Objective
The objective of the study was to describe baseline data from Addressing the Need for Advanced HPV Diagnostics, a prospective, multicenter US cervical cancer screening trial.
Study Design
A total of 47,208 women aged 21 years or older undergoing routine screening were enrolled; liquid-based cytology and human papillomavirus (HPV) testing were performed. Women with abnormal cytology underwent colposcopy, as did high-risk HPV (hrHPV)–positive women and a random subset of women negative by both tests aged 25 years or older. Verification bias adjustment was applied; 95% confidence intervals were computed by the bootstrap method.
Results
The prevalence of cytologic abnormalities was 7.1%. hrHPV, HPV 16, and HPV 18 were detected using the cobas HPV Test in 12.6%, 2.8%, and 1.0% of women, respectively. Both cytologic abnormalities and hrHPV positivity declined with increasing age. The adjusted prevalence of cervical intraepithelial neoplasia grade 2 (CIN2) or greater in women aged 25-34 years was 2.3%, decreasing to 1.5% among older women.
Conclusion
The Addressing the Need for Advanced HPV Diagnostics study provides important estimates of the prevalence of cytologic abnormalities, hrHPV positivity, and CIN2 or greater in a US screening population.
Over the last 50 years, cytology-based cervical cancer screening has dramatically reduced the burden of invasive cervical cancer in the United States; whereas the incidence in the 1940s was estimated to be 32.6 per 100,000, today it is only 8.1 per 100,000. However, despite intensive cytologic screening, cervical cancer remains a significant cause of morbidity and mortality in the United States with more than 12,000 incident cases of cervical cancer annually and more than 4000 deaths. Moreover, approximately 500,000 women in the United States are diagnosed with high-grade cervical cancer precursors (cervical intraepithelial neoplasia grades 2 and 3 [CIN2, CIN3]) annually.
Cervical cancer is caused by infection with 1 of 14 high-risk types of human papillomavirus (hrHPV), with just 2 hrHPV genotypes (HPV 16 and HPV 18) causing approximately 70% of all cases. This has led to considerable interest in determining the optimal strategies for incorporating testing for hrHPV (14 pooled types) and genotyping for HPV 16 and HPV 18 into the US cervical cancer screening program to further reduce the burden of cervical disease. However, ensuring appropriate adoption of hrHPV testing into these strategies will require comprehensive assessments of the performance of cytology, hrHPV testing, and the burden of cervical disease in large US screening populations.
A recently initiated clinical trial, referred to as Addressing the Need for Advanced HPV Diagnostics (ATHENA), was designed to prospectively evaluate the performance of the cobas HPV Test, a new polymerase chain reaction–based deoxyribonucleic acid (DNA) amplification test that simultaneously identifies a pooled result for 12 hrHPV types (HPV 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) and individual results for HPV 16 and HPV 18. This trial evaluated 46,887 eligible women aged 21 years and older undergoing routine screening, of whom 8637 women underwent colposcopy, including a randomly selected subset of women aged 25 years and older who were negative by both Papanicolaou and hrHPV testing.
This manuscript describes the ATHENA study design and methods as well as the baseline characteristics of our study population, including the distribution of cytology results, hrHPV prevalence, and cervical disease status by age and HPV status.
Materials and Methods
Objectives
Specific objectives of the ATHENA HPV trial included determining the performance of the cobas HPV Test both as a triage test for women with abnormal cytology (atypical squamous cells of undetermined significance [ASC-US]) and as an adjunctive test to guide clinical management in women with cytology results negative for intraepithelial lesions or malignancies (NILM). A third objective was to evaluate the performance of the cobas HPV Test as a potential first-line test in the screening of women aged 25 years and older, regardless of cytology result.
Study design
The study is being conducted in 2 phases: a baseline (cross-sectional) phase and a 3 year follow-up (longitudinal) phase; data from only the baseline phase are reported here because the follow-up phase is ongoing and will be completed in December 2012. The process used to select women for colposcopy and biopsy based on age, HPV test result, and cytology result is shown in the Figure and described in detail below.
HPV tests used for subject selection were first-generation Roche HPV tests (AMPLICOR HPV test and LINEAR ARRAY HPV genotyping test; Roche Molecular Systems, Pleasanton, CA); all HPV results are based on the second-generation Roche HPV test (cobas HPV Test). The primary study endpoint for disease detection was high-grade cervical disease defined as CIN2 or greater (CIN2, CIN3, adenocarcinoma in situ, and invasive cervical cancer), as determined by a central pathology review panel (described in the following text), and the secondary study endpoint for disease detection was CIN3 or greater. Reporting of the study endpoints was based on the highest grade lesion identified by the central pathology review panel.
Sample size was determined by the need for a sufficient number of women with CIN2 or greater in the ASC-US population to adequately evaluate the performance of the cobas HPV Test. In accordance with the sample size in similar registration trials, it was determined that approximately 70 women with CIN2 or greater would be needed. This estimate was used, along with published rates of ASC-US cytology and HPV positivity in the overall population, to arrive at a sample size of approximately 45,000 women.
Participants were recruited from among women presenting for routine cervical cancer screening at 61 clinical sites across 23 states between May 2008 and August 2009. Clinical centers were predominantly general obstetrics and gynecology practices that routinely perform colposcopy. The inclusion and exclusion criteria are described in detail elsewhere.
The study was approved by Independent Investigational Review Board, Inc. (Plantation, FL) for the clinical sites and by Independent Investigational Review Board, Inc., the local institutional review board, or Copernicus Group Investigational Review Board (Research Triangle Park, NC) for the clinical laboratories. The study was conducted according to the International Conference on Harmonization Guideline for Good Clinical Practice.
Baseline phase (cross-sectional phase)
Participating women underwent 1 or 2 study visits at baseline, as follows.
Study visit 1 (enrollment visit [all participants])
After informed consent was obtained, a brief medical and the women’s obstetrics and gynecology history were taken. A speculum examination was then performed during which 2 cervical samples (A and B) were collected using a plastic spatula and cytobrush according to the manufacturer’s instructions and placed into 2 separate vials of PreservCyt solution (Hologic, Inc., Bedford, MA) ( Figure ). Sample A was processed for cytologic examination and HPV testing with the aforementioned Roche tests. Sample B was used to test for HPV DNA with the Hybrid Capture 2 assay according to the manufacturer’s instructions in women with ASC-US cytology (QIAGEN, Gaithersburg, MD) as well as for DNA sequencing in a subset of women selected for an HPV sequencing study (not reported here) and for long-term storage for future testing.
Study visit 2 (colposcopy visit [selected participants])
Prior to reporting screening test results back to the clinical sites, results were entered into a subject selection and randomization database that generated a subset of women selected for colposcopy. Selection/randomization was based on the results of cervical cytology and HPV testing with the first-generation AMPLICOR and LINEAR ARRAY tests (Roche).
This subset included all women aged 21 years or older with abnormal cervical cytology (ASC-US or greater), irrespective of HPV test results (n = 3259); women aged 25 years or older with NILM cervical cytology and a positive HPV test result by either of the first-generation HPV tests (n = 5726) and randomly selected women aged 25 years or older with NILM cytology who were negative for HPV by both first-generation HPV tests (n = 1041). Women who were not selected for colposcopy, or who decided to exit the study after the enrollment visit, were subsequently provided with the results of their enrollment cytology and HPV tests. The results of the cobas HPV Test were not used to select women for colposcopy because the test cutoff value had not been finalized at the start of enrollment into ATHENA.
Nonpregnant women selected for colposcopy underwent the procedure within 12 weeks of the enrollment visit. At the time of colposcopy, both study participants and colposcopists were blinded to cytology and HPV test results except, for safety reasons, in women with a cytologic diagnosis of cervical carcinoma or other malignant neoplasm. A standardized colposcopy protocol was followed as described in detail elsewhere and in the Supplemental Table . Women who met the primary clinical endpoint (CIN2 or greater by consensus pathology) exited the study.
Follow-up phase (3 year longitudinal follow-up)
Women who underwent colposcopy but did not meet the primary endpoint of CIN2 or greater by consensus pathology continued to the follow-up phase of the study (3 year longitudinal follow-up). Women diagnosed by the clinical laboratory with CIN2 or greater that was downgraded to less than CIN2 by consensus pathology were included in the follow-up phase. Women requiring additional procedures (eg, loop electrosurgical excision procedure, cervical conization) were managed according to standard of care at the clinical site. If available, cervical specimens collected during such treatment procedures were submitted for consensus pathology review.
During the follow-up phase (ongoing), women are being scheduled for annual follow-up examinations at years 1, 2, and 3. At each visit a liquid-based cytology (LBC) specimen (ThinPrep Papanicolaou test; Hologic, Inc, Bedford, MA) is obtained for cytology and cobas HPV testing. The residual specimen is stored for future testing. Nonpregnant women in whom cervical cytology is abnormal (ASC-US or greater) are referred for colposcopy with biopsy and/or endocervical curettage (ECC) according to the same protocol utilized during the baseline phase. Women found to have a diagnosis of CIN2 or greater will exit the study; those who do not will continue in the follow-up phase.
To optimize disease ascertainment at the end of the 3 year follow-up phase, an exit colposcopy and ECC will be offered to all nonpregnant women. This colposcopy will use the same protocol that was utilized at baseline with the exception that all participants will have an ECC.
Laboratory testing
Cytology and HPV testing
Cytology was conducted at four clinical laboratories and carried out as described in detail elsewhere ; cytologic evaluation was performed without computerized imaging. HPV testing was performed at these 4 laboratories and 1 additional laboratory. Cycle threshold cutoff values for the cobas HPV Test were established using samples from the first approximately 29,000 women enrolled; subsequent cross-validation of the test cutoff was achieved using samples from the remaining approximately 18,000 participants.
Consensus pathology review
The consensus pathology review panel consisted of 3 study pathologists blinded to all subject and laboratory information. Each biopsy and ECC was initially evaluated by 2 pathologists and reported using the 3 grades of CIN (CIN1, CIN2, CIN3) as well as adenocarcinoma in situ or carcinoma. If the diagnoses were concordant, it was recorded as the central pathology review panel diagnosis; if discordant, the biopsy/ECC was reviewed by the third study pathologist.
In cases in which all 3 diagnoses were discordant, the slides were reviewed in conference between the 3 pathologists to arrive at a consensus pathology diagnosis. Pathology specimens obtained at an unscheduled visit (a visit after study visit 2 for a gynecologic procedure or for a study colposcopy performed outside the 12 week window) could be used to determine the histologic stage of disease at baseline, provided the specimen was obtained within 28 days of the colposcopy at study visit 2. If more than 1 pathology specimen was obtained (either as biopsy or unscheduled visit specimen), the highest grade of disease was considered the consensus pathology diagnosis. Pathology results were categorized as CIN2 or greater, less than CIN2, CIN3 or greater, and less than CIN3 for determination of study endpoints as defined in Supplemental Figure 2 .
Statistical analyses
Prevalence estimates of Papanicolaou and HPV results were calculated based on all eligible women with valid Papanicolaou or HPV test results. Crude prevalence estimates of cervical disease were calculated based on women who underwent colposcopy/biopsy. The crude estimates of prevalence can result in bias because all women with positive Papanicolaou/HPV results were selected to undergo colposcopy, whereas only a small subset of women with negative test results were randomly selected to undergo colposcopy.
Verification bias adjustment was applied to account for the difference in rates of selection to colposcopy. This was accomplished by calculating the likely number of cases that would have been found if all women had undergone colposcopy and been disease verified.
In brief, the data were divided into strata of combined age group, Papanicolaou test results, and HPV test results. Disease prevalence in each stratum was assumed to be independent of whether the women underwent biopsy. Stratum-specific probabilities were then applied to the remainder of the women who had not undergone biopsy; this permitted an estimate of the number of cases that would have been found if all women had undergone colposcopy.
Verification bias-adjusted prevalence was calculated by collapsing strata by age groups. The 95% confidence intervals (CIs) were computed by bootstrap method with 1000 bootstrap samples. The 2.5th and 97.5th percentile of the bootstrap distribution of prevalence were used as the lower and upper limits of the 95% CIs.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
