third year of follow-up, including colposco- pies for those with ASC-US or more severe cytology. Final disease status for women meeting the second criterion was based on their final consensus histology result or they were considered to have a normal cervix if they had NILM cytol- ogy at the last visit. Women with CIN2 or worse did not have further follow-up in the study.

Participants with an ASC-US or more severe cytology during follow-up who did not have a colposcopy or who attended the colposcopy visit but biopsy specimens were not collected, were lost, or the slides were inadequate to determine disease status were classified as indeterminate for cervical disease status.

Statistical Analysis Test performance was evaluated with participants hav-

ing a consensus histology result of CIN2+ (CIN2, CIN3, carcinoma in situ, or invasive cancer) classified as positive for cervical disease. A diagnosis of CIN1 or normal disease status classified participants as negative for cervical disease. In addition, test performance was evaluated using a more definitive disease end point where a consensus histology result of CIN3+ (CIN3, carcinoma in situ, or invasive can- cer) classified participants as positive for cervical disease, and CIN2, CIN1, or normal classified participants as nega- tive for cervical disease.

For the baseline risk analysis, a disproportionately smaller subset (3.4%) of HPV-negative vs HPV-positive women had disease status determined from the baseline colposcopy visit, resulting in verification bias. To adjust for this bias, a multiple imputation method41 was used to impute

missing disease status based on the observed consensus histology results and AHPV and HC2 assay results from women who had a baseline colposcopy. Verification bias- adjusted risk estimates and 95% confidence intervals (CIs) were generated using these imputed results.

The follow-up risk analysis included disease identified at baseline and during follow-up. Because all women did not have a colposcopy at baseline, individual by-year risk estimates may reflect disease that was either present but not detected at baseline or incident or progressive disease. Cumulative risks with 95% CIs were generated using the life table method, with participants not completing the study censored after the follow-up year last attended.

Sensitivity and specificity estimates with 95% CIs were generated including women who completed the study. The McNemar exact test of discordant matched pairs was per- formed to compare the assays, including only participants with results for both assays. All statistical tests and 95% CIs were two-tailed and performed at the 5% significance level, using SAS version 9.1 or higher (SAS Institute, Cary, NC).

Results

Participant Disposition and Demographic Information A total of 13,495 women were included in this clinical

study (Figure 1). Of the 12,869 women 30 years or older, 227 had an unsatisfactory or missing cytology result and 1,001 had an abnormal Pap result: ASC-US (5.7%), low-grade squamous intraepithelial lesion (1.5%), high-grade squa- mous intraepithelial lesion (0.2%), ASC-H (0.1%), atypical granular cells, atypical granular cells favor neoplastic, or

❚Table 1❚ Demographics of Evaluable Participants (n = 10,860) Characteristic Value

Age, y Mean 44.2 Median 43 Minimum-maximum 30-89 IQR 15 Age groups, No. (%), y 30 to <40 4,192 (38.6) ≥40 6,668 (61.4) Race/ethnicity, No. (%) Non-Hispanic white 4,774 (44.0) Hispanic white 1,814 (16.7) Black 1,354 (12.5) Asian 622 (5.7) Othera 488 (4.5) Unknown 1,808 (16.6)

IQR, interquartile range. a Other includes American Indian, Alaska Native, Native Hawaiian, Pacific Islander,

and multiple races.

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 477 DOI: 10.1309/AJCPHVD7MIP3FYVV

“other” (0.2% combined prevalence). The remaining 11,641 women 30 years or older with a NILM cytology at baseline were enrolled in the Adjunct Study and tested with the AHPV and HC2 tests. In total, 10,860 women were available for the baseline analysis, including 864 women (525 HPV+ and 339 HPV–) with a baseline colposcopy (781 women withdrew; see Figure 1 for reasons). Approximately 50% of the women referred to colposcopy had an ECC biopsy only, and approximately 50% had ECC plus one or more directed biopsies, resulting in the identification of 20 cases of CIN2+.

After the baseline evaluation, 10,509 women were eligi- ble for follow-up (331 women withdrew for various reasons; see Figure 1). During follow-up, 7,247 women returned for an annual cytology visit during year 1, 6,517 returned during year 2, and 6,339 returned during year 3, with 6,201 women completing the study. Of the women who completed the study, 4,452 returned during all 3 years; the remaining returned only once during the first 2 years and in year 3 or had CIN2+ and exited the study prior to year 3. In each follow-up year, 4% to 6% of the women had ASC-US or greater cytology.

Demographics are presented in ❚Table 1❚. The median age was 43 years, with 61.4% age 40 years or older; 44.0% were non-Hispanic white, 16.7% were Hispanic white,

12.5% were black, 5.7% were Asian, and 21.1% were cat- egorized as “other” race or unknown.

HPV and Disease Prevalence Cervical disease and HPV status are shown in ❚Table 2❚

for the baseline evaluation and cumulatively after 3 years of follow-up. Of the 10,860 evaluable participants with NILM cytology at baseline, 512 were positive for AHPV, yield- ing a prevalence of 4.7% for HR-HPV E6/E7 oncogenic mRNA, whereas prevalence of HR-HPV DNA was 6.5% among 10,229 women with HC2 results. A total of 845 HPV RNA-positive or DNA-positive women and 556 randomly selected HPV-negative women were referred to colposcopy at baseline (Figure 1).

At baseline, the percentage of colposcopy attendance was similar between HPV-positive (62%, n = 526) and ran- domly selected HPV-negative (61%, n = 339) women with 29 cases of CIN1, nine cases of CIN2, eight cases of CIN3, and three cases of adenocarcinoma in situ (AIS) identified (Table 2). Four of the CIN2 cases and two of the AIS cases were identified based on an ECC biopsy specimen only.

In total, 6,271 women completed the 3-year follow- up with a known disease status (Table 2). Of these, 6,098

❚Table 2❚ Disease Status at Baseline and After 3 Years of Follow-up and Corresponding AHPV and HC2 Test Results at Baseline

Participants at Baseline, No.

AHPV+, No. AHPV–, No.

HC2+ HC2– HC2 Missinga HC2+ HC2– HC2 Missinga

Disease status at baselineb

Total No. 10,860 383 97 32 282 9,467 599 Verified Normal 769 211 19 12 170 353 4 CIN1 29 12 0 1 7 9 0 CIN2 9 4 0 0 2 2 1 CIN3 8 7 0 0 1 0 0 AIS 3 2 1 0 0 0 0 CIN2+ 20 13 1 0 3 2 1 CIN3+ 11 9 1 0 1 0 0 Unverified 10,042 147 77 19 102 9,103 594 Disease status after 3-y follow-upc

Total No. 10,843d 383 97 32 281 9,452 599 Normal 6,098 161 48 10 123 5,440 316 CIN1 56 10 0 0 6 36 4 CIN2 24 7 0 1 3 12 1 CIN3 20 14 0 1 2 3 0 AIS 3 2 1 0 0 0 0 CIN2+ 47 23 1 2 5 15 1 CIN3+ 23 16 1 1 2 3 0 Missing 4,378 167 44 17 130 3,756 264 Indeterminate 264 21 4 3 17 205 14

AHPV, Aptima HPV (Hologic, San Diego, CA); AIS, adenocarcinoma in situ; CIN1, cervical intraepithelial neoplasia grade 1; CIN2, cervical intraepithelial neoplasia grade 2; CIN2+, CIN2 or higher; CIN3, cervical intraepithelial neoplasia grade 3; CIN3+, CIN3 or higher; HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD).

a In total, 631 women with AHPV assay results did not have HC2 test results primarily due to insufficient volume of the cytology specimen. b Verified disease status was determined for women who attended colposcopy at baseline and had a consensus histology result. Women without a consensus histology result

have an unverified disease status. c Disease status after 3-year follow-up is based on completing 3-year follow-up with cytology performed at least once and colposcopy attendance for ASC-US or higher results

during the first 2 years and during the third year. d Seventeen women were determined ineligible after completion of baseline; results are excluded from follow-up analyses.

Reid et al / HPV Oncogenic mRNA Detection for Cervical Cancer Screening

478 Am J Clin Pathol 2015;144:473-483 DOI: 10.1309/AJCPHVD7MIP3FYVV

(97.2%) women had normal (negative) disease status, and 56 (0.9%) had low-grade lesions (CIN1). In addition to the 20 women with CIN2+ identified at baseline, 15 (0.2%) women had CIN2 and 12 (0.2%) women had CIN3 identified during follow-up, with two cases identified from an ECC biopsy specimen only.

Of the 27 women with CIN2+ identified during follow- up, two had CIN1 at baseline, with CIN3 identified during year 1. Ten women had no disease found at baseline, with five cases of CIN2+ identified during year 1, one case of CIN2+ identified during year 2, and four cases of CIN2+ identified during year 3. The remaining 15 women with CIN2+ identified during follow-up did not have a baseline colposcopy; among them, two cases of CIN2+ were identi- fied during year 1, six cases of CIN2+ during year 2, and seven cases of CIN2+ during year 3.

AHPV Assay Performance Baseline risk and prevalence estimates adjusted for

verification bias are provided in ❚Table 3❚. The prevalence of

CIN2+ was 0.9% in the overall population. CIN2+ occurred in 4.5% (95% CI, 2.7%-7.4%) of women with positive AHPV results and in 0.6% (95% CI, 0.2%-1.9%) of women with negative AHPV results, yielding a relative risk of 7.5 (95% CI, 2.1-26.3). This indicates that women with a posi- tive AHPV result are at significantly greater risk of CIN2+ than women with a negative AHPV result. The CIN2+ rela- tive risk obtained for the HC2 test at baseline was similar (7.3; 95% CI, 1.6-33.5). For CIN3+ diagnosis, the overall prevalence was 0.4%. The AHPV relative risk was 24.9 (95% CI, 2.0-307.0), again with a similar relative risk for HC2 (21.0; 95% CI, 1.0-423.8).

Cumulative absolute and relative risks for AHPV and HC2 over the 3-year follow-up period for HPV-positive and HPV-negative women are shown in ❚Table 4❚. Women with an HPV-negative result with either test had very low cervical disease risk after 3 years of follow-up (<0.3%). Comparatively, 5% to 6% of women with an HPV-positive result had CIN2+ and 3% to 4% had CIN3+, with overall cumulative absolute and relative risks slightly higher for

❚Table 4❚ Cumulative Absolute and Relative Risk of CIN2+ and CIN3+ Disease by Age Group After 3-Year Follow-up (Life Table Analysis)

Disease Status/Age Group, y Assay Result

AHPV Assay HC2 Test

Absolute Risk (95% CI), %

Relative Risk (95% CI)

Prevalence, %

Absolute Risk (95% CI), %

Relative Risk (95% CI)

Prevalence, %

CIN2+ Overall Positive 6.32 (4.29-9.27) 23.94 (13.59-42.18) 0.55 5.12 (3.53-7.41) 22.39 (12.19-41.12) 0.55

Negative 0.26 (0.17-0.41) 0.23 (0.14-0.38) 30-39 Positive 7.76 (4.81-12.40) 31.11 (13.04-74.21) 0.76 6.46 (3.99-10.39) 27.36 (10.88-68.80) 0.79

Negative 0.25 (0.12-0.53) 0.24 (0.10-0.54) ≥40 Positive 4.51 (2.34-8.63) 16.57 (7.26-37.82) 0.42 3.77 (2.10-6.71) 16.85 (7.21-39.35) 0.40

Negative 0.27 (0.16-0.46) 0.22 (0.12-0.42) CIN3+ Overall Positive 4.42 (2.76-7.03) 67.87 (25.32-181.88) 0.27 3.43 (2.14-5.48) 59.14 (20.09-74.12) 0.28

Negative 0.07 (0.03-0.16) 0.06 (0.02-0.16) 30-39 Positive 5.74 (3.22-10.11) 102.84 (23.17-456.51) 0.44 4.78 (2.67-8.48) 171.50 (22.39-1,313.63) 0.45

Negative 0.06 (0.01-0.22) 0.03 (0.00-0.20) ≥40 Positive 2.81 (1.27-6.16) 41.80 (10.53-166.00) 0.16 2.05 (0.93-4.52) 28.46 (7.15-113.20) 0.17

Negative 0.07 (0.02-0.21) 0.07 (0.02-0.22)

AHPV, Aptima HPV (Hologic, San Diego, CA); CI, confidence interval; CIN2+, cervical intraepithelial neoplasia grade 2 or higher; CIN3+, cervical intraepithelial neoplasia grade 3 or higher; HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD).

❚Table 3❚ Absolute and Relative Risk of CIN2+ and CIN3+ Disease at Baseline (Verification Bias Adjusted)

Disease Status/Assay Result

AHPV Assay HC2 Test

Absolute Risk (95% CI), %

Relative Risk (95% CI)

Prevalence, %

Absolute Risk (95% CI), %

Relative Risk (95% CI)

Prevalence, %

CIN2+ Positive 4.5 (2.7-7.4) 7.5 (2.1-26.3) 0.9 3.7 (2.3-6.1) 7.3 (1.6-33.5) 0.9 Negative 0.6 (0.2-1.9) 0.5 (0.1-2.1) CIN3+ Positive 3.0 (1.6-5.5) 24.9 (2.0-307.0) 0.4 2.3 (1.3-4.1) 21.0 (1.0-423.8) 0.4 Negative 0.1 (0.0-1.7) 0.1 (0.0-2.4)

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 479 DOI: 10.1309/AJCPHVD7MIP3FYVV

the AHPV assay than for HC2. Younger women aged 30 to 39 years who were HPV positive had twice the prevalence of disease but a similar increase in relative risk of cervical disease compared with HPV-positive women 40 years and older (Table 4). Risk of cervical disease in HPV-negative women did not vary by age group.

❚Figure 2❚ and ❚Figure 3❚ show the cumulative absolute risk of CIN2+ and CIN3+, respectively, by year according to AHPV or HC2 positivity status at baseline. Both assays show a similar trend, with consistent slightly higher risk for the AHPV assay each year.

After 3 years of follow-up, the specificity of AHPV for CIN2 or lower was 96.3% (95% CI, 95.8%-96.7%), signifi- cantly greater (P < .001) compared with HC2 specificity of 94.8% (95% CI, 94.3%-95.4%) ❚Table 5❚. AHPV specificity for CIN3 or lower (96.2%; 95% CI, 95.5%-96.5%) was also significantly greater (P < .001) than HC2 specificity (94.7%; 95% CI, 94.1%-95.2%). Estimated sensitivities for detection of CIN2+ and CIN3+ were similar between the two assays (P = .219 and P = 1.0, respectively). For detection of CIN2+, AHPV sensitivity was 55.3% (95% CI, 41.2%-68.6%), and HC2 sensitivity was 63.6% (95% CI, 48.9%-76.2%). For CIN3+ detection, AHPV sensitivity was 78.3% (95% CI, 58.1%-90.3%), and HC2 sensitivity was 81.8% (95% CI, 61.5%-92.7%) (Table 5).

Discussion

This study presents the results of a 3-year longitudinal evaluation of the AHPV assay as an adjunctive method for screening women 30 years and older who have NILM Pap cytology results. Consistent with previously published data,28,29 these results demonstrate that HR-HPV oncogenic E6/E7 mRNA testing has a sensitivity similar to an HR-HPV DNA-based test for detection of CIN2+ and CIN3+ and

slightly, but significantly, improved specificity compared with HR-HPV DNA testing for both end points. We found that use of AHPV as an adjunctive method for HPV-induced cervical disease screening provided disease detection capa- bility similar to HC2 while reducing the false-positive rate (from 5.2% to 3.7%) relative to the HPV DNA-based test. Reduction of HPV detection in women without cervical disease minimizes the anxiety and burden associated with spurious positive HPV molecular test results in women with NILM cytology, decreases health care costs, and reduces unnecessary follow-up procedures, thereby improving the safety of cervical cancer screening (unnecessary colposcopy is considered a significant “harm” in the recent American Cancer Society guidelines16).

Importantly, we show that women with a NILM cytol- ogy result who also had a positive AHPV result are approxi- mately 24 times more likely to have CIN2+ disease after 3 years than women with a negative AHPV result. This risk increased to approximately 68-fold for detection of CIN3+ disease. Similar but slightly lower risk estimates were obtained with HC2, demonstrating comparable accuracy of the AHPV and HC2 for identifying participants with CIN2+ and CIN3+ in this respect.

After 3 years of follow-up, women in this study who were HPV negative at baseline using any test method had very low risk for CIN2+ (<0.3%), a result similar to previously published studies with HC2.42,43 These findings reinforce evi- dence from previous studies showing that HR-HPV nucleic acid testing should be performed as an adjunctive test to routine Pap for cervical cancer screening of women 30 years or older to increase sensitivity of disease detection.28 Corre- spondingly, compared with annual cytology-only screening, this study supports longer screening intervals for women negative for both abnormal cytology and HPV E6/E7 mRNA, due to the high NPV and low risk of disease afforded by this screening algorithm for 3 years following a test-negative

0 0 1 2 3

1 2 3 4

Year of Follow-Up

A b

s o

lu te

R is

k f

o r

C IN

3 +

AHPV+ AHPV– HC2+ HC2–5

6 7 8

❚Figure 3❚ Cumulative absolute risk of cervical intraepithelial neoplasia grade 3 or higher (CIN3+) by year. AHPV, Aptima HPV (Hologic, San Diego, CA); HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD).

0 0 1 2 3

1 2 3 4

Year of Follow-Up

A b

s o

lu te

R is

k f

o r

C IN

2 +

AHPV+ AHPV– HC2+ HC2–

5 6 7 8 9

❚Figure 2❚ Cumulative absolute risk of cervical intraepithelial neoplasia grade 2 or higher (CIN2+) by year. AHPV, Aptima HPV (Hologic, San Diego, CA); HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD).

Reid et al / HPV Oncogenic mRNA Detection for Cervical Cancer Screening

480 Am J Clin Pathol 2015;144:473-483 DOI: 10.1309/AJCPHVD7MIP3FYVV

baseline visit. Extension of cervical cancer screening intervals following negative HPV and cytology test results in women 30 years or older is a key recommendation of current US screening guidelines from both the American Cancer Society and the US Preventive Services Task Force.16

Conversely, since the positive predictive value of any HPV test in women with NILM cytology is low, additional AHPV testing to detect persistent HR-HPV infection dur- ing follow-up care in women with an initial AHPV-positive result is likely a better option than direct referral to colpos- copy. Alternatively, genotyping with referral for HPV 16– or HPV 18–positive women can optimize referral and minimize loss to follow-up.44

Several design features were employed in the CLEAR study to achieve accurate determination of the performance characteristics for both AHPV and HC2 assays. First, all biopsy samples were subjected to adjudicated review by three independent expert pathologists. Second, molecular test performance was compared with a consensus histology diagnosis, the gold standard for determining cervical disease status. Third, AHPV performance was compared directly with HC2 performance, the most broadly used and charac- terized HPV DNA test. Fourth, performance characteristics of both assays obtained from baseline results were adjusted for verification bias by conducting colposcopy and biopsy in

3.4% of HPV-negative women. This process is recommend- ed in low-prevalence populations to avoid overestimating assay sensitivity and underestimating assay specificity.45-47 Finally, women were followed for 3 years with annual cytol- ogy testing and referral to colposcopy for abnormal results.

A limitation of this study was that a portion of HPV- negative women with normal cytology were subjected to colposcopy and biopsy at the baseline visit but not at the subsequent follow-up visits. Thus, the relative risk estimates reported here for disease in HPV-positive vs HPV-negative women evaluated during years 1, 2, and 3 of the follow-up period may be overstated. This potential bias is present in previously reported longitudinal cotesting studies17,19,48 and is unavoidable, since implementation of invasive procedures on thousands of women with normal cytology and negative HPV test results presents a burden to study participants and is not supported by current US and European practice guide- lines. However, as in previous longitudinal cotesting studies, women enrolled in CLEAR who exited at the final (third) year of follow-up had yielded negative cytology and/or negative HC2 and AHPV results from four consecutive examinations. Thus, their risk of harboring an occult CIN lesion is likely to be exceedingly small,42,43 such that any potential error encountered here most likely constitutes a very small fraction of the overall magnitude of the risks reported.

❚Table 5❚ Clinical Sensitivity and Specificity for CIN2+ and CIN3+ Disease After 3-Year Follow-upa

HC2

AHPV, No. Sensitivity/Specificity, % [No./Total No.] (95% CI)b Difference (95% CI) P ValuePositive Negative Total AHPV HC2

CIN2+ 55.3 [26/47] (41.2 to 68.6) 63.6 [28/44] (48.9 to 76.2) –9.1 (–21.9 to 3.8) .219 Positive 23 5 28 Negative 1 15 16 Missing/equivocal 2 1 3 Total 26 21 47 <CIN2 96.3 [5,925/6,154]

(95.8 to 96.7) 94.8 [5,524/5,824] (94.3 to 95.4)

1.4 (0.9 to 1.9) <.001

Positive 171 129 300 Negative 48 5,476 5,524 Missing/equivocal 10 320 330 Total 229 5,925 6,154 CIN3+ 78.3 [18/23] (58.1 to 90.3) 81.8 [18/22] (61.5 to 92.7) –4.5 (–24.4 to 15.3) 1.000 Positive 16 2 18 Negative 1 3 4 Missing/equivocal 1 0 1 Total 18 5 23 <CIN3 96.2 [5,941/6,178]

(95.5 to 96.5) 94.7 [5,536/5,846] (94.1 to 95.2)

1.4 (1.0 to 1.9) <.001

Positive 178 132 310 Negative 48 5,488 5,536 Missing/equivocal 11 321 332 Total 237 5,941 6,178

AHPV, Aptima HPV (Hologic, San Diego, CA); CI, confidence interval; CIN2+, cervical intraepithelial neoplasia grade 2 or higher; <CIN2, cervical intraepithelial neoplasia grade 2 or lower; CIN3+, cervical intraepithelial neoplasia grade 3 or higher; <CIN3, cervical intraepithelial neoplasia grade 3 or lower; HC2, Hybrid Capture 2 (Qiagen, Gaithersburg, MD).

a Differences (95% CI) and P values (McNemar exact test) are calculated including only women with both AHPV and Digene HC2 assay results (excluding samples with missing or equivocal Digene HC2 results).

b Sensitivity is reported for CIN2+ and CIN3+; specificity is reported for <CIN2 and <CIN3.

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 481 DOI: 10.1309/AJCPHVD7MIP3FYVV

Another limitation of this study was that colposcopists were aware of the women’s HPV test status during the first half of the baseline portion of the study, because during that period, only women who tested positive in the AHPV or HC2 were referred to colposcopy. When the colposcopists were unmasked, they may have been more diligent to find cervical disease with prior knowledge of current HPV infec- tion status. However, after randomly selected HPV-negative women were referred to colposcopy, the colposcopists were masked to HPV status, and throughout the entire study, colposcopists were masked as to which HPV assay caused the referral. Thus, any potential “colposcopy bias” would be identical for both molecular tests.

In summary, these results demonstrate that the clinical performance of HR-HPV E6/E7 mRNA testing using the AHPV is consistent with current US cervical cancer screen- ing guidelines for women with a NILM cytology result who are 30 years of age or older. There was a significantly greater risk of CIN2+ in AHPV-positive vs AHPV-negative participants, as well as a statistically and clinically signifi- cant improvement in specificity for detection of CIN2+ by the AHPV compared with HPV DNA testing with the HC2 assay. Thus, these data confirm the clinical utility of AHPV testing in an adjunct cervical cancer screening setting.

Corresponding author: Damon Getman, PhD, Hologic, 10210 Genetic Center Dr, San Diego, CA, 92121; damon.getman@ hologic.com.

CLEAR Study Central Pathology Review Panel: Thomas C. Wright Jr., MD, Department of Pathology, Columbia University School of Medicine, New York, NY; Mark Stoler, MD, Department of Pathology, University of Virginia Health System, Charlottesville, VA; and Alex Ferenczy MD, BioVision, Incorporated, Outremont, Canada.

The CLEAR study testing sites and participants are as follows: Laboratory testing sites: B. A. Body, Laboratory Corporation of America, Burlington, NC; B. Yen-Lieberman, Cleveland Clinic Foundation, Cleveland, OH; C. Ginocchio, North Shore Long Island Jewish Health System Laboratory, Lake Success, NY; and T. Davis, Eskenazi Health Services, Indianapolis, IN. Collection sites (principal investigators and institutions): M. Rabinowitz, Community Medical Research, Miami Beach, FL; A. T. Hanna, Diverse Research Solutions, Oxnard, CA; K. Ault, Emory University School of Medicine, Atlanta, GA; G. M. Grossman, HealthCare Partners, Los Angeles, CA; W. Campbell, HealthCare Partners, Torrance, CA; M. Stripling, NEA Baptist Clinic Women’s Clinic, Jonesboro, AR; D. Ronk, Planned Parenthood of Eastern Arkansas and Oklahoma, Tulsa, OK; R. Muckerman, PPS Clinical Research, Chesterfield, MO; B. Naghi, Peninsula Research Associates, Rolling Hills Estates, CA; T. Easter, REMEK Research, Montclair, CA; E. Guzman, Guzman MD Research, Downey, CA; A. Wagner, Saginaw Valley Medical Research, Saginaw, MI; C. Saffer MD, West Coast OB/GYN, San Diego, CA; R. Healy, Adams Patterson OB-GYN, Memphis, TN; and W. Wilkerson, Insignia Clinical Research, Tampa, FL.

This study was funded by Hologic.

Disclosures: P. E. Castle has received commercial HPV tests for research at a reduced or no cost from Roche, QIAGEN, Norchip, and MTM. He is a paid consultant for BD, GE Healthcare, Roche, Hologic, and Cepheid. He is compensated as a member of a Merck Data and Safety Monitoring Board for HPV vaccines. M. H. Stoler has been a consultant in clinical trial and HPV DNA test development for Merck, Roche Molecular Systems, Becton Dickinson, Qiagen, Hologic, and Ventana Medical Systems. T. C. Wright is a consultant and/or speaker for Hologic, Roche, Becton Dickinson, and Ikonisys. J. Cuzick is on advisory boards for Hologic, Roche, Qiagen, Abbott, Becton Dickinson, and Merck, and his institution receives research funding from these companies. J. L. Reid, J. Dockter, C. Giachetti, and D. Getman are employees of Hologic.

References 1. Jemal A, Bray F, Center M, et al. Global cancer statistics. CA

Cancer J Clin. 2011;61:69-90. 2. Sasieni PD, Cuzick J, Lynch-Farmery E. Estimating the efficacy

of screening by auditing smear histories of women with and without cervical cancer. Br J Cancer. 1996;73:1001-1005.

3. Gustafsson L, Ponten J, Zack M, et al. International incidence rates of invasive cervical cancer after introduction of cytological screening. Cancer Causes Control. 1997;8:755-763.

4. US Cancer Statistics Working Group. United States Cancer Statistics: 2004 Incidence and Mortality. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; 2007. www.cdc.gov/cancer/npcr/npcrpdfs/ US_Cancer_Statistics_2004_Incidence_and_Mortaliy.pdf. Accessed October 13, 2014.

5. Franco E, Syrjanen K, de-Wolf C, et al. New developments in cervical cancer screening and prevention. Geneva, Switzerland, June 17-19, 1996. Workshop. Cancer Epidemiol Biomarkers Prev. 1996;5:853-856.

6. Miller AB, Nazeer S, Fonn S, et al. Report on consensus conference on cervical cancer screening and management. Int J Cancer. 2000;86:440-447.

7. Stoler MH, Schiffman M; Atypical Squamous Cells of Undetermined Significance–Low-Grade Squamous Intraepithelial Lesion Triage Study (ALTS) Group. Interobserver reproducibility of cervical cytologic and histologic interpretations: realistic estimates from the ASCUS-LSIL Triage Study. JAMA. 2001;285:1500-1505.

8. Arbyn M, Bergeron C, Klinkhamer P, et al. Liquid compared with conventional cervical cytology: a systematic review and meta-analysis. Obstet Gynecol. 2008;111:167-177.

9. Syrjänen KJ, Shabalova IP, Ivanchenko O, et al. Reproducibility of classification and correction for verification bias as determinants of performance of Papanicolaou smear cytology in the screening setting: experience from the new independent states of the former Soviet Union cohort study. Acta Cytol. 2009;53:548-557.

10. Walboomers JM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol. 1999;189:12-19.

11. Kulasingam SL, Hughes JP, Kiviat NB, et al. Evaluation of human papillomavirus testing in primary screening for cervical abnormalities: comparison of sensitivity, specificity, and frequency of referral. JAMA. 2002;288:1749-1757.

Reid et al / HPV Oncogenic mRNA Detection for Cervical Cancer Screening

482 Am J Clin Pathol 2015;144:473-483 DOI: 10.1309/AJCPHVD7MIP3FYVV

12. Cuzick, JC, Clavel KU, Petry CJ, et al. Overview of the European and North American studies on HPV testing in primary cervical cancer screening. Int J Cancer. 2006;119:1095-1101.

13. Mayrand MH, Duarte-Franco E, Rodrigues I, et al; Canadian Cervical Cancer Screening Trial Study Group. Human papillomavirus DNA versus Papanicolaou screening tests for cervical cancer. N Engl J Med. 2007;357:1579-1588.

14. Ronco G, Giorgi-Rossi P, Carozzi F, et al; New Technologies for Cervical Cancer Screening Working Group. Results at recruitment from a randomized controlled trial comparing human papillomavirus testing alone with conventional cytology as the primary cervical cancer screening test. J Natl Cancer Inst. 2008;100:492-501.

15. Baseman JG, Kulasingam SL, Harris TG, et al. Evaluation of primary cervical cancer screening with an oncogenic human papillomavirus DNA test and cervical cytologic findings among women who attended family planning clinics in the United States. Am J Obstet Gynecol. 2008;199:26-28.

16. Saslow D, Solomon D, Lawson H, et al. American Cancer Society, American Society for Colposcopy and Cervical Pathology, and American Society for Clinical Pathology screening guidelines for the prevention and early detection of cervical cancer. Am J Clin Pathol. 2012;137:516-542.

17. Khan M, Castle PE, Lorincz AT, et al. The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice. J Natl Cancer Inst. 2005;97:1072-1079.

18. Ronco G, Giorgi-Rossi P, Carozzi F, et al; New Technologies for Cervical Cancer screening (NTCC) Working Group. Efficacy of human papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial neoplasia: a randomised controlled trial. Lancet Oncol. 2010;11:249-257.

19. Rijkaart DC, Berkhof J, Rozendaal L, et al. Human papillomavirus testing for the detection of high-grade cervical intraepithelial neoplasia and cancer: final results of the POBASCAM randomised controlled trial. Lancet Oncol. 2012;13:78-88.

20. Moscicki A-B, Ma Y, Jonte J, et al. The role of sexual behavior and human papillomavirus persistence in predicting repeated infections with new human papillomavirus types. Cancer Epidemiol Biomarkers Prev. 2010;19:2055-2065.

21. Goodman M, Shvetsov Y, McDuffie K, et al. Prevalence, acquisition, and clearance of cervical human papillomavirus infection among women with normal cytology: Hawaii Human Papillomavirus Cohort Study. Cancer Res. 2008;68:8813-8824.

22. Münger K, Baldwin A, Edwards KM, et al. Mechanisms of human papillomavirus–induced oncogenesis. J Virol. 2004;78:11451-11460.

23. Moscicki A-B, Ma Y, Wibbelsman C, et al. Rate of and risks for regression of CIN-2 in adolescents and young women. Obstet Gynecol. 2010;116:1373-1380.

24. Koliopoulos G, Arbyn M, Martin-Hirsch P, et al. Diagnostic accuracy of human papillomavirus testing in primary cervical screening: a systematic review and meta-analysis of non-randomized studies. Gynecol Oncol. 2007;104:232- 246.

25. Szarewski A, Ambroisine L, Cadman L, et al. Comparison of predictors for high-grade cervical intraepithelial neoplasia in women with abnormal smears. Cancer Epidemiol Biomarkers Prev. 2008;17:3033-3042.

26. Clad A, Reuschenbach M, Weinschenk J, et al. Performance of the Aptima high-risk human papillomavirus mRNA assay in a referral population in comparison with Hybrid Capture 2 and cytology. J Clin Microbiol. 2011;49:1071-1076.

27. Ratnam S, Coutlee F, Fontaine D, et al. Aptima HPV E6/E7 mRNA test is as sensitive as Hybrid Capture 2 assay but more specific at detecting cervical precancer and cancer. J Clin Microbiol. 2011;49:557-564.

28. Monsonego J, Hudgens MG, Zerat L, et al. Evaluation of oncogenic human papillomavirus RNA and DNA tests with liquid-based cytology in primary cervical cancer screening: the FASE study. Int J Cancer. 2010;129:691-701.

29. Wu R, Belinson SE, Du H, et al. Human papillomavirus messenger RNA assay for cervical cancer screening: the Shenzhen Cervical Cancer Screening Trial I. Int J Gynecol Cancer. 2010;20:1411-1414.

30. Stoler MH, Wright TC Jr, Cuzick J, et al. APTIMA HPV assay performance in women with atypical squamous cells of undetermined significance cytology results. Am J Obstet Gynecol. 2013;208:144-148.

31. zur Hausen H. Molecular pathogenesis of cancer of the cervix and its causation by specific human papillomavirus types. Curr Top Microbiol Immunol. 1994;186:131-156.

32. Sotlar K, Stubner A, Diemer D, et al. Detection of high-risk human papillomavirus E6 and E7 oncogene transcripts in cervical scrapes by nested RT-polymerase chain reaction. J Med Virol. 2004;74:107-116.

33. Klingelhutz AJ, Foster SA, McDougall JK. Telomerase activation by the E6 gene product of human papillomavirus type 16. Nature. 1996;380:79-82.

34. Schreiber K, Cannon R, Karrison T, et al. Strong synergy between mutant ras and HPV16 E6/E7 in the development of primary tumors. Oncogene. 2004;23:3972-3979.

35. Liu X, Clements A, Zhao K, et al. Structure of the human papillomavirus E7 oncoprotein and its mechanism for inactivation of the retinoblastoma tumor suppressor. J Biol Chem. 2006;281:578-586.

36. Solomon D, Davey D, Kurman R, et al. The Bethesda System 2001: terminology for reporting the results of cervical cytology. JAMA. 2002;287:2114-2119.

37. Castle P, Solomon D, Wheeler CM, et al. Human papillomavirus genotype specificity of Hybrid Capture 2. J Clin Microbiol. 2008;46:2595-2604.

38. Hologic. APTIMA HPV assay [package insert]. San Diego, CA: Hologic; 2008.

39. Qiagen Hybrid Capture® 2 High-Risk HPV DNA Test® [package insert]. Gaithersburg, MD: Qiagen; 2007.

40. Wright TC, Ronnett BM, Kurman RJ, et al. Precancerous lesions of the cervix. In: Kurman RJ, Ellenson LH, Ronnett BM, eds. Blaustein’s Pathology of the Female Genital Tract. 6th ed. New York, NY: Springer; 2011:193-252.

41. Little RJA, Rubin DB. Statistical Analysis With Missing Data. 2nd ed. New York, NY: John Wiley; 2002:85-89.

42. Sherman ME, Lorincz AT, Scott DR. Baseline cytology, human papillomavirus testing, and risk for cervical neoplasia: a 10-year cohort analysis. J Natl Cancer Inst. 2003;95:46-52.

43. Dillner J, Rebolj M, Birembaut P, et al; Joint European Cohort Study. Long term predictive values of cytology and human papillomavirus testing in cervical cancer screening: joint European cohort study. BMJ. 2008;337:a1754.

AJCP / Original Article

Am J Clin Pathol 2015;144:473-483 483 DOI: 10.1309/AJCPHVD7MIP3FYVV

44. Castle PE, Cuzick J, Stoler M, et al. Detection of human papillomavirus 16, 18, and 45 in women with ASC-US cytology and the risk of cervical precancer: results from the CLEAR HPV study. Am J Clin Pathol. 2015;143:160-167.

45. Zhou X. Correcting for verification bias in studies of a diagnostic test’s accuracy. Stat Methods Med Res. 1998;7:337-353.

46. Zhou X. Comparing accuracies of two screening tests in a two-phase study for dementia. Appl Stat. 1998;47:135-147.

47. Begg CB, Greenes RA. Assessment of diagnostic tests when disease is subject to selection bias. Biometrics. 1983;39:207-216.

48. Katki H, Kinney W, Fetterman B, et al. Cervical cancer risk for women undergoing concurrent testing for human papillomavirus and cervical cytology: a population-based study in routine clinical practice. Lancet Oncol. 2011;12:663-672.

Copyright of American Journal of Clinical Pathology is the property of American Society of Clinical Pathologists and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.