Abstract
Detection of human papillomavirus (HPV) E6/E7 oncogene expression may be more predictive of cervical cancer risk than testing for HPV DNA. The Aptima HPV test (Gen-Probe) detects E6/E7 mRNA of 14 oncogenic types. Its clinical performance was compared with that of the Hybrid Capture 2 DNA test (HC2; Qiagen) in women referred for colposcopy and those routinely screened. Aptima was also compared with the PreTect HPV-Proofer E6/E7 mRNA assay (Proofer; Norchip) in the referral population. Cervical specimens collected in PreservCyt (Hologic Inc.) were processed for HPV detection and genotyping with the Linear Array (LA) method (Roche Molecular Diagnostics, Laval, Quebec, Canada). Histology-confirmed high-grade cervical intraepithelial neoplasia (CIN 2) or worse (CIN 2+) served as the disease endpoint. On the basis of 1,418 referral cases (CIN 2+, n = 401), the sensitivity of Aptima was 96.3% (95% confidence interval [CI], 94.4, 98.2), whereas it was 94.3% (95% CI, 92.0, 96.6) for HC2. The specificities were 43.2% (95% CI, 40.2, 46.2) and 38.7% (95% CI, 35.7, 41.7), respectively (P < 0.05). In 1,373 women undergoing routine screening (CIN 2+, n = 7), both Aptima and HC2 showed 100% sensitivity, and the specificities were 88.3% (95% CI, 86.6, 90.0) and 85.3% (95% CI, 83.5, 87.3), respectively (P < 0.05); for women ≥30 years of age (n = 845), the specificities were 93.9% (95% CI, 92.3, 95.5) and 92.1% (95% CI, 90.3, 93.9), respectively (P < 0.05). On the basis of 818 referral cases (CIN 2+, n = 235), the sensitivity of Aptima was 94.9% (95% CI, 92.1, 97.7) and that of Proofer was 79.1% (95% CI, 73.9, 84.3), and the specificities were 45.8% (95% CI, 41.8, 49.8) and 75.1% (95% CI, 71.6, 78.6), respectively (P < 0.05). Both Aptima and Proofer showed a higher degree of agreement with LA genotyping than HC2. In conclusion, the Aptima test is as sensitive as HC2 but more specific for detecting CIN 2+ and can serve as a reliable test for both primary cervical cancer screening and the triage of borderline cytological abnormalities.
Persistent infection with oncogenic human papillomavirus (HPV) is the underlying cause of cervical cancer (34, 42), and therefore, testing for oncogenic HPV infection could serve as an accurate means of detecting women at risk for cervical cancer. There are also indications that testing for HPV might be the most effective method of cervical cancer screening in developing countries (32). Moreover, HPV testing would be warranted as a primary screening tool in the era of HPV vaccination (13). Numerous studies have established that testing for HPV DNA is significantly more sensitive than Pap cytology for the detection of high-grade cervical intraepithelial neoplasia (CIN 2) or worse (CIN 2+, i.e., CIN 2, CIN 3, squamous cell carcinoma, endocervical adenocarcinoma in situ, and endocervical adenocarcinoma) (1, 4, 15, 20, 27, 30) and is recommended in primary cervical cancer screening and for the triage of borderline cytological abnormalities (33, 43, 44). However, HPV testing lacks specificity due to the ubiquitous and transient nature of HPV infection in women, and therefore, the positive predictive value (PPV) tends to be lower than that obtained by cytology (10, 15). The above observation nonetheless has been based on HPV DNA testing, with most studies utilizing the Hybrid Capture 2 DNA test (HC2; Qiagen) (10, 15). While HC2 is highly sensitive for the detection of 13 high-risk oncogenic types targeted by the test (10, 11, 15), it is also known to cross-react with untargeted nononcogenic types, thus potentially contributing to a reduction in the test's specificity (5, 28, 31).
The oncogenic process in cervical cancer is initiated and mediated by the upregulation of HPV E6/E7 oncoproteins, and thus, overexpression of these oncoproteins is a marker for an increased risk of cervical cancer (26, 38, 45). Therefore, detection of E6/E7 oncogene expression could be more specific and a better predictor of cervical cancer risk than the detection of HPV DNA, and E6/E7 oncogenic expression can be detected by testing for E6/E7 mRNA transcripts (38). There are presently two E6/E7 mRNA-based tests developed and commercialized, with the indication that the detection of E6/E7 mRNA could improve the specificity of HPV testing (12, 17, 18, 23).
The Aptima HPV assay (Gen-Probe) is a recently developed qualitative nucleic acid amplification test that detects E6/E7 mRNA collectively from 14 high-risk oncogenic types, HPV-16, -18, -31, -33, -35, -39, -45, -51, -52, -56, -58, -59, -66, and -68. There is indication that the Aptima test has clinical sensitivity similar to that of HC2 but improved clinical specificity over that of HC2, which is designed to detect DNA of the same HPV types as Aptima except for type 66 (12, 40). The PreTect HPV-Proofer test (Proofer; Norchip) is the other E6/E7 mRNA-based assay, which detects E6/E7 mRNA individually from five high-risk oncogenic types, HPV-16, -18, -31, -33, and -45, and this has been in use in Europe for some time. There have been several studies on the clinical performance of Proofer in Europe (16, 17, 18, 19, 22, 23, 24), with a single North American study in Canada (29). Proofer has been shown to be significantly more specific than both Aptima and HC2, but it lacks sensitivity for the detection of CIN 2+ (29, 40).
As there are a number of other HPV tests developed and currently available for comparative studies and clinical application, additional data on the relative clinical performance of Aptima would be useful. In this regard, a further assessment of the clinical performance of Aptima in comparison with that of HC2 is warranted, as the latter has been extensively validated and is recommended as a reference test to evaluate any newly developed HPV tests (21). Notwithstanding the differences between Aptima and Proofer, there have been limited data comparing the clinical performance of these two E6/E7 mRNA-based tests. Further, the relative performance of these tests from the standpoint of genotype specificity has not been analyzed using a standardized genotyping method. The aim of the present study was to assess the clinical performance of Aptima for the detection of CIN 2+ in comparison with that of HC2 and Proofer along with genotyping using a standardized commercially available method. This study was carried out utilizing cervical specimens from a random sample of women with a history of abnormal cytology referred for colposcopic assessment as well as those routinely screened, as indicated in the recently published guidelines for validating HPV tests (21). The data in this paper present the relative clinical sensitivity, specificity, and predictive values of Aptima for the detection of CIN 2+ on the basis of cross-sectional data.
(Preliminary data of this study were presented in part at the EUROGIN Meeting, Nice, France, 2008 [29a].)
MATERIALS AND METHODS
Study population.
The study population was comprised of a random sample of women referred for colposcopy for further assessment of cervical cancer risk and follow-up and those routinely screened. The colposcopy referral population consisted of women either newly diagnosed with abnormal Pap cytology of any grade who were referred for colposcopy or those with a history of abnormal cytology who were being followed up in colposcopy clinics according to the routine standard of care. The referral cases were enrolled from five tertiary care centers in five provinces across Canada for a longitudinal study. The prompting Pap test had been performed using conventional cytology at various sites served by the referral centers. Women 15 years of age or older who had had any grade of cytological abnormality within the previous 2 years and who had not received treatment were eligible. The routine screen participants were those receiving their normal Pap cytology screening in two centers. Those consenting to participate in the study signed a written informed consent. The study was approved by the institutional ethics review boards of all participating study centers.
The study population was comprised of 1,418 referral cases and 1,373 routinely screened women for the Aptima-HC2 study. The mean age of the referral group participants was 30.6 years (standard deviation [SD], 10.35 years; range, 15 to 80 years; median, 27 years). The proportion of women <30 years of age was 59%. The interval of time between the initial cytological diagnosis and enrollment at the colposcopy referral visit ranged from 1 to 3 months for new cases and 6 to 24 months for colposcopy follow-up cases with a history of abnormal cytology. The mean age of the routine screen participants was 36.3 years (SD, 13.75 years; range, 16 to 81 years; median, 35 years). In this population, the proportion of women <30 years of age was 38.3%. The Aptima-Proofer comparative study was limited to a proportion of the referral population and did not include the routine screen population.
Study procedures and testing methods.
Upon enrollment, a single cervical specimen was collected from all participants using a Cervex broom-type brush (Rovers Medical Devices, Oss, Netherlands) and suspended into PreservCyt collection medium (Hologic Inc., Marlborough, MA), according to the manufacturer's instructions. For the referral group, cervical specimens were taken immediately prior to colposcopic examination. All specimens were couriered to the Public Health Laboratory, St. John's, Newfoundland, Canada, for processing. Liquid-based cytology (LBC) was performed using the ThinPrep (Hologic Inc.) method in a central laboratory, and the results were reported according to the 2001 Bethesda system (37). Only the LBC results obtained at the time of enrollment were used for study purposes.
Residual PreservCyt samples from the referral group were tested simultaneously with Aptima, HC2, and Proofer, followed by HPV genotyping using the Linear Array (LA) HPV genotyping test. Aptima, HC2, and Proofer were carried out using fresh specimens within 2 weeks of collection, and LA genotyping was performed with frozen aliquots. Fresh residual PreservCyt samples from the routine screen population were tested simultaneously with Aptima and HC2. The procedures used for these tests are briefly summarized below. Researchers and their technologists performing these tests were blinded to results obtained in the other tests and also to cytology, colposcopy, and histology results.
Aptima.
Aptima (Gen-Probe Inc., San Diego, CA) detects HPV E6/E7 mRNA from 14 high-risk oncogenic types, HPV-16, -18, -31, -33, -35, -39, -45, -51, -52, -56, -58, -59, -66, and -68, collectively from liquid cervical specimens. The assay incorporates an internal control to control for nucleic acid capture, amplification, and detection, as well as operator or instrument error. An aliquot of 1 ml of each PreservCyt sample was transferred to 2.9 ml buffered detergent solution, and a 400-μl aliquot of the mixture was then tested on a semiautomated Direct Tube Sampling (DTS) system (Gen-Probe), according to the manufacturer's instructions. Assay results are interpreted on the basis of the signal-to-cutoff ratio (S/CO) for the analyte, and specimens with S/CO values of ≥0.5 were considered positive.
HC2.
HC2 (Qiagen, Mississauga, Ontario, Canada) detects DNA of 13 high-risk oncogenic types, HPV-16, -18, -31, -33, -35, -39, -45, -51, -52, -56, -58, -59, and -68, collectively through hybridization of a RNA probe cocktail with the target DNA, capture of the resultant RNA-DNA hybrids by antibodies specific for the hybrids, and detection of the antibodies with a chemiluminescence substrate. This test was performed with 4 ml of PreservCyt samples using conversion buffer, according to the manufacturer's instructions. Specimens with relative light unit/cutoff ratio (RLU/CO) values of ≥1.0 were considered positive.
Proofer.
Proofer (Norchip, Klokkarstua, Norway) is a real-time multiplex nucleic acid sequence-based amplification assay (NASBA) for isothermal amplification and detection of E6/E7 mRNA from five high-risk oncogenic types, HPV-16, -18, -31, -33, and -45, using molecular beacon probes. Five milliliters of PreservCyt sample was processed for the extraction of HPV RNA using a Magnapure instrument (Roche Molecular Diagnostics), and HPV E6/E7 mRNA was detected according to the manufacturer's instructions and as previously described (17, 23). An FLx900I fluorescence reader (Bio-Tek, Winooski, VT) was used for the detection of the accumulated mRNA product with PreTect analysis software (Norchip) for analyzing the fluorescence profiles. To verify the integrity of RNA in the specimen, the test includes a primer set and a probe directed against the human U1 small nuclear ribonucleoprotein-specific mRNA. Standardized artificial oligonucleotides corresponding to the respective viral sequences are provided in the test kit, and these were included as positive controls for each of the five HPV types targeted by the test. Water was used as the negative control.
LA genotyping test.
The LA assay (Roche Molecular Diagnostics, Laval, Canada) is a qualitative nucleic acid amplification test that detects the L1 DNA sequence of 36 high- and low-risk mucosal HPV types (HPV-6, -11, -16, -18, -26, -31, -33, -34, -35, -39, -40, -42, -44, -45, -51, -52, -53, -54, -56, -58, -59, -61, -62, -66, -67, -68, -69, -70, -71, -72, -73, -81, -82, -83, -84, and -89), with genotyping of specific HPV types (9). The LA assay uses PCR for the amplification of the target sequence, with the β-globin DNA sequence coamplified to control for sample adequacy and DNA quality. The biotinylated amplicons generated are denatured and hybridized to an array containing immobilized probes for 36 HPV genotypes. HPV DNA was extracted from 250 μl of PreservCyt sample using an AmpliLute liquid medium extraction kit (Roche Molecular Diagnostics), according to the manufacturer's instructions. Extracted DNA was then tested using the LA assay as previously described (9). Because the HPV-52-specific probe cross-reacts with types 33, 35, and 58, samples positive with the HPV-52-specific probe were further tested with a validated real-time PCR assay specific for type 52 (8). Only the samples reactive in the HPV-52 real-time PCR assay were considered HPV-52 positive.
Colposcopy and histology.
Participating obstetrician/gynecologist specialists at the study sites carried out colposcopy and, if warranted, cervical biopsy on the day of patient enrollment according to the standard of care. In some cases, biopsy specimens were taken in subsequent follow-up visits, and in such instances, histology results for biopsy specimens taken no later than 6 months following enrollment were included in the study analysis. Cervical histology was read and reviewed independently by pathologists at the respective study sites, and the results served as the disease endpoint for study purposes. The pathologists were blinded to HPV results.
Data analysis.
The clinical performance of the HPV tests was assessed on the basis of histological diagnosis, with CIN 2+ serving as the disease endpoint. The cross-sectional HPV data based on a single cervical brush specimen collected at the time of enrollment and histology results for cervical biopsy specimens obtained either at enrollment or during follow-up for up to 6 months were utilized in this evaluation. Sensitivity, specificity, and predictive values were calculated using the conventional contingency tables, and 95% confidence intervals (CIs) were computed using the binomial method. McNemar's chi-square test was used to test the differences between sensitivities and specificities, as they were calculated using the same set of samples. Accuracy of HPV detection was calculated as the percentage of the correct results by the respective HPV test compared to the histology results. The HPV results were also studied analytically with LA genotyping results. A significance level of 0.05 was used in comparing performance characteristics.
RESULTS
Comparison of Aptima and HC2 for detecting CIN 2+ in referral population.
The clinical performance of Aptima in comparison with that of HC2 was assessed with cervical specimens collected from 1,418 women referred for colposcopy. Of these, 401 (28.3%) women had CIN 2+, which included 13 cases of invasive cervical cancer. There were 366 women with CIN 1 and 651 having either a normal colposcopy result with no biopsy or negative histology, for a total of 1,017 women with a diagnosis of ≤CIN 1 (Table 1). Overall, Aptima was positive for 964 (68%) of the total of 1,418 women and 386 of the 401 women with CIN 2+, for a sensitivity of 96.3%. HC2 was positive for 1,001/1,418 women (70.6%) and 378/401 women with CIN 2+, for a sensitivity of 94.3%. A further analysis of the tests' sensitivities for CIN 2 and ≥CIN 3 is shown in Table 1. The sensitivity values were not significantly different between the two tests. In terms of specificity, Aptima was negative for 439 of the 1,017 women with ≤CIN 1, for a specificity of 43.2%, and HC2 was negative for 394/1,017 women with ≤CIN 1, for a specificity of 38.7% (P < 0.05). Specificities based on CIN 1 and negative histology also showed significantly higher values for Aptima (Table 1). Table 2 provides the details of test results for the 13 cases of invasive cervical cancer. The overall agreement between Aptima and HC2 was 89.6% (1,271/1,418; kappa, 0.75). Considering only the 401 women with CIN 2+, the agreement between Aptima and HC2 was 95% (381/401); there were 372 positive and 9 negative results for both tests, and 20 samples yielded discordant results. Of the nine CIN 2+ cases testing negative by both tests, LA genotyping identified the oncogenic types targeted by both tests in four and nononcogenic types in two, and LA genotyping was negative for three. Among the 20 discordant results, there were 14 Aptima positive (Aptima+) and HC2-negative (HC2−) specimens, 12 of which contained the genotypes targeted by both tests and 2 were not tested by LA genotyping. The remaining six specimens with discordant results were Aptima negative (Aptima−) and HC2 positive (HC2+), and five of these contained the genotypes targeted by both tests and one contained untargeted nononcogenic types.
Table 1.
Comparison of Aptima and HC2 for detection of cervical intraepithelial lesions in referral population by histological gradea
| Diagnostic index | Histological category | Aptima |
HC2 |
||
|---|---|---|---|---|---|
| No. of relevant results/total no. tested (%) | 95% confidence interval | No. of relevant results/total no. tested (%) | 95% confidence interval | ||
| Sensitivity | CIN 2+ (n = 401) | 386/401 (96.3) | 94.4, 98.2 | 378/401 (94.3) | 92.0, 96.6 |
| CIN 2 (n = 120) | 112/120 (93.3) | 88.8, 97.8 | 106/120 (88.3) | 82.6, 94.0 | |
| ≥CIN 3 (n = 281) | 274/281 (97.5) | 95.7, 99.3 | 272/281 (96.8) | 94.7, 98.9 | |
| Specificity | ≤CIN 1 (n = 1,017) | 439/1,017 (43.2) | 40.2, 46.2 | 394/1,017 (38.7)b | 35.7, 41.7 |
| CIN 1 (n = 366) | 93/366 (25.4) | 20.9, 29.9 | 81/366 (22.1)b | 17.8, 26.4 | |
| Negative (n = 651) | 346/651 (53.1) | 49.3, 56.9 | 313/651 (48.1)b | 44.3, 51.9 | |
| Positive predictive value | CIN 2+ (n = 401) | 386/964 (40.0) | 36.9, 43.1 | 378/1,001 (37.8) | 34.8, 40.8 |
| Negative predictive value | ≤CIN 1 (n = 1,017) | 439/454 (96.7) | 95.1, 98.3 | 394/417 (94.5) | 92.3, 96.7 |
n = 1,418.
Specificities for detection of CIN 2+ between Aptima and HC2 significant at P < 0.05.
Table 2.
Comparison of LBC, Aptima, HC2, Proofer, and LA genotyping results for invasive cervical cancer
| Case no. | Age (yr) | LBC findinga | Aptima resultb | Aptima S/CO | HC2 resultb | HC2 RLU/CO | Proofer resultc | Proofer genotype | LA genotype(s) | Histologya |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 77 | UNS | Positive | 13.55 | Positive | 31.58 | Positive | 33 | 33 | SCC |
| 2 | 46 | ASCUS | Positive | 14.22 | Negative | <1 | Positive | 45 | 16, 45 | SCC |
| 3 | 75 | UNS | Positive | 10.29 | Positive | 27.35 | NTd | 31 | SCC | |
| 4 | 39 | ASCUS | Positive | 10.85 | Positive | 771.23 | NT | 16 | SCC | |
| 5 | 43 | ASCUS | Positive | 11.21 | Positive | 13.96 | NT | 16, 61 | SCC | |
| 6 | 24 | HSIL | Positive | 12.27 | Positive | 292.38 | Positive | 16 | 16, 39, 53 | SCC |
| 7 | 28 | LSIL | Positive | 11.56 | Positive | 718.77 | Positive | 16 | 16 | SCC |
| 8 | 42 | HSIL | Positive | 11.49 | Positive | 3.32 | Positive | 16 | 16 | SCC |
| 9 | 30 | HSIL | Positive | 16.98 | Positive | 3.13 | Positive | 45 | 45 | SCC |
| 10 | 45 | ASCUS | Positive | 18.78 | Positive | 138.93 | NT | 16 | SCC | |
| 11 | 51 | ASCUS | Positive | 13.62 | Positive | 42.88 | NT | 52, 58 | SCC | |
| 12 | 34 | LSIL | Positive | 11.94 | Positive | 1.22 | Positive | 18 | 18, 62 | SCC |
| 13 | 32 | ASC-H | Positive | 17.52 | Positive | 541.52 | Positive | 18 | 18 | Adenocarcinoma |
UNS, unsatisfactory; ASCUS, atypical squamous cells of undetermined significance; ASC-H, atypical squamous cells of undetermined significance, favor high grade; LSIL, low-grade squamous intraepithelial lesion; HSIL, high-grade squamous intraepithelial lesion; SCC, squamous cell carcinoma.
On the basis of a total of 1,418 referral specimens tested.
On the basis of a subset of 818 referral specimens tested.
NT, not tested due to lack of sufficient specimen.
Aptima and HC2 results were further analyzed for agreement with LA genotyping results for the total of 1,418 specimens. LA genotyping was performed on 940 of the 964 Aptima+ specimens, of which 936 tested positive for HPV DNA. Of these, 918 (98.1%) contained at least 1 of the 14 types targeted by Aptima. In the remaining 18, the following types, alone or in combination, were detected, in descending order of frequency: 70, 54, 42, 53, 55, 82, 61, 62, 67, 6, 83, and 89. In comparison, LA genotyping was performed on 975 of the 1,001 HC2+ specimens, of which 969 tested positive for HPV DNA. Of these, 907 (93.6%) contained at least 1 of the 13 types targeted by HC2. In the remaining 62 (6.4%), the following types, alone or in combination, were detected, in descending order of frequency: 53, 66, 62, 70, 42, 54, 6, 55, 61, 89, 73, 67, 82, 40, 83, 84, 72, and 81. Since HPV-66 is a type included in Aptima but not in HC2, its prevalence was specifically analyzed. Type 66 was detected in 21 CIN 2+ cases but always in conjunction with multiple-type infections involving other oncogenic HPV types. All 21 CIN 2+ cases tested positive with both Aptima and HC2. Type 66 was also detected in 81 cases of ≤CIN 1; 68 were multiple-type infections and 13 were single-type infections. Aptima and HC2 were positive for 56 and 62 of the 68 cases of ≤CIN 1 with multiple-type infections, respectively, and 9 and 10 of the 13 cases of ≤CIN 1 with single-type infections, respectively.
The Aptima and HC2 results were also analyzed taking into account cytology results, as shown in Table 3. The proportion of cases of CIN 2+ at histology increased with increasing severity of cytological abnormalities, i.e., 28.9% for atypical squamous cells of undetermined significance (ASCUS) category, 29.9% for low-grade squamous intraepithelial lesions (LSILs), and 75.0% for high-grade squamous intraepithelial lesions (HSILs). The comparison of the sensitivities of the tests for detection of CIN 2+ in each of the cytological categories showed no significant differences. However, the specificity of Aptima was significantly higher than that of HC2 for detection of CIN 2+ in the ASCUS and LSIL categories (Table 3).
Table 3.
Correlation of Aptima and HC2 results with cytological and histological diagnosesa
| Cytology | Histology | No. of specimens |
|||
|---|---|---|---|---|---|
| Aptima |
HC2 |
||||
| Positive (% sensitivity) | Negative (% specificity) | Positive (% sensitivity) | Negative (% specificity) | ||
| Negative (n = 467) | ≤CIN 1 (n = 398) | 196 | 202 (50.8) | 205 | 193 (48.5) |
| ≥CIN 2 (n = 69) | 62 (89.9) | 7 | 58 (84.1) | 11 | |
| ASCUSb (n = 419) | ≤CIN 1 (n = 298) | 161 | 137 (46.0) | 178 | 120 (40.3)c |
| ≥CIN 2 (n = 121) | 116 (95.9) | 5 | 114 (94.2) | 7 | |
| LSIL (n = 324) | ≤CIN 1 (n = 227) | 165 | 62 (27.3) | 181 | 46 (20.3)c |
| ≥CIN 2 (n = 97) | 95 (97.9) | 2 | 93 (95.9) | 4 | |
| HSIL (n = 124) | ≤CIN 1 (n = 31) | 27 | 4 (12.9) | 27 | 4 (12.9) |
| ≥CIN 2 (n = 93) | 92 (98.9) | 1 | 92 (98.9) | 1 | |
n = 1,334. Eighty-four cases with unsatisfactory Pap smears were excluded from the total of 1,418.
Includes 22 cases of ASCUS, favor high grade (ASC-H), and two cases of atypical glandular cells.
Specificities for detection of CIN 2+ between Aptima and HC2 significant at P < 0.05
Comparison of Aptima and Proofer for detecting CIN 2+ in referral population.
Due to the lack of sufficient volume of residual PreservCyt samples, the performance of Aptima could be compared with that of Proofer on only 818 of the referral case specimens. Of these, histology identified 235 (28.7%) cases of CIN 2+, which included 8 cases of invasive cervical cancer. There were 204 women with CIN 1 and 379 with either a normal colposcopy finding with no biopsy or negative histology, for a total of 583 with a diagnosis of ≤CIN 1. Aptima showed significantly higher sensitivity and significantly lower specificity than Proofer across histological grades (Table 4). However, Proofer detected all eight cases of invasive cervical cancer, and this included the case which tested negative by HC2 (Table 2). The overall agreement between Aptima and Proofer was 71.1% (582/818). The agreement between the two tests for the detection of CIN 2+ was 81.7% (192/235); there were 183 positive results and 9 negative results for both tests, and 43 specimens yielded discordant results. Of the nine CIN 2+ cases testing negative by both tests, LA genotyping detected at least one of the genotypes targeted by Aptima in four and nononcogenic types in three, and two were negative. However, none of these specimens contained the five types targeted by Proofer. The analysis of the 43 discordant results is shown in Table 5.
Table 4.
Comparison of Aptima and Proofer tests for detection of cervical intraepithelial lesion in referral population by histological gradea
| Diagnostic index | Histological category | Aptima |
Proofer |
||
|---|---|---|---|---|---|
| No. of relevant results/no. tested (%) | 95% confidence interval | No. of relevant results/no. tested (%) | 95% confidence interval | ||
| Sensitivity | CIN 2+ (n = 235) | 223/235 (94.9) | 92.1, 97.7 | 186/235 (79.1)b | 73.9, 84.3 |
| CIN 2 (n = 63) | 58/63 (92.1) | 85.4, 98.8 | 40/63 (63.5)b | 51.6, 75.4 | |
| ≥CIN 3 (n = 172) | 165/172 (95.9) | 92.9, 98.9 | 146/172 (84.9)b | 79.5, 90.3 | |
| Specificity | ≤CIN 1 (n = 583) | 267/583 (45.8) | 41.8, 49.8 | 438/583 (75.1)c | 71.6, 78.6 |
| CIN 1 (n = 204) | 54/204 (26.5) | 20.4, 32.6 | 135/204 (66.2)c | 59.7, 72.7 | |
| Negative (n = 379) | 213/379 (56.2) | 51.2, 61.2 | 303/379 (79.9)c | 75.9, 83.9 | |
| Positive predictive value | CIN 2+ (n = 235) | 223/539 (41.4) | 37.2, 45.6 | 186/331 (56.2) | 50.9, 61.5 |
| Negative predictive value | ≤CIN 1 (n = 583) | 267/279 (95.7) | 93.3, 98.1 | 438/487 (89.9) | 87.2, 92.6 |
n = 818.
Sensitivities for detection of CIN 2+ between Aptima and Proofer significant at P < 0.05.
Specificities for detection of CIN 2+ between Aptima and Proofer significant at P < 0.05.
Table 5.
Analysis of 43 Aptima-Proofer discordant results in CIN 2+ cases
| Result | na | No. of specimens with the following LA genotyping result: |
|||||
|---|---|---|---|---|---|---|---|
| Positive for Aptima |
Positive for Proofer |
Negative for Proofer targeted types | Not testedb | ||||
| Targeted types | Untargeted types | Targeted types | Untargeted types | ||||
| Aptima+/Proofer− | 40 | 19 | 18 | 17 | 2 | 18 | 3 |
| Aptima−/Proofer+ | 3 | 3 | 3 | ||||
n, total number of specimens tested.
Not tested due to lack of sufficient specimen.
Aptima and Proofer results were further analyzed for agreement with LA genotyping results on the total of 881 specimens. Of the 331 Poofer-positive (Proofer+) specimens, LA genotyping identified at least one of the five types targeted by Proofer in 320 (96.7%). Considering type-specific results, there was a 96.1% (318/331) agreement between Proofer and LA genotyping. Of the 235 CIN 2+ cases, there were 200 (85.1%) containing at least one of the five types targeted by Proofer, and 197 (98.5%) and 182 (91.0%) of these were positive by Aptima and Proofer, respectively. Of the 18 Proofer-negative CIN 2+ cases in this subset, Aptima was positive for 17. Among the 35 CIN 2+ cases that did not have the five types targeted by Proofer, LA genotyping detected other oncogenic types in 24 (77.4%) of the 31 tested, with the most common types being 52, 51, 58, and 39. The Aptima and Proofer results were also analyzed taking into account the cytology results, as shown in Table 6. The comparison of sensitivities and specificities of these tests for the detection of CIN 2+ in each of the cytological grades showed significant differences, with Aptima showing a higher sensitivity and a lower specificity than Proofer (Table 6).
Table 6.
Correlation of Aptima and Proofer results with cytological and histological diagnosesa
| Cytology | Histology | No. of specimens |
|||
|---|---|---|---|---|---|
| Aptima |
Proofer |
||||
| Positive (% sensitivity) | Negative (% specificity) | Positive (% sensitivity) | Negative (% specificity) | ||
| Negative (n = 222) | ≤CIN 1 (n = 193) | 76 | 117 (60.6) | 36 | 157 (81.3)b |
| ≥CIN 2 (n = 29) | 25 (86.2) | 4 | 21 (72.4)d | 8 | |
| ASCUSc (n = 226) | ≤CIN 1 (n = 167) | 90 | 77 (46.1) | 42 | 125 (74.9)b |
| ≥CIN 2 (n = 59) | 54 (91.5) | 5 | 44 (74.6)d | 15 | |
| LSIL (n = 226) | ≤CIN 1 (n = 161) | 112 | 49 (30.4) | 48 | 113 (70.2)b |
| ≥CIN 2 (n = 65) | 63 (96.9) | 2 | 50 (76.9)d | 15 | |
| HSIL (n = 91) | ≤CIN 1 (n = 23) | 20 | 3 (13.0) | 11 | 12 (52.2)b |
| ≥CIN 2 (n = 68) | 67 (98.5) | 1 | 58 (85.3)d | 10 | |
n = 765. Fifty-three cases with unsatisfactory Pap smears were excluded from the total of 818.
Specificities for detection of CIN 2+ between Aptima and Proofer significant at P < 0.05.
Includes 19 cases of ASCUS, favor high grade (ASC-H), and 1 case of atypical glandular cells.
Sensitivities for detection of CIN 2+ between Aptima and Proofer significant at P < 0.05.
Comparison of Aptima and HC2 for detecting CIN 2+ in routine screened population.
In the 1,373 women participating in routine cervical cancer screening, 7 were found to have CIN 2+, including a case of adenocarcinoma. Both Aptima and HC2 were positive in all seven cases of CIN 2+. The remaining 1,366 with ≤CIN 1 included 30 with CIN 1 and 1,336 with either normal colposcopy findings with no biopsy or negative histology. The specificities on the basis of the results obtained with the 1,366 women with ≤CIN 1 were 88.4% (95% CI, 86.6, 90.0) for Aptima and 85.2% (95% CI, 83.3, 87.1) for HC2 (P < 0.05). As this study population comprised women 16 to 81 years of age and HPV-based screening is indicated for women ≥30 years, specificity was reassessed on the basis of the results for women in this age group (n = 845). Aptima showed a specificity of 93.9% (95% CI, 92.3, 95.5), whereas HC2 showed a specificity of 92.1% (95% CI, 90.3, 93.9) (P < 0.05). The corresponding figures for women <30 years of age (n = 521) were 79.5% (95% CI, 75.8, 82.8) and 74.1% (95% CI, 70.3, 77.9), respectively (P < 0.05).
DISCUSSION
The main goal of this study was to compare the clinical performance of the Aptima mRNA assay with that of HC2 for the detection of cervical precancerous lesions and cancer in both colposcopy referral and routinely screened populations. The Aptima assay was also compared with Proofer in the referral population. We chose referral patients because this population selects out women with the highest prevalence of cervical oncogenic HPV infection and the associated cervical precancerous lesions and cancer. This allowed us to demonstrate the relative sensitivity of Aptima with a large number of CIN 2+ cases within a limited sample size of the study population and to also assess its performance for the triage of borderline cytological abnormalities. On the other hand, the majority of women undergoing routine cervical cytology screening do not have HPV infection and thus have negligible prevalent cervical lesions, and this served as an ideal population with which to define the relative specificity of the Aptima test. With a sample size of 1,418 referral women with 401 cases of CIN 2+ (CIN ≥3, n = 281), this study is the largest cross-sectional study to date to assess the performance of Aptima in comparison with that of HC2 and the second study to compare Aptima with Proofer in this population. In addition, our study is one of the few studies to assess the performance of Aptima in a routinely screened population.
The Aptima-HC2 comparative study with the referral population showed a slightly higher sensitivity for Aptima than HC2, i.e., 96.3% versus 94.3%, for the detection of CIN 2+. Recently, published studies with similar referral populations showed a sensitivity of 95.5% for Aptima versus 99.6% for HC2 in the United Kingdom (40) and sensitivities of 90.8% and 95.0%, respectively, in France (12). A study in Germany has shown similar results with cervical specimens tested from cytologically abnormal cases stored for up to 3 years (5a). While our observed sensitivity rate in a Canadian population was generally in agreement with those from the studies described above, it should be noted that the manufacturer has recently revised the test's cutoff from ≥1.0, which was used in the above studies, to the cutoff of ≥0.5 that we used in our study. The slightly increased sensitivity of Aptima over that of HC2 that we observed may be, to some extent, attributable to this lower test cutoff. It is also worth noting that Aptima was positive for all 13 cases of cervical cancer, compared to 12 detected by HC2. Our study showed that the specificity of Aptima was significantly higher than that of HC2, i.e., 43.2% and 38.7%, respectively, consistent with the findings of other studies (12, 40). The higher specificity of Aptima could be attributable to the targeting of mRNA as well as the absence of cross-reactivity. Our data indicated that the lower specificity of HC2 is due in part to its cross-reactivity with untargeted nononcogenic types. The HC2 cross-reactivity rate of 6.4% that we observed in the present study is similar to the rate of 6.5% noted in a previous study (29) and the rate of 7.9% reported in a larger study (5). Regardless, the overall agreement of Aptima with HC2 was good, reaching 95.0% for the detection of CIN 2+.
The Aptima-Proofer comparative study utilizing specimens from a subset of the referral population showed a significantly higher sensitivity for Aptima than Proofer, i.e., 94.9% and 79.1%, respectively, and a lower specificity for Aptima than Proofer, i.e., 45.8% and 75.1%, respectively. In a study similar to ours conducted in the United Kingdom, Proofer showed a sensitivity of 73.6% and a specificity of 73.1% for detecting CIN 2+ (40). The higher sensitivity and lower specificity of Aptima compared with those of Proofer are attributable to the wider spectrum of HPV types targeted by Aptima and the higher prevalence of oncogenic types other than the five targeted by Proofer in ≤CIN 1 lesions. Conversely, the lower sensitivity and higher specificity of Proofer are mainly due to the limited number of HPV genotypes targeted by Proofer, as previously reported (29). While a higher specificity of an HPV test is desirable, targeting only the most common oncogenic types in cervical cancer may have adverse clinical implications. Regardless, Proofer, targeting the five most common oncogenic types in cervical cancer, still might identify the majority of women whose cervical lesions are likely to progress, and there is an indication that Proofer is a better predictor of CIN 2+ than DNA-based tests (17, 19, 24). Further, as opposed to Aptima, Proofer provides simultaneous type-specific HPV identification, which is useful for risk stratification and better clinical management, especially for infections with HPV types 16 and 18 (14).
Our study of the referral population utilized LA genotyping to assess false-negative and discordant results with the HPV tests. It should be noted that while LA genotyping is a standardized commercially available test, the failure of this test to detect some or all of the targeted HPV types in a sample has been reported (41) and cannot be ruled out in our study. Our data indicated a better overall agreement of both Aptima and Proofer results with LA genotyping results (98.1% and 96.7%, respectively) than HC2 results with LA genotyping results (93.6%). The lower agreement of HC2 is attributable to its cross-reactivity with untargeted nononcogenic types, as noted above. In contrast, Aptima+/Proofer− results were attributable to oncogenic types not targeted by Proofer. In a small proportion of Aptima- and Proofer-positive specimens, LA genotyping identified nononcogenic types exclusively. It is not known whether this is due to the difference in the tests' targets or the limitation of LA genotyping to detect coexisting oncogenic types. The genotypic analysis highlights the relative analytical specificities of the three tests, with Aptima showing a better specificity. Our data also provided confirmation that the five types targeted by Proofer accounted for 85.1% of CIN 2+ cases in the North American setting, and this is similar to the 82.1% fraction we previously reported (29) and consistent with the recognized prevalence rates (2, 6, 36). This implies that the remaining CIN 2+ fraction is attributable to other oncogenic types. LA genotyping confirmed that a majority of Proofer-negative CIN 2+ cases, in fact, contained oncogenic HPV types not targeted by that assay; most of these tested positive by Aptima. If we restrict Proofer sensitivity analysis to the five types targeted by the test, the sensitivity of Proofer would improve significantly, reaching 91% compared with a sensitivity of 98.5% for Aptima (data not shown). The above difference in sensitivity may be attributable to a higher threshold of the Proofer cutoff for the detection of E6/E7 mRNA, as previously reported (29). Further, we specifically analyzed our genotypic data to determine the relevance of including type 66 in the Aptima test. The data indicated that type 66 was detected only in association with other oncogenic types in CIN 2+ cases and as a single type only in ≤CIN 1 cases, the majority of which tested positive by Aptima and also by HC2 due to its known cross-reactivity. Considering these results and the high probability of regression of CIN 1 lesions, our data appear to suggest that the inclusion of type 66 in the Aptima test is more likely to reduce specificity than increase sensitivity. While this is a speculation based on a limited number of cases and requires verification, it is worth noting that although type 66 is considered an oncogenic type and indicated for inclusion in HPV tests (7, 21, 39), there is only limited evidence for its oncogenic potential in cervical cancer (3).
Our analysis of the relative performance of Aptima stratified by cytological grades in the referral population provides data on the potential application of this test for the triage of borderline cytological abnormalities. Due to its higher specificity, Aptima would reduce colposcopy referrals by nearly 6.0% compared to HC2 for women with ASCUS cytology while retaining the same level of sensitivity as HC2 for the detection of CIN 2+. This reduction would remain about the same if these tests are used for women with LSIL cytology as well (Table 3). A similar analysis of Aptima with Proofer indicated that the Aptima referral rate would be significantly higher than the Proofer referral rate due to the lower specificity of Aptima compared with that of Proofer. But the higher specificity of Proofer and the attendant greater reduction in referral rate is achieved at the cost of lowered sensitivity (Table 4). This is to be taken into account when balancing sensitivity and specificity and could be influenced by specific needs and settings.
We chose to include women <30 years of age in our routine screened population to further define the potential for increased specificity of an mRNA-based test as this age group is known to have a higher prevalence of transient HPV infection. Aptima showed a significantly higher specificity than HC2 in this age group as well as those ≥30 years of age, while it maintained the same level of sensitivity as HC2 for detecting CIN 2+. This has significance in population-based adjunct or primary HPV screening, considering that even a small reduction in test specificity can result in a large number of women being unnecessarily targeted for intensified clinical management. A recent French study based on 4,429 routinely screened women 20 to 65 years of age has reported sensitivities of 95.7% and 95.3% for Aptima and HC2, respectively, for the detection of CIN 3 or worse, with Aptima showing a significantly higher specificity, i.e., 90.31% versus 84.9% (25). A study in China based on 2,097 routinely screened women 25 to 29 years of age has shown similar results (44a).
The recently published guidelines for validating new HPV assays for primary cervical screening advise that this be conducted with a representative population-based screening cohort of women ≥30 years of age and that the sensitivity of a candidate test be at least 90% of the sensitivity of HC2 and achieve a statistical power of over 99% when 100 cases of CIN 2+ are tested (21). Although there were only a very few CIN 2+ cases in our routine screened population due to the limited sample size, there were 401 cases of CIN 2+ in the referral population, with Aptima showing sensitivity equal to that of HC2 for detecting CIN 2+. Considering that ≥CIN 3 is a better marker of risk of progression to invasive cervical cancer than ≤CIN 2 (21, 35), an equivalent sensitivity of Aptima was also observed on the basis of 281 cases of ≥CIN 3 (Table 1). Regarding specificity, the guidelines advise that a candidate test should have specificity at least 98% of the specificity of HC2 and that this be demonstrated with a minimum random sample of 800 women ≥30 years of age to achieve a power of 80%. In our sample size of 845 routinely screened women in this age group, Aptima showed a significantly higher specificity than HC2. It should also be noted that the performance indices of HC2 in our studies of both referral and routine screened populations were consistent with those previously reported (10, 21, 40). In conclusion, our data indicate that the Aptima test is as sensitive as HC2 but more specific for detecting CIN 2+ and has the potential to serve as a reliable test for both primary cervical cancer screening and the triage of borderline cytological abnormalities.
ACKNOWLEDGMENTS
This study was supported by a research grant from Merck Frosst Canada Ltd., and we thank James Mansi for his support and enthusiasm.
Clinical collaborators (all in Canada) included Justice Arthur (Burin Peninsula Health Care Centre, Burin), Carol Greene (Medical West Clinic, St. John's), Andrea Singleton (Churchill Square Medical Clinic, St. John's), Lesa Dawson, Catherine Popadiuk, and Patti Power (Dr. H. Bliss Murphy Cancer Centre, St. John's), Thomas Baskett, Catherine Craig, Isabelle Delisle, Jeffery Dempster, Robert Grimshaw, Katharina Kieser, Winifred Lee, Barbara Parish, and Khalid Sait (Queen Elizabeth II Health Sciences Centre, Halifax), and Cheryl Algers, Nisrin Anfinan, Pam Chu, Jennifer Hilton, Jalene Mannerfeldt, Jill Nation, and Gregg Nelson (Tom Baker Cancer Centre, Calgary).
Research staff (all in Canada) included Bettina Bentley (Queen Elizabeth II Health Sciences Centre, Halifax), Carol Blady and Bonnie Kozark (Regional Health Sciences Centre, Thunder Bay), Erin Breit and Danielle Arseneault (Tom Baker Cancer Centre, Calgary), Pierre Forest (Centre Hospitalier de l'Université de Montréal, Montreal), Elizabeth Oates, Adam Byrne, Samantha Ratnam, Laura Gilbert, Claire Press, and Debbie McGrath, (Public Health Laboratory, St. John's), and Patsy Francis and Mary Paul (Regional Cytology Laboratory, St. John's).
We thank Laura Gilbert for assistance with the preparation of the manuscript.
Footnotes
Published ahead of print on 8 December 2010.
REFERENCES
- 1. Anttila A., et al. 2010. Rate of cervical cancer, severe intraepithelial neoplasia, and adenocarcinoma in situ in primary HPV DNA screening with cytology triage: randomised study within organized screening programme. BMJ 340:c1804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Bosch F. X., et al. 2008. Epidemiology and natural history of human papillomavirus infections and type-specific implications in cervical neoplasia. Vaccine 26S:K1–K16 [DOI] [PubMed] [Google Scholar]
- 3. Bouvard V., et al. 2009. A review of human carcinogens. Part B. Biological agents. Lancet Oncol. 10:321–322 [DOI] [PubMed] [Google Scholar]
- 4. Bulkmans N. W. J., et al. 2007. Human papillomavirus DNA testing for the detection of cervical intraepithelial neoplasia grade 3 and cancer: 5-year follow-up of a randomised controlled implementation trial. Lancet 370:1764–1772 [DOI] [PubMed] [Google Scholar]
- 5. Castle P. E., et al. 2008. Human papillomavirus genotype specificity of Hybrid Capture 2. J. Clin. Microbiol. 46:2595–2604 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5a. Clad A., et al. 2010. EUROGIN 2010, abstr. P HT-1 [Google Scholar]
- 6. Clifford G. M., Smith J. S., Aguado T., Franceschi S. 2003. Comparison of HPV type distribution in high-grade cervical lesions and cervical cancer: a meta-analysis. Br. J. Cancer 89:101–105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Cogliano V., et al. 2005. Carcinogenicity of human papillomaviruses. Lancet Oncol. 6:204. [DOI] [PubMed] [Google Scholar]
- 8. Coutlée F., et al. 2007. Confirmatory real-time PCR assay for human papillomavirus (HPV) type 52 infection in anogenital specimens screened for HPV infection with the Linear Array HPV genotyping test. J. Clin. Microbiol. 45:3821–3823 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Coutlée F., et al. 2006. Enhanced detection and typing of human papillomavirus DNA in anogenital samples with PGMY primers and the linear array HPV genotyping test. J. Clin. Microbiol. 44:1998–2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cuzick J., et al. 2008. Overview of human papillomavirus-based and other novel options for cervical cancer screening in developed and developing countries. Vaccine 26S:K29–K41 [DOI] [PubMed] [Google Scholar]
- 11. Cuzick J., et al. 2006. Overview of the European and North American studies on HPV testing in primary cervical cancer screening. Int. J. Cancer 199:1095–1101 [DOI] [PubMed] [Google Scholar]
- 12. Dockter J., et al. 2009. Clinical performance of the APTIMA® HPV assay for the detection of high-risk HPV and high-grade cervical lesions. J. Clin. Virol. 45:S55–S61 [DOI] [PubMed] [Google Scholar]
- 13. Franco E. L., Cuzick J., Hildesheim A., De Sanjosé S. 2006. Issues in planning in cervical cancer screening in the era of HPV vaccination. Vaccine 24S3:S3/171–S3/177 [DOI] [PubMed] [Google Scholar]
- 14. Khan M. J., et al. 2005. 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. 97:1072–1079 [DOI] [PubMed] [Google Scholar]
- 15. Koliopoulos G., et al. 2007. Diagnostic accuracy of human papillomavirus testing in primary cervical screening: a systematic review and meta-analysis of non-randomized studies. Gynecol. Oncol. 104:232–246 [DOI] [PubMed] [Google Scholar]
- 16. Kraus I., et al. 2004. Human papillomavirus oncogenic expression in the dysplastic portio; an investigation of biopsies from 190 cervical cones. Br. J. Cancer 90:1407–1413 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Kraus I., et al. 2006. Presence of E6 and E7 mRNA from human papillomavirus types 16, 18, 31, 33, and 45 in the majority of cervical carcinomas. J. Clin. Microbiol. 44:1310–1317 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Lie A. K., Kristensen G. 2008. Human papillomavirus E6/E7 mRNA testing as a predictive marker for cervical carcinoma. Expert Rev. Mol. Diagn. 8:405–415 [DOI] [PubMed] [Google Scholar]
- 19. Lie A. K., et al. 2005. DNA- versus RNA-based methods for human papillomavirus detection in cervical neoplasia. Gynecol. Oncol. 97:908–915 [DOI] [PubMed] [Google Scholar]
- 20. Mayrand M. H., et al. 2007. Human papillomavirus DNA versus Papanicolaou screening tests for cervical cancer. N. Engl. J. Med. 357:1579–1588 [DOI] [PubMed] [Google Scholar]
- 21. Meijer C. J. L. M., et al. 2009. Guidelines for human papillomavirus DNA test requirements for primary cervical cancer screening in women 30 years and older. Int. J. Cancer 124:516–520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Molden T., et al. 2005. Comparison of human papillomavirus messenger RNA and DNA detection: a cross sectional study of 4,136 women >30 years of age with a 2-year follow-up of high-grade squamous intraepithelial lesion. Cancer Epidemiol. Biomarkers Prev. 14:367–372 [DOI] [PubMed] [Google Scholar]
- 23. Molden T., Kraus I., Skomedal H., Nordstrøm T., Karlsen F. 2007. PreTect™ HPV-Proofer: real-time detection and typing of E6/E7 mRNA from carcinogenic human papillomaviruses. J. Virol. Methods 142:204–212 [DOI] [PubMed] [Google Scholar]
- 24. Molden T., et al. 2005. Predicting CIN2+ when detecting HPV mRNA and DNA by PreTect HPV-Proofer and consensus PCR: a 2-year follow-up of women with ASCUS or LSIL Pap smear. Int. J. Cancer 114:973–976 [DOI] [PubMed] [Google Scholar]
- 25. Monsonego J., et al. 12 October 2010, posting date Evaluation of oncogenic human papillomavirus RNA and DNA tests with liquid based cytology in primary cervical cancer screening (the FASE study). Int. J. Cancer. doi:10.1002/ijc.25726 [DOI] [PubMed] [Google Scholar]
- 26. Münger K., et al. 2004. Mechanisms of human papillomavirus-induced oncogenesis. J. Virol. 78:11451–11460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Naucler P., et al. 2007. Human papillomavirus and Papanicolaou tests to screen for cervical cancer. N. Engl. J. Med. 357:1589–1597 [DOI] [PubMed] [Google Scholar]
- 28. Poljak M., Marin I. J., Seme K., Vince A. 2002. Hybrid Capture II HPV test detects at least 15 human papillomavirus genotypes not included in its current high-risk probe cocktail. J. Clin. Virol. 25:S89–S97 [DOI] [PubMed] [Google Scholar]
- 29. Ratnam S., et al. 2010. Clinical performance of the Pre-Tect HPV-Proofer E6/E7 mRNA assay in comparison with that of the Hybrid Capture 2 test for identification of women at risk of cervical cancer. J. Clin. Microbiol. 48:2779–2785 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29a. Ratnam S., et al. 2010. Clinical correlation of Aptima HPV E6/E7 mRNA test in cervical cancer screening: preliminary results from a multicentre Canadian study. EUROGIN, abstr. SS 8–6 [Google Scholar]
- 30. Ronco G., et al. 2008. 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. 100:492–501 [DOI] [PubMed] [Google Scholar]
- 31. Safaeian M., et al. 2007. Comparison of the SPF10-LiPA system to the Hybrid Capture 2 assay for detection of carcinogenic human papillomavirus genotypes among 5,683 young women in Guanacaste, Costa Rica. J. Clin. Microbiol. 45:1447–1454 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Sankaranarayanan R., et al. 2009. HPV screening for cervical cancer in rural India. N. Engl. J. Med. 360:1385–1394 [DOI] [PubMed] [Google Scholar]
- 33. Saslow D., et al. 2002. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J. Clin. 52:342–362 [DOI] [PubMed] [Google Scholar]
- 34. Schiffman M. H., et al. 1993. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J. Natl. Cancer Inst. 85:958–964 [DOI] [PubMed] [Google Scholar]
- 35. Schiffman M., Castle P. E., Jeronimo J., Rodriquez A. C., Wacholder S. 2007. Human papillomavirus and cervical cancer. Lancet 370:890–907 [DOI] [PubMed] [Google Scholar]
- 36. Smith J. S., et al. 2007. Human papillomavirus type distribution in invasive cervical cancer and high-grade cervical lesions: a meta-analysis update. Int. J. Cancer 121:621–632 [DOI] [PubMed] [Google Scholar]
- 37. Solomon D., et al. 2002. The 2001 Bethesda system: terminology for reporting results of cervical cytology. JAMA 287:2114–2119 [DOI] [PubMed] [Google Scholar]
- 38. Sotlar K., et al. 2004. Detection of high-risk human papillomavirus E6 and E7 oncogene transcripts in cervical scrapes by nested RT-polymerase chain reaction. J. Med. Virol. 74:107–116 [DOI] [PubMed] [Google Scholar]
- 39. Stoler M. H., Castle P. E., Solomon D., Schiffman M. 2007. The expanded use of HPV testing in gynecologic practice per ASCCP-guided management requires the use of well-validated assays. Am. J. Clin. Pathol. 127:335–337 [DOI] [PubMed] [Google Scholar]
- 40. Szarewski A., et al. 2008. Comparison of predictors for high-grade cervical intraepithelial neoplasia in women with abnormal smears. Cancer Epidemiol. Biomarkers Prev. 17:3033–3042 [DOI] [PubMed] [Google Scholar]
- 41. van Hamont D., van Ham M. A. P. C., Bakkers J. M. J. E., Massuger L. F. A. G., Melchers W. J. G. 2006. Evaluation of the SPF10-INNO LiPA human papillomavirus (HPV) genotyping test and the Roche Linear Array HPV genotyping test. J. Clin. Microbiol. 44:3122–3129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Walboomers J. M. M., et al. 1999. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 189:12–19 [DOI] [PubMed] [Google Scholar]
- 43. Wright T. C., et al. 2007. 2006 consensus guidelines for the management of women with abnormal cervical cancer screening tests. Am. J. Obstet. Gynecol. 197:346–355 [DOI] [PubMed] [Google Scholar]
- 44. Wright T. C., et al. 2007. 2006 consensus guidelines for the management of women with cervical intraepithelial neoplasia or adenocercinoma in situ. Am. J. Obstet. Gynecol. 197:340–345 [DOI] [PubMed] [Google Scholar]
- 44a. Wulan N., et al. 2010. EUROGIN 2010, abstr. SS 23/24-13 [Google Scholar]
- 45. zur Hausen H. 1994. Molecular pathogenesis of cancer of the cervix and its causation by specific human papillomavirus types. Curr. Top. Microbiol. Immunol. 186:131–156 [DOI] [PubMed] [Google Scholar]