Objective
The objective of the study was to determine the sensitivity and specificity of the APTIMA human papillomavirus (AHPV) assay for high-grade cervical intraepithelial neoplasia (CIN) in women 21 years old and older with atypical squamous cells of undetermined significance (ASC-US) cytology.
Study Design
Women 21 years old and older with ASC-US cytology had colposcopy/biopsy and molecular human papillomavirus testing. Performance of the AHPV and Hybrid Capture 2 assays was compared with a clinical diagnosis of CIN grade 2, CIN grade 3, or adenocarcinoma in situ (CIN2 or greater).
Results
Of 939 evaluable subjects, CIN2 or greater and CIN3 or greater prevalence was 9.7% and 4.4%, respectively. AHPV sensitivity and specificity was 86.8% and 62.9% for CIN2 or greater detection and 90.2% and 60.2% for CIN3 or greater, respectively. AHPV had a similar sensitivity to Hybrid Capture 2 but a significantly higher specificity (62.9% vs 55.8%, P < .001) for CIN2 or greater detection.
Conclusion
Among women with ASC-US cytology, detection of high-risk human papillomavirus E6/E7 oncogenic messenger ribonucleic acid is an effective triage method for colposcopy referral.
Cervical cancer screening programs using cervical cytology (traditional Papanicolaou smear or liquid-based cytology [LBC]) have substantially reduced the mortality associated with cervical cancer. However, it is widely recognized that a single cytology-based screening test, be it a conventional Papanicolaou or a liquid-based cytology specimen, has relatively low sensitivity for detecting cervical precancer and cancer.
Almost all cervical cancers are caused by infection with 1 or more of 14 high-risk human papillomavirus (HR-HPV) types, leading to widespread recommendations for combining HPV molecular testing for these types with cytology to improve the sensitivity of screening programs. Furthermore, because of the improvement in disease detection offered by combining molecular HPV testing with cervical cytology, reflex molecular HPV testing has been recommended as the preferred approach for atypical squamous cells of undetermined significance (ASC-US) cytology triage by the American Society for Colposcopy and Cervical Pathology in its 2006 guidelines.
Such testing can be conveniently performed on the residual LBC specimen collected during the screening visit (which is stored for this purpose). This approach provides rapid results, is cost effective, and eliminates the need for women to return to the office or clinic to collect another cervical specimen for a repeat Papanicolaou or HPV test.
The first generation of Food and Drug Administration (FDA)–approved commercial HPV tests detect deoxyribonucleic acid (DNA) from most or all of the recognized 14 HR-HPV types. The specificity of these tests for the detection of high-grade cervical disease is governed by the natural history of HPV infection, which tends to be transient in nature and to resolve spontaneously without resulting in invasive cervical disease. Because the development of HPV-induced cervical cancer is dependent on activity of the viral E6 and E7 oncoproteins, tests that detect the corresponding E6/E7 oncogenic messenger ribonucleic acids (mRNAs) have the potential to be more specific than DNA-based tests for the detection of clinically significant disease. Studies comparing the E6/E7 oncogene mRNA-based APTIMA HPV assay (AHPV) with the HPV DNA-based Digene Hybrid Capture 2 (HC2) assay in screening and ASC-US triage settings have previously demonstrated the improved specificity of HPV ribonucleic acid (RNA)–based molecular testing for the detection of cervical disease, defined as a diagnosis of cervical intraepithelial neoplasia (CIN) grade 2 or worse (ie, CIN grade 2, CIN grade 3, adenocarcinoma in situ, or invasive cervical cancer). To confirm and extend these findings in a US population–based setting, the clinical performance of AHPV was evaluated in a pivotal, prospective, multicenter US clinical study (the CLEAR [Clinical Evaluation of APTIMA mRNA] study) for the triage of women with ASC-US cytology for colposcopy referral.
Materials and Methods
Study design, conduct, and participants
The CLEAR clinical study consisted of 2 parts: the ASC-US Study ( Figure ) and the Adjunct Study. Eligible women invited to participate were 21 years old or older who were undergoing routine Papanicolaou testing and who had an ASC-US cytology result. Women were recruited from 19 US family planning, and obstetric/gynecological clinics (private and academic), 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 subjects. 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, reported prior vaccination against HPV, had a history of cervical disease (cancer or precancerous) or an abnormal Papanicolaou test result in the previous 12 months, or had a history of illness that could interfere with the study or create an unacceptable risk to the subject. Demographic information and relevant medical information (cervical cancer history, prior HPV diagnosis, and any abnormal cytology history) were collected from each subject.
Cytology (referral Papanicolaou)
Cervical specimens were 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 Papanicolaou test vial containing PreservCyt (Hologic Inc, Bedford, MA) solution (referral Papanicolaou specimen). Referral Papanicolaou specimens were processed locally using the ThinPrep 2000 system (Hologic Inc) and evaluated for routine screening cytology. Cytology results were classified using the 2001 Bethesda System for reporting cervical cytology. Subjects meeting study selection criteria with ASC-US cytology results were referred for a colposcopy examination ( Figure ).
Disease ascertainment (colposcopy and biopsy/consensus histology)
Most colposcopy visits (>98%) were completed within 12 weeks from the initial visit (median, 4 weeks; interquartile range, 3 weeks). Colposcopists were masked to HPV results and were instructed to obtain 4 cervical punch biopsy specimens (1 specimen from each of 4 quadrants) and an endocervical curettage (ECC) biopsy specimen. Quadrants with visible lesion(s) were biopsied at the most severe area of any lesion; quadrant(s) without a visible lesion were biopsied at the squamocolumnar junction (random biopsy). Thus, each subject had 5 specimens for disease ascertainment.
The biopsy specimens were processed according to each site’s normal procedures to produce hematoxylin and eosin–stained specimen slides. After local pathologist review, slides were reviewed by 2 central panel pathologists using the 3-tiered CIN terminology. Slides with discordant central diagnoses were reviewed by a third central pathologist to reach a consensus diagnosis (2 of 3 agreements). If agreement was not achieved, the 3 panel pathologists reviewed the slides in conference to reach a consensus.
A subject’s cervical disease status was based on the highest-grade consensus histology result from colposcopy biopsy (clinical reference). Review pathologists were masked to all other pathologists’ diagnoses, the punch biopsy method, the subjects’ clinical status, enrollment status (ASC-US or Adjunct study), and HPV test results.
A subject with a consensus histology result of CIN2 or worse was considered positive for cervical disease. A subject who had CIN1 or a normal consensus histology result was considered negative for cervical disease. In addition, test performance was evaluated using a more definitive disease endpoint in which a subject with a consensus histology result of CIN3 or worse was considered positive for cervical disease and CIN2, CIN1, or normal considered negative for cervical disease. Subjects who attended the colposcopy visit were classified as indeterminate for cervical disease status if biopsies were not collected or were lost or if the slides were inadequate to determine disease status.
HPV testing
After the colposcopy visit was completed, PreservCyt specimens (1 mL aliquot) were tested with the AHPV assay (Gen-Probe Inc, San Diego, CA) on the automated TIGRIS DTS system (Gen-Probe Inc). AHPV is an isothermal 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 samples with AHPV. The majority of the PreservCyt specimens were also tested at 1 laboratory for HR-HPV DNA using the HC2 assay (Qiagen, Gaithersburg, MD), an FDA-approved test that detects HR-HPV DNA of 13 of the 14 HR-HPV types detected by AHPV. Testing and results interpretation of both HR-HPV tests were done according to the manufacturer’s instructions.
AHPV results were not used for determining treatment or patient care. Technicians performing HC2 and AHPV assays were masked to the other HPV test result and the subjects’ clinical status and colposcopic/histology results.
Statistical analysis
AHPV and HC2 performances were evaluated against the consensus histology diagnoses of CIN2 or worse and CIN3 or worse based on all biopsies (random, directed, and ECC) using sensitivity, specificity, positive and negative predictive values, and absolute and relative risk with 95% confidence intervals (CIs). Sensitivity and specificity were also evaluated against consensus histology diagnoses of CIN2 or worse and CIN3 or worse based only on the histology of the directed biopsies. Direct comparison of the sensitivity and specificity of the 2 assays was performed using McNemar’s test of discordant matched pairs including only subjects with results for both assays. Subjects with indeterminate cervical disease status were excluded from all performance analyses. Score CIs were used for sensitivity and specificity; exact CIs were used for positive predictive values, negative predictive values, and absolute risk; normal approximations were used for relative risk CIs. All statistical tests and CIs were 2 tailed and performed at the 5% significance level, using SAS version 9.1 or higher (SAS Institute, Cary, NC).
Results
Subject disposition and demographic information
A total of 13,495 subjects were included in this clinical study; 1351 had an ASC-US cytology result from routine Papanicolaou testing and were enrolled into the ASC-US study ( Figure ). Of the 1351 subjects, 99 were excluded due to age younger than 21 years and 313 were not evaluable for various reasons, as listed in the Figure . The demographic characteristics of the remaining 939 evaluable subjects are presented in Table 1 . The median age was 31.0 years (range, 21–71 years), with 44.2% of subjects being younger than 30 years of age. The majority of subjects were white (57.6%), 22.7% were Black or African American, and the remaining were of other/unknown race.
| Characteristic | All subjects | AHPV positive | AHPV negative |
|---|---|---|---|
| (n = 939) | (n = 394) | (n = 545) | |
| Age, y | |||
| Mean | 33.6 | 29.2 | 36.8 |
| Median | 31.0 | 27.0 | 36.0 |
| Minimum-maximum | 21–71 | 21–60 | 21–71 |
| IQR | 16 | 10 | 17 |
| Age groups, y | n (column percentage) | n (row percentage) | n (row percentage) |
| 21 to <30 | 415 (44.2%) | 250 (60.2%) | 165 (39.8%) |
| 30 to <40 | 262 (27.9%) | 95 (36.3%) | 167 (63.7%) |
| ≥40 | 262 (27.9%) | 49 (18.7%) | 213 (81.3%) |
| Race/ethnicity | |||
| White, not Hispanic | 434 (46.2%) | 182 (41.9%) | 252 (58.1%) |
| White, Hispanic | 107 (11.4%) | 33 (30.8%) | 74 (69.2%) |
| Black | 213 (22.7%) | 113 (53.1%) | 100 (46.9%) |
| Asian | 27 (2.9%) | 5 (18.5%) | 22 (81.5%) |
| Other a | 53 (5.6%) | 25 (47.2%) | 28 (52.8%) |
| Unknown | 105 (11.2%) | 36 (34.3%) | 69 (65.7%) |
a Other: American Indian, Alaska Native, Native Hawaiian, Pacific Islander, and multiple races.
HPV and disease prevalence
Of the 939 evaluable subjects with ASC-US, 394 were positive by AHPV, yielding a prevalence of 42.0% for HR-HPV oncogenic RNA. AHPV positivity rates decreased with increasing age and were 60.2% for subjects aged 21 to younger than 30 years, 36.3% for subjects aged 30 to younger than 40 years, and 18.7% for subjects aged 40 years or older ( Table 1 ).
Table 2 presents AHPV and HC2 results by the distribution of cervical disease status based on consensus histology review. Of the 939 evaluable subjects, 848 (90.3%) had a normal or CIN1 histology diagnosis, whereas 50 (5.3%) had CIN2, 40 (4.3%) had CIN3, and 1 (0.1%) had adenocarcinoma in situ. Thus, the prevalence of CIN2 or worse and CIN3 or worse cervical disease was 9.7% and 4.4%, respectively.
| Disease status a | All subjects (n = 939) | AHPV positive (n = 394) | AHPV negative (n = 545) | ||||
|---|---|---|---|---|---|---|---|
| HC2 positive | HC2 negative | HC2 miss b | HC2 positive | HC2 negative | HC2 miss b | ||
| Normal | 649 | 170 | 7 | 14 | 47 | 371 | 40 |
| CIN1 | 199 | 113 | 0 | 11 | 13 | 55 | 7 |
| CIN2 | 50 | 41 | 1 | 0 | 2 | 6 | 0 |
| CIN3 | 40 | 32 | 2 | 2 | 3 | 1 | 0 |
| AIS | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
| CIN2 or worse | 91 | 74 | 3 | 2 | 5 | 7 | 0 |
| CIN3 or worse | 41 | 33 | 2 | 2 | 3 | 1 | 0 |
a Disease status is based on the consensus histology result from all biopsies;
b Seventy-four women with AHPV results did not have HC2 results, primarily because of insufficient volume of the cytology specimen.
Sensitivity and specificity analysis
AHPV was positive for 79 of 91 specimens of CIN2 or worse, yielding a sensitivity of 86.8% (95% CI, 78.4–92.3%). For CIN3 or worse, AHPV was positive for 37 of the 41 specimens, yielding a sensitivity of 90.2% (95% CI, 77.5–96.1%). HC2 had sensitivity estimates of 88.8% (95% CI, 80.5–93.8%) and 92.3% (95% CI, 79.7–97.3%) for CIN2 or worse and CIN3 or worse, respectively ( Table 3 ).
| Disease status b | Assay | n | Prevalence, % | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) |
|---|---|---|---|---|---|---|---|
| CIN2 or worse | APTIMA | 939 | 9.7 | 86.8% (78.4–92.3%) | 62.9% (59.6–66.0%) | 20.1% (18.1–22.0%) | 97.8% (96.5–98.8%) |
| HC2 | 865 | 10.3 | 88.8% (80.5–93.8%) | 55.8% (52.3–59.3%) | 18.7% (17.0–20.4%) | 97.7% (96.2–98.8%) | |
| CIN3 or worse | APTIMA | 939 | 4.4 | 90.2% (77.5–96.1%) | 60.2% (57.0–63.4%) | 9.4% (8.1–10.4%) | 99.3% (98.3–99.8%) |
| HC2 | 865 | 4.5 | 92.3% (79.7–97.3%) | 53.3% (49.9–56.6%) | 8.5% (7.4–9.4%) | 99.3% (98.3–99.8%) |
a Includes directed, random, and endocervical curettage biopsy results;
b Disease status based on the consensus histology result from all biopsies.
Of the 848 subjects with consensus histology result of less than CIN2, 315 (37.2%) were positive by AHPV. Among 776 less than CIN2 subjects with HC2 results, 343 (44.2%) were positive by HC2. The specificity of AHPV for CIN2 or worse was 62.9% (95% CI, 59.6–66.0%) and for CIN3 or worse was 60.2% (95% CI, 57.0–63.4%). In comparison, HC2 had significantly lower specificity estimates for CIN2 or worse and CIN3 or worse: 55.8% (95% CI, 52.3–59.3%) and 53.3% (95% CI, 49.9–56.6%), respectively ( Table 3 ).
Of the 91 cases with CIN2 or worse observed in this ASC-US study, 60 (65.9%) were identified from directed biopsy specimens, 28 (30.8%) were identified from random biopsy specimens (not identified in directed biopsies), and an additional 3 (3.3%) were identified from ECC biopsy specimens only. Using results from directed biopsies only, sensitivity values for detection of CIN2 or worse and CIN3 or worse increased for both assays. For the detection of CIN2 or worse, sensitivities for both assays were similar (AHPV, 93.3%; HC2, 93.2%). Specificity values were slightly decreased (by approximately 1%) for both assays, although AHPV again had a significantly higher specificity than HC2 (AHPV, 61.5%; HC2, 54.5%) ( Table 4 ).
| Disease status a | Assay | n | Prevalence, % | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) |
|---|---|---|---|---|---|---|---|
| CIN2 or worse | APTIMA | 936 | 6.4 | 93.3% (84.1–97.4%) | 61.5% (58.3–64.7%) | 14.2% (12.7–15.6%) | 99.3% (98.3–99.8%) |
| HC2 | 863 | 6.8 | 93.2% (83.8–97.3%) | 54.5% (51.0–57.9%) | 13.1% (11.7–14.2%) | 99.1% (97.9–99.7%) | |
| CIN3 or worse | APTIMA | 937 | 3.1 | 93.1% (78.0–98.1%) | 59.6% (56.4–62.7%) | 6.9% (5.8–7.6%) | 99.6% (98.8–100%) |
| HC2 | 864 | 3.2 | 96.4% (82.3–99.4%) | 52.8% (49.4–56.1%) | 6.4% (5.5–7.0%) | 99.8% (98.9–100%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
