- Email: [email protected]
Prevalence of Human Papillomavirus Infection in Unselected SurePath Samples Using the APTIMA HPV mRNA Assay
The Journal of Molecular Diagnostics, Vol. 15, No. 5, September 2013
jmd.amjpathol.org
Prevalence of Human Papillomavirus Infection in Unselected SurePath Samples Using the APTIMA HPV mRNA Assay Matejka Rebolj,* Sarah Preisler,yz Ditte M. Ejegod,yz Jesper Bonde,yz Carsten Rygaard,y and Elsebeth Lynge* From the Department of Public Health,* University of Copenhagen, Copenhagen; and the Department of Pathologyy and the Clinical Research Center,z Hvidovre University Hospital, Hvidovre, Denmark Accepted for publication April 10, 2013. Address correspondence to Matejka Rebolj, Ph.D., Department of Public Health, University of Copenhagen, Øster Farimagsgade 5, DK 1014 Copenhagen K, Denmark. E-mail: [email protected].
The APTIMA Human Papillomavirus (HPV) Assay detects E6/E7 mRNA from 14 human papillomavirus genotypes. Horizon was a population-based split-sample study among well-screened women, with an aim to compare APTIMA, Hybrid Capture 2 (HC2), and liquid-based cytology (LBC) using SurePath samples. APTIMA testing on the PANTHER platform, and HC2 testing on the Rapid Capture System were performed in accordance with protocols agreed on with the manufacturers before the study, on 5070 consecutive, routine, cervical cytology samples from Copenhagen, Denmark. In this high-risk population, 17% of all samples tested positive on APTIMA, 20% of samples tested positive on HC2, and 7% of samples had abnormal cytology. Among the 4411 samples without recent abnormalities, 15% tested positive on APTIMA, 19% tested positive on HC2, and 5% had abnormal cytology. The k coefficient of 0.75 suggested substantial agreement between APTIMA and HC2. This is the first APTIMA study using SurePath samples on the PANTHER platform. The trends in positivity rates on SurePath samples for APTIMA, HC2, and LBC were consistent with studies based on PreservCyt samples, and the agreement between the two HPV assays was substantial. The high proportions of women testing positive suggest that in countries with a high HPV prevalence, caution will be needed if HPV tests, including mRNA-based tests, are to replace LBC. (J Mol Diagn 2013, 15: 670e677; http://dx.doi.org/10.1016/ j.jmoldx.2013.04.002)
Screening with human papillomavirus (HPV) DNA tests offers excellent protection from cervical cancer,1,2 but many women with transient HPV infections test positive.3 Integration of HPV into the host cell genome leads to expression of E6 and E7 oncogenes, causing disruption in p53 and pRb cell-cycle control pathways.4 Therefore, detection of E6/E7
mRNA (mRNA) from HPV may be more specific for the detection of high-grade cervical intraepithelial neoplasia (CIN) than HPV DNA testing.5 The Food and Drug Administrationeapproved APTIMA HPV Assay (Hologic/Gen-Probe, San Diego, CA) is an in vitro nucleic acid amplification test for qualitative
Supported by the Danish Strategic Research Council (10-092793 to M.R. and S.P.). None of the funders had any role in the study design; in the collection, analysis, or interpretation of the data; or in the writing of the report or the decision to submit the article for publication. Disclosures: Hologic/Gen-Probe, the manufacturer of the APTIMA HPV Assay; Qiagen, the manufacturer of the Hybrid Capture 2 assay; and Genomica, the manufacturer of the CLART HPV2 assay, provided assays, instrumentation, and limited co-funding for laboratory material for the study. None of the authors was compensated for their work on this project, hold stock in, or received bonuses from any of the manufacturers. According to the contract between the manufacturers, Hvidovre University
Hospital, and the University of Copenhagen, Hologic/Gen-Probe, Qiagen, and Genomica had the right to comment on a draft version of this manuscript but had no editorial rights. The authors alone retained the right to submit the manuscript. All authors have attended meetings with manufacturers of HPV assays. S.P. received honoraria from Hologic/Gen-Probe for lectures; D.M.E. received honoraria from Genomica for lectures; J.B. has served as a paid advisor to Roche and received honoraria from Hologic/Gen-Probe, Roche, Qiagen, and Genomica for lectures; C.R. has served as an unpaid advisor to Roche; and E.L. has served as an unpaid advisor to Hologic/Gen-Probe and Norchip. Hvidovre Hospital holds a recompense agreement with Genomica on a KRAS/BRAF diagnostic system.
Copyright ª 2013 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2013.04.002
APTIMA mRNA HPV on SurePath Samples detection of E6/E7 viral mRNA from 14 HPV genotypes (genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68; Hologic/Gen-Probe, San Diego, CA) detected in combination. Of these, 13 are known to be carcinogenic in humans, and 1 (genotype 66) is possibly carcinogenic in humans.7 The assay has an internal control spiked in each reaction, and these controls monitor each step of the process. APTIMA’s sensitivity for high-grade CIN appears to be similar to that of HPV DNA tests, and much better than that of cytology, which historically has been the standard cervical screening test. At the same time, studies indicate that APTIMA has better clinical specificity than HPV DNA tests.7e16 APTIMA has been validated for use only on samples stored in PreservCyt (Hologic, Boxborough, MA). Arbyn et al16 have called for evaluation of APTIMA using other storage media. SurePath (BD DiagnosticseTriPath, Burlington, NC) is the other widely used medium; however, the formaldehydebased fixative in SurePath represents a special challenge for molecular-based testing because it leads to extensive crosslinking of cellular proteins and nucleic acids, rendering the genomic material less easily accessible for extraction and subsequent analysis.17,18 To some extent, this cross-linking can be reversed, and the amount of amplifiable RNA increased, by proteinase K treatment,18,19 which is part of the pre-analytical treatment for APTIMA (Schröder AI, Guzenski LI, Dockter J: Poster handout. Eurogin, February 17e20, 2010, Monte Carlo). Previous studies performed the testing using either the TIGRIS platform8,10,20 or a semiautomated Direct Tube Sampling system.9,10,12 A new, smaller, fully automated and Food and Drug Administrationeapproved platform PANTHER now also is available (all Hologic/Gen-Probe). An aim of Horizon, a population-based split-sample study, was to compare APTIMA with the Hybrid Capture 2 High-Risk HPV assay (HC2; Qiagen, Hilden, Germany) and liquid-based cytology (LBC) on unselected samples from Copenhagen, Denmark. Danish women are well screened but have a high prevalence of HPV infection and a high background risk of cervical cancer.21,22 In Denmark, SurePath has an 80% market share, and gradual implementation of HPVbased screening therefore has to be based on SurePath samples. Unlike previous studies evaluating the APTIMA assay, the Horizon study was performed on samples stored in SurePath, and testing was performed on the PANTHER platform.
Materials and Methods Setting The Department of Pathology at Hvidovre University Hospital in Copenhagen is accredited with Joint Commission International, and is the largest cervical cytology laboratory in Denmark. While the present study was ongoing, the Department of Pathology at Hvidovre University Hospital in Copenhagen evaluated approximately 66,000 cervical
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
SurePath samples per year. It handles all cervical cytology from women living in the municipalities of Copenhagen and Frederiksberg (denoted as Copenhagen in the remainder of the article), regardless of the reason for sample collection. Since the 1960s, Copenhagen has been covered by an organized cervical screening program. At present, women aged 23 to 49 years are targeted for screening every 3 years, and women aged 50 to 65 years are targeted for screening every 5 years. In 2011, 76% of women in the Capital Region, which includes Copenhagen, were screened at least once within the recommended interval.23
Sample Collection Horizon was nested into routine laboratory practice. Samples were identified by the internal laboratory specimen identifier and a barcode before they were divided sequentially into racks of 48 samples. For this study, samples were collected from June 10, 2011, to August 25, 2011, equally from Monday to Friday. Based on capacity and processing considerations at the molecular pathology laboratory, the target number of samples was set to 5000. After completion of routine LBC evaluation [which includes HC2 triage of women aged 30 years and older with atypical squamous cells of undetermined significance (ASCUS)], approximately 2 mL of residual cellular suspension of SurePath per sample was collected for Horizon from each of the tubes in, at most, the first four racks processed on the sampling days. This method mimicked collection of unselected consecutive LBC samples, assuming that the time of day that the sample arrived in the laboratory was not associated with any specific characteristics. In accordance with the protocol agreed on with the assay manufacturers before the study, collected samples were diluted with 2 mL of SurePath (dilution factor, approximately 1:1) to obtain enough volume for additional testing, which will be reported separately (unpublished data). All testing was performed in the same laboratory and by the same staff.
Cytology Cytologic evaluation of SurePath samples was performed by cytoscreeners, in collaboration with a pathologist in cases of abnormal findings. Reading of the specimens was assisted by the FocalPoint GS Imaging System (SlideWizard; BD, Burlington, NC). Cytologic findings were reported using the Bethesda 2001 system, and abnormalities were classified as ASCUS, low-grade squamous intraepithelial lesions, or highgrade squamous intraepithelial lesions or worse. The last category includes atypical squamous cells, which cannot exclude high-grade squamous intraepithelial lesions, atypical glandular cells, adenocarcinoma in situ, and carcinoma. Samples with ASCUS or worse (ASCUS) were considered to have abnormal cytology. Cytoscreeners and pathologists were blinded to the results of HPV testing in Horizon.
671
Rebolj et al
APTIMA HPV mRNA Testing For handling of SurePath samples, we followed a protocol provided for this study by Hologic/Gen-Probe. Diluted sample (1 mL) was aliquoted into an APTIMA Specimen Transfer Tube containing 2.9 mL of buffered solution (Hologic/Gen-Probe). Before APTIMA testing, proteinase K treatment was performed using the Pace 2 Fast Expression Kit containing 1 mL diluent and lyophilized reagent (all from Hologic/Gen-Probe). Reconstituted proteinase K (100 mL) was added to each Specimen Transfer Tube and incubated at 65 C for 2 hours. The treated specimen tube was stored at 2 C to 8 C until testing. Testing was performed on the PANTHER platform. After lysis of the cells, mRNA was captured using oligonucleotide probes attached to magnetic beads. The released target mRNA molecules were amplified using transcription-mediated amplification, in which two different primers, nucleotides, the T7 RNA polymerase and the Moloney murine leukemia virus reverse transcriptase with RNase H activity produced mRNA amplicons in an isothermal reaction. The amplicons were hybridized to DNA probes labeled with an acridinium ester detector molecule, leading to photon emission. The light signal (relative light units) was detected on a luminometer and a signal to cut-off ratio (S/CO) was calculated from the positive and negative calibrators loaded onto the machine. An internal control was included in each sample as supplied in the kit.6 Positive APTIMA samples were defined in accordance with the manufacturer’s specifications as an S/CO of 0.5 or greater. We also evaluated APTIMA’s intralaboratory reproducibility in a routine setting, based on a selection of 361 negative and 375 positive APTIMA samples from different Horizon batches. These samples were retested once within an average of 8 days (range, 1 to 50 days). For samples with nonreproducible APTIMA results (disagreement between initial and retest outcomes), viral DNA was extracted from separate 1-mL aliquots using a LC Total Nucleic Assay kit on the automated LC96 MagnaPure platform (Roche Diagnostics, Rotkreuz, Switzerland) and subsequently analyzed using the CLART HPV2 genotyping assay (Genomica, Madrid, Spain) according to the manufacturer’s specifications. CLART targets L1 genomic sequences of the DNA from 35 HPV genotypes.
Hybrid Capture 2 HPV DNA Testing HC2 is widely considered the standard HPV DNA assay.24 It is a hybridization assay detecting 13 high-risk genotypes (genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) in combination. The assay has no internal control for sufficiency of test material. Sample DNA was either denatured before testing by pretreating manually according to the manufacturer’s protocol (n Z 5043), or DNA was isolated and purified using the DSP AXpH DNA kit on the QIASymphony SP platform (Qiagen, Hilden, Germany) (n Z
672
1086). Testing of these samples was performed on the automated Rapid Capture System (Qiagen, Gaithersburg, MD) using different scripts depending on pretreatment. A minority of samples used for routine triage of women aged 30 years and older with ASCUS were denatured and tested manually (n Z 106). The remaining 23 samples were not tested with HC2. Positive HC2 samples were defined as those having a relative light unit/cut-off value of 1 or greater.
Screening History The screening history of women from January 1, 2000, onward was retrieved from the Danish Pathology Data Bank.25 Horizon samples with an earlier diagnosis of cervical cancer, a CIN diagnosis up to 3 years earlier, ASCUS in the previous 15 months, or more severe cytologic abnormalities or a positive HPV test in the past 12 months were considered follow-up samples. Horizon samples with no recent abnormality were considered primary samples. Reflecting routine practice, primary samples included screening samples and a small proportion of samples taken by indication.
Statistical Analysis Differences in the distributions of age, screening history, cytology, and HC2 results between the included and excluded samples were tested with the c2 test. Trends in HPV positivity by age were tested with the Mantel-Haenszel c2 test for trend. Agreement between the APTIMA and HC2 assays was measured by the k coefficient. The 95% CIs for the differences in the relative proportions (RPs) of women testing positive on APTIMA or HC2 by cytology grade were calculated by assuming that the logarithms of RPs were approximately normally distributed.
Ethical Considerations Horizon was designed as a quality development study, using only residual material that otherwise would have been discarded. According to Danish regulations of biomedical research, published on May 5, 2011, in the Guidelines about Notification of a Biomedical Research Project to the Committee System on Biomedical Research Ethics no. 9154 section 2.5, quality development studies do not require ethical approval.
Results Study Sample Of the 12,138 samples processed by the laboratory during the sample collection period, 6258 (52%) were collected and processed for the Horizon study. All collected samples were tested with LBC, and all but 23 samples (0.4%, for technical reasons) were tested with HC2. In total, 1188 (19.0%) of the 6258 samples were not tested with APTIMA owing to insufficient volume for the complete study (n Z 1165) or because of technical reasons
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples (n Z 23), and were excluded from the analysis. There was no significant difference in the distribution of age (P Z 0.08), cytology (P Z 0.46), or HC2 result (P Z 0.31) between the 5070 included and the 1188 excluded samples. There was a small but significant difference in women’s screening history, with 87.0% of the included and 89.9% of the excluded samples being primary (P Z 0.01). Among the 5070 samples, 5011 (98.8%) women had one sample collected during the course of the study, whereas 59 (1.2%) samples were from 29 women, reflecting consecutive sampling from routine practice. The mean age of the women was 37.3 years (SD, 12.3 years; range, 16 to 89 years), and 64% of the samples were obtained from women younger than age 40 years (Table 1), reflecting a predominantly young source population (Statistics Denmark, http://www.statbank.dk/folk1; last accessed, April 11, 2013). The mean number of days between the receipt of the specimen in the laboratory and storage in the APTIMA Specimen Transport Tubes was 2 (range, 1 to 5 days). The mean number of days between storing and proteinase K treatment was 81 (range, 45 to 112 days); and the mean number of days between proteinase K treatment and testing with APTIMA was 12 (range, 0 to 25 days). All of these times were in accordance with the protocol agreed on with the manufacturer before the study.
seven of these, whereas for the remaining four samples there were errors in the testing process (data analysis error, failure in sample dispense verification, and/or incorrect sample volume). One sample (<0.1%) remained invalid after the testing had been repeated. APTIMA was positive in 849 (16.7%) samples. For women targeted by the screening program (age, 23 to 65 years), the proportion testing positive was 16.1%. This proportion decreased from 27.8% at ages 23 to 29 years to 3.8% at ages 60 to 65 years (P < 0.0001 for trend across age categories) (Table 1). In women ages 16 to 22 and older than age 65, the proportions testing positive on APTIMA were 44.4% and 6.2%, respectively. Women in this age range are not invited routinely for screening, but may have presented for medical conditions or may have sought screening themselves. In total, 38 (0.7%) samples tested positive on APTIMA at S/CO values of 0.50 or greater and less than 1.00. From the 5064 samples that were tested with HC2, 1035 (20.4%) tested positive. Although the proportion testing positive was higher overall than with APTIMA, it showed a similar decrease with increasing age. Among women ages 16 to 22 years, 56.8% of samples tested positive on HC2, 33.0% of samples tested positive at ages 23 to 29 years, 17.4% of samples tested positive at ages 30 to 39 years, 11.3% of samples tested positive at ages 40 to 49 years, 7.3% of samples tested positive at ages 50 to 59 years, 6.0% of samples tested positive at ages 60 to 65 years, and 6.2% of samples tested positive at age older than 65 years (data not shown). Of the 846 APTIMA-positive samples for which HC2 results were
HPV Testing Overall, APTIMA testing had to be repeated on 11 (0.2%) samples to obtain valid results. The internal control failed in Table 1
APTIMA Results for 5070 Primary and Follow-Up Samples by Age, Screening History, Cytology, and HC2 Results Results on APTIMA, N (%)
Characteristics
Positive
Negative
Invalid
Total
Total Age, years 16e22 23e29 30e39 40e49 50e59 60e65 >65 Screening history Primary sample Follow-up sample Cytology Normal ASCUS LSIL HSIL ASCUS Inadequate HC2* Positive Negative
849 (16.7%)
4220 (83.2%)
1 (0.0%)
5070 (100%)
72 427 225 82 27 9 7
90 1108 1301 911 479 225 106
0 0 0 0 1 0 0
162 1535 1526 993 507 234 113
(44.4%) (27.8%) (14.7%) (8.3%) (5.3%) (3.8%) (6.2%)
(55.6%) (72.2%) (85.3%) (91.7%) (94.5%) (96.2%) (93.8%)
(0.0%) (0.0%) (0.0%) (0.0%) (0.2%) (0.0%) (0.0%)
(100%) (100%) (100%) (100%) (100%) (100%) (100%)
676 (15.3%) 173 (26.3%)
3734 (84.7%) 486 (73.7%)
1 (0.0%) 0 (0.0%)
4411 (100%) 659 (100%)
587 58 110 93 261 1
4083 65 33 14 112 25
1 0 0 0 0 0
4671 123 143 107 373 26
(12.6%) (47.2%) (76.9%) (86.9%) (70.0%) (3.8%)
747 (72.2%) 99 (2.5%)
(87.4%) (52.8%) (23.1%) (13.1%) (30.0%) (96.2%)
288 (27.8%) 3929 (97.5%)
(0.0%) (0.0%) (0.0%) (0.0%) (0.0%) (0.0%)
0 (0.0%) 1 (0.0%)
(100%) (100%) (100%) (100%) (100%) (100%)
1035 (100%) 4029 (100%)
*Six samples were not tested with HC2. HSIL, high-grade intraepithelial lesions or worse; LSIL, low-grade squamous intraepithelial lesions.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
673
Rebolj et al available, 747 (88.3%) were also HC2 positive, and of the 4217 APTIMA-negative samples with HC2 results, 3929 (93.2%) were also HC2 negative. The k coefficient of 0.75 indicated substantial agreement in these samples between the APTIMA and HC2 assays. Cytology was abnormal in 373 (7.4%) of 5070 samples. The proportion of samples testing positive on APTIMA and HC2 increased with cytologic grade. Among samples with normal cytology, 12.6% tested positive on APTIMA and 15.6% tested positive on HC2 (RP, 0.81; 95% CI, 0.73 to 0.89); among samples showing ASCUS 47.2% and 64.2% were positive, respectively (RP, 0.73; 95% CI, 0.58 to 0.92); among samples showing low-grade squamous intraepithelial lesions, 76.9% and 89.5% were positive, respectively (RP, 0.86; 95% CI, 0.77 to 0.96); and among samples with highgrade squamous intraepithelial lesions or worse, 86.9% and 90.7% were positive, respectively (RP, 0.96; 95% CI, 0.87 to 1.05) (for both assays, P < 0.0001 for trend). These results were fairly similar in the subset of 4411 primary samples (Table 2). In the primary samples, 676 (15.3%) tested positive on APTIMA, 822 (18.6%) tested positive on HC2 (k coefficient for agreement between APTIMA and HC2, 0.75), and 242 (5.5%) samples had ASCUS. Among women ages 23 to 29 years, 26.8% tested positive on APTIMA, 31.7% tested positive on HC2, and 7.4% tested positive on LBC (Table 3). Among women ages 30 to 65 years, these proportions were 9.4%, 11.7%, and 4.4%, respectively. Because women from Copenhagen are on average younger than Danish women in general, the proportion of women with positive test results would be somewhat lower if testing were performed for all women in Denmark. Table 2
APTIMA and LBC were both positive in 166 (3.8%) primary samples. The proportions of APTIMA-positive, cytology-normal primary samples strongly decreased with age (Figure 1): from 20.5% at ages 23 to 29 years, 9.7% at ages 30 to 39 years, 5.9% at ages 40 to 49 years, to 3.1% at ages 50 to 65 years (P < 0.0001 for trend).
Intralaboratory Reproducibility of APTIMA Among the 375 APTIMA-positive samples in the initial run, 318 (84.8%) had reproducible results on retesting (positive reproducibility). The average initial S/CO value of the 57 nonreproducible samples was 4.99 (range, 0.56 to 11.55), compared with 11.17 (range, 0.63 to 40.89) for reproducible samples. By using the CLART genotyping assay, 37 (64.9%) of the 57 samples contained at least 1 of the 14 HPV genotypes targeted by APTIMA, 5 (8.8%) samples contained at least 1 of the other 21 genotypes detectable by CLART but not included among the 14 APTIMA-targeted genotypes, whereas none of the 35 CLART HPV genotypes was detected in 15 (26.3%) of the 57 samples. Among the 361 negative samples in the initial run, 355 (98.3%) had reproducible results (negative reproducibility). The average S/CO value of the six nonreproducible samples was 5.01 (range, 0.58 to 10.59) in the second run.
Discussion General Findings This is the first study to report APTIMA results using SurePath samples tested on the PANTHER platform. Whether this combination affects APTIMA’s sensitivity and/or specificity
APTIMA Results for 4411 Primary Samples by Age, Screening History, Cytology, and HC2 Results Results on APTIMA, N (%)
Characteristics
Positive
Negative
Invalid
Total
Total Age, years 16e22 23e29 30e39 40e49 50e59 60e65 >65 Cytology Normal ASCUS LSIL HSIL ASCUS Inadequate HC2* Positive Negative
676 (15.3%)
3734 (84.7%)
1 (0.0%)
4411 (100%)
53 345 172 71 23 6 6
(39.3%) (26.8%) (13.2%) (7.9%) (4.9%) (2.8%) (5.7%)
82 942 1128 833 441 209 99
(60.7%) (73.2%) (86.8%) (92.1%) (94.8%) (97.2%) (94.3%)
0 0 0 0 1 0 0
(0.0%) (0.0%) (0.0%) (0.0%) (0.2%) (0.0%) (0.0%)
135 1287 1300 904 465 215 105
(100%) (100%) (100%) (100%) (100%) (100%) (100%)
510 40 65 61 166 0
(12.3%) (47.1%) (74.7%) (87.1%) (68.6%) (0.0%)
3637 45 22 9 76 21
(87.7%) (52.9%) (25.3%) (12.9%) (31.4%) (100.0%)
1 0 0 0 0 0
(0.0%) (0.0%) (0.0%) (0.0%) (0.0%) (0.0%)
4148 85 87 70 242 21
(100%) (100%) (100%) (100%) (100%) (100%)
592 (72.0%) 82 (2.3%)
230 (28.0%) 3502 (97.7%)
0 (0.0%) 1 (0.0%)
822 (100%) 3585 (100%)
*Four samples were not tested with HC2. HSIL, high-grade intraepithelial lesions or worse; LSIL, low-grade squamous intraepithelial lesions.
674
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples Table 3
APTIMA in Primary Screening: Comparison of the Horizon Data With Previously Published Studies
Age group 30 years Horizon Horizon Monsonego et al11 Ratnam et al12 Wu et al14 <30 years Horizon Monsonego et al11 Ratnam et al12
Area
N
Age group, years
Copenhagen* Denmark* Parisy Canadax Shenzhenk
2884 3320 845 2000
30e65 30e65 30e65 30e81 25e59**
Copenhagen* Parisy Canadax
1287 1109 521
23e29 20e29 16e29
Positive on APTIMA
Positive on HC2
Positive on LBC
9.4% 7.7% 8.5% >6.1%{ 10.1%
11.7% 9.7% 13.0% >7.9%{ 16.5%
4.4% 4.3% 8.3%z NR 5.4%
26.8% 15.6% >20.5%{
31.7% 23.5% >25.9%{
7.4% 13.5%z NR
*Inclusion criteria were as follows: routine primary samples, see Materials and Methods. Copenhagen: the results refer to samples included in Horizon. Denmark: the results refer to proportions from Horizon, age-standardized to all Danish women. y FASE trial. Inclusion criteria were as follows: women seen for their annual examination in 17 private gynecology practices, not having undergone total hysterectomy, nonpregnant, no cytologic abnormality in the past 6 months. Women with unsatisfactory cytology results (10.5%) were excluded from analyses. z Adjudicated cytology results. x Inclusion criteria were as follows: women undergoing routine Papanicolaou screening at two screening centers. { Lower bound, approximated as 100% specificity for CIN1. jj Inclusion criteria were as follows: women eligible for the Shenzhen Cervical Cancer Screening Trial I (nonpregnant, no cervical cancer screening for at least 3 years, no prior hysterectomy, no prior pelvic irradiation). Of the 2098 enrolled women, 3 did not have complete data for APTIMA, HC2, or LBC, and 95 women with at least one positive test result did not return for colposcopy. These 98 women were excluded from the study. Women were medically underserved and predominantly urban. **Reported as 24 to 59 years in tables from the original publication. Data were not reported separately by age group. NR, not reported.
16−22 Yrs N=135 Cytology − Cytology + 23−29 Yrs N=1287
16−89 Yrs N=4411
30−39 Yrs N=1300 40−49 Yrs N=904 50−59 Yrs 60−65 Yrs N=465 N=215
>65 Yrs N=105
30−65 Yrs N=2884
-
jmd.amjpathol.org
10% 20%
Although 19% of the collected samples had to be discarded, there was no significant difference between the included and the excluded samples in terms of women’s age, cytology, or the HC2 result. All samples were tested in one laboratory by the same staff that is involved in routine screening. In a 2004 to
% Cytology Positive
0
Strengths and Weaknesses
The Journal of Molecular Diagnostics
20%
% HPV−Positive
30%
40%
2005 study, a total of 11,617 (primary) consecutive SurePath samples from the same area evaluated in the same laboratory were tested with HC2.22 The median age of the women in that study was 36.4 years (range, 15 to 93 years), and 6% had abnormal cytology. The proportion of women aged 25 to 64 years testing positive on HC2 was approximately 17%, which is similar to the 16% in Horizon. Horizon samples therefore can be considered representative for the source population. Following routine recommendations from the Danish cervical screening program, women with abnormal cytology were referred for colposcopy or for repeat testing. As a follow-up to Horizon, we will invite women with positive HPV tests and normal cytology for repeat testing in about a year. The standard measures of sensitivity and specificity of the screening tests will be reported once this histologic follow-up evaluation of the study population has been completed.
10%
for detecting high-grade CIN has not yet been described in the literature. In Horizon, the overall prevalence of HPV infection measured by APTIMA was 17%, compared with 20% on HC2, whereas 7% of all samples had abnormal cytology. The prevalence of HPV infection on APTIMA decreased with age and increased with cytology grade. These findings all conform to expectations based on prior studies using PreservCyt samples. The results suggest that APTIMA performed well on SurePath samples, but this remains to be confirmed with histologic outcomes collected in a planned follow-up study to Horizon. The positive reproducibility observed for APTIMA was only 85%, which might signal a need to further optimize the protocol for storing and handling SurePath samples with respect to molecular target stability. It also is noteworthy that the nonreproducible positive samples had on average a lower S/CO value to start. A lack of positive reproducibility may pose problems for laboratories, for example, in terms of their ability to verify the initial results on issues of patient safety, and to show reliable quality assurance and control procedures. However, a low positive reproducibility to some degree might be balanced by a low frequency of positive tests. Furthermore, in routine settings, action is taken only on a single test, therefore low positive reproducibility would not constitute a problem for an assay’s sensitivity for CIN.
HPV + HPV −
Figure 1 Proportion of women with a positive primary APTIMA test and abnormal primary cytology by age.
675
Rebolj et al
Comparison With Other Studies and Implications for Screening A majority of studies published to date on APTIMA testing was performed in women referred for colposcopy because of prior abnormal cytology (ASCUS). In these disease-enriched populations, the proportions of women whose PreservCyt samples tested positive on APTIMA were typically on the order of 50% to 80%.10,12,13,15,20,26 This was comparable with our study, in which 69% of primary samples with ASCUS tested positive (Table 2). In women with ASCUS aged 30 years and older from Jutland, Denmark, 31.7% tested positive on APTIMA.8 Similarly, in a Norwegian study using 6- to 12month follow-up samples of women with ASCUS or lowgrade squamous intraepithelial lesions, of whom 80% were aged older than 30 years, 32.0% tested positive on APTIMA.9 Among the 86 women with ASCUS aged 30 years and older in Horizon, 39.5% tested positive on APTIMA. The findings of the Horizon study using SurePath samples thus correspond well with the findings of previous studies based on PreservCyt samples. Although HPV DNA screening has been strongly discouraged for women younger than age 30 years because of its lack of specificity for high-grade CIN,1,27 conclusions from the French APTIMA Screening Evaluation study (FASE) supported the use of APTIMA in women younger than age 30 years.11 In this age group, FASE researchers estimated APTIMA’s specificity for CIN3 at 84.9% (95% CI, 82.8 to 87.0), almost the same as that of LBC (86.9%; 95% CI, 84.9 to 88.9), and better than that of HC2 (76.9%; 95% CI, 74.4 to 79.4). In Horizon, the proportion of women aged 23 to 29 years testing positive on LBC was about half of that in FASE, whereas the proportion testing positive on APTIMA was about twice as high (Table 3). This large difference between the proportions testing positive on APTIMA and on LBC in Horizon probably will lead to a larger difference in the specificity for the two screening tests than was observed in FASE. In FASE, APTIMA’s good specificity compared with LBC may reflect, among other factors, a relatively high proportion of women with ASCUS. Proportions of women aged 30 years and older testing positive on APTIMA were similar in all four reported screening studies, including Horizon (range, >6% to 10%).11,12,14 The proportion of women with a positive LBC was about 4% to 5% in the Chinese Shenzhen and Horizon studies, but again was higher, at 8%, in FASE. Nonetheless, all available screening trials observed a lower proportion of positive tests by APTIMA than by HC2. The only two studies to date that used a manufacturerrevised cut-off value for positive APTIMA resultsd0.50 or greater S/CO instead of 1.00 or greater S/COdwere a Canadian screening study12 and Horizon. A wider definition of a positive result could help improve the assay’s sensitivity for high-grade CIN, but also could decrease its specificity. In Horizon, the proportion of samples with S/CO values between 0.50 and less than 1.00 was only 0.7% (1.0% in women aged 23 to 29 years, and 0.6% in women aged 30 to 65 years).
676
Based on these baseline data it does not appear that the lowered cut-off value for a positive APTIMA test substantially decreased the assay’s specificity. In light of the high cost and limited capacity for colposcopy in many countries, including Denmark, the high proportion of women with positive HPV DNA tests calls for triage testing. It appears that triage also will be necessary for women with positive mRNA tests. Cytology may be a reasonable candidate triage test. In Horizon, 6.3% of women with a primary test at age 23 to 29 years tested positive on both LBC and APTIMA (Figure 1). Among women aged 30 to 65 years, this was 2.5%. The remaining 20.5% (26.8% 6.3%) among young women, and 6.9% (9.4% 2.5%) among older women, who tested positive on APTIMA and negative on LBC, would wait several months for repeat testing. As stated earlier, histologic verification of positive screening tests in Horizon is underway. Nevertheless, some preliminary estimates of the frequency of false-positive test results, defined as positive tests without histologic CIN3, can be calculated based on recent historical data for Copenhagen. These data, retrieved from the Pathology Data Bank, showed that 1.3% of all screened women had CIN3 detected after abnormal cytology (ASCUS). If HPV testing is 32% more sensitive for CIN3 than cytology,28 about 1.7% of women are expected to have CIN3 detected after a positive HPV test. Therefore, 7.7% (9.4% 1.7%) of women aged 30 to 65 years would have a false-positive APTIMA test result with respect to the clinical outcome. This would be 10.0% (11.7% 1.7%) for HC2, and 3.1% (4.4% 1.3%) for LBC. Among women aged 23 to 29 years, the proportions would be 25.5% (27.2% 1.7%) for APTIMA, 30.5% (32.2% 1.7%) for HC2, and 6.2% (7.5% 1.3%) for LBC. Because of a high expected frequency of false-positive test results in countries with a high prevalence of HPV infections, caution therefore is needed when introducing HPV-based screening, even in the case of highly sensitive26 mRNA assays. Optimization of the LBC media with molecular HPV testing in mind, or vice versa, will be a critical issue for future implementation of primary HPV mRNA or DNA testing with cytology triage, and studies using SurePath are highly relevant given its substantial market share. An alternative could be to introduce a two-tier sampling process, requiring either that two samples are taken per examination (one in molecular collection media, and one in LBC media), or that initially one sample is taken from all women but that those who require triage are recalled for an additional test. However, a two-tier sampling process would be associated with extra costs; the need for a recall also might induce unnecessary anxiety, and reduce the effectiveness of screening as a result of loss to follow-up evaluation.
Conclusions Based on data from Copenhagen, an area with a high prevalence of HPV infections, fewer women tested positive on
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples APTIMA than on HC2. Baseline data furthermore suggest that APTIMA performed well using SurePath samples on the PANTHER platform.
13.
Acknowledgments We thank Miki Hansen and Maria Louise Deistler (Hvidovre University Hospital) for excellent laboratory assistance; Bryan Howard Goldman (University of Copenhagen) for visual representation of the data and help with statistical analysis; the Danish Hologic/Gen-Probe, Qiagen, and Genomica representatives for enabling the collaboration between their parent companies; Hvidovre University Hospital; and the University of Copenhagen.
14.
15.
16.
References 1. Ronco G, Giorgi-Rossi P, Carozzi F, Confortini M, Dalla Palma P, Del Mistro A, Ghiringhello B, Girlando S, Gillio-Tos A, De Marco L, Naldoni C, Pierotti P, Rizzolo R, Schincaglia P, Zorzi M, Zappa M, Segnan N, Cuzick J: Efficacy of human papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial neoplasia: a randomised controlled trial. Lancet Oncol 2010, 11:249e257 2. Rijkaart DC, Berkhof J, Rozendaal L, van Kemenade FJ, Bulkmans NW, Heideman DA, Kenter GG, Cuzick J, Snijders PJ, Meijer CJ: 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, 13:78e88 3. Rebolj M, Pribac I, Lynge E: False-positive human papillomavirus DNA tests in cervical screening: it is all in a definition. Eur J Cancer 2011, 47:255e261 4. Doorbar J: Molecular biology of human papillomavirus infection and cervical cancer. Clin Sci (Lond) 2006, 110:525e541 5. Cuschieri K, Wentzensen N: Human papillomavirus mRNA and p16 detection as biomarkers for the improved diagnosis of cervical neoplasia. Cancer Epidemiol Biomarkers Prev 2008, 17:2536e2545 6. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, IbrahimTallaa L, Guha N, Freeman C, Galichet L, Cogliano V: A review of human carcinogensepart B: biological agents. Lancet Oncol 2009, 10:321e322 7. Szarewski A, Ambroisine L, Cadman L, Austin J, Ho L, Terry G, Liddle S, Dina R, McCarthy J, Buckley H, Bergeron C, Soutter P, Lyons D, Cuzick J: Comparison of predictors for high-grade cervical intraepithelial neoplasia in women with abnormal smears. Cancer Epidemiol Biomarkers Prev 2008, 17:3033e3042 8. Waldstrom M, Ornskov D: Comparison of the clinical performance of an HPV mRNA test and an HPV DNA test in triage of atypical squamous cells of undetermined significance (ASC-US). Cytopathology 2012, 23:389e395 9. Ovestad IT, Vennestrom U, Andersen L, Gudlaugsson E, Munk AC, Malpica A, Feng W, Voorhorst F, Janssen EA, Baak JP: Comparison of different commercial methods for HPV detection in follow-up cytology after ASCUS/LSIL, prediction of CIN2-3 in follow up biopsies and spontaneous regression of CIN2-3. Gynecol Oncol 2011, 123:278e283 10. Dockter J, Schroder A, Hill C, Guzenski L, Monsonego J, Giachetti C: Clinical performance of the APTIMA HPV Assay for the detection of high-risk HPV and high-grade cervical lesions. J Clin Virol 2009, 45(Suppl 1):S55eS61 11. Monsonego J, Hudgens MG, Zerat L, Zerat JC, Syrjanen K, Halfon P, Ruiz F, Smith JS: 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, 129:691e701 12. Ratnam S, Coutlee F, Fontaine D, Bentley J, Escott N, Ghatage P, Gadag V, Holloway G, Bartellas E, Kum N, Giede C, Lear A: Aptima
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
17.
18.
19.
20.
21. 22.
23.
24.
25. 26.
27. 28.
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, 49:557e564 Clad A, Reuschenbach M, Weinschenk J, Grote R, Rahmsdorf J, Freudenberg N: Performance of the APTIMA high-risk HPV mRNA assay in a referral population in comparison with Hybrid Capture 2 and cytology. J Clin Microbiol 2011, 49:1071e1076 Wu R, Belinson SE, Du H, Na W, Qu X, Wu R, Liu Y, Wang C, Zhou Y, Zhang L, Belinson JL: Human papillomavirus messenger RNA assay for cervical cancer screening: the Shenzhen Cervical Cancer Screening Trial I. Int J Gynecol Cancer 2010, 20:1411e1414 Reuschenbach M, Clad A, von Knebel DC, Wentzensen N, Rahmsdorf J, Schaffrath F, Griesser H, Freudenberg N, von Knebel DM: Performance of p16INK4a-cytology. HPV mRNA, and HPV DNA testing to identify high grade cervical dysplasia in women with abnormal screening results. Gynecol Oncol 2010, 119:98e105 Arbyn M, Roelens J, Cuschieri K, Cuzick J, Szarewski A, Ratnam S, Reuschenbach M, Belinson S, Belinson JL, Monsonego J: The APTIMA HPV assay versus the hybrid capture 2 test in triage of women with ASC-USor LSIL cervical cytology: a meta-analysisof the diagnostic accuracy. Int J Cancer 2013, 132:101e108 Horvath CA, Boulet G, Sahebali S, Depuydt C, Vermeulen T, van den Broeck D, Vereecken A, Bogers J: Effects of fixation on RNA integrity in a liquid-based cervical cytology setting. J Clin Pathol 2008, 61: 132e137 Powell N, Smith K, Fiander A: Recovery of human papillomavirus nucleic acids from liquid-based cytology media. J Virol Methods 2006, 137:58e62 Murphy PG, Henderson DT, Adams MD, Horlick EA, Dixon EP, King LM, Avissar PL, Brown CA, Fischer TJ, Malinowski DP: Isolation of RNA from cell lines and cervical cytology specimens stored in BD SurePath preservative fluid and downstream detection of housekeeping gene and HPV E6 expression using real time RT-PCR. J Virol Methods 2009, 156:138e144 Waldstrom M, Ornskov D: Clinical performance of a human papillomavirus messenger RNA test (Aptima HPV Assay) on residual material from archived 3-year-old PreservCyt samples with low-grade squamous intraepithelial lesion. Arch Pathol Lab Med 2011, 135:1052e1056 Clemmesen J, Nielsen A: The social distribution of cancer in Copenhagen, 1943 to 1947. Br J Cancer 1951, 5:159e171 Kjaer SK, Breugelmans G, Munk C, Junge J, Watson M, Iftner T: Population-based prevalence, type- and age-specific distribution of HPV in women before introduction of an HPV-vaccination program in Denmark. Int J Cancer 2008, 123:1864e1870 Styregruppen for DKLS: Årsrapport DKLS 2011: Dansk Kvalitetsdatabase for Livmoderhalskræftscreening [annual report 2011: Danish Quality Assurance database for cervical cancer screening]. Hvidovre, DKLS, 2012, pp 1e122 Meijer CJ, Berkhof J, Castle PE, Hesselink AT, Franco EL, Ronco G, Arbyn M, Bosch FX, Cuzick J, Dillner J, Heideman DA, Snijders PJ: Guidelines for human papillomavirus DNA test requirements for primary cervical cancer screening in women 30 years and older. Int J Cancer 2009, 124:516e520 Bjerregaard B, Larsen OB: The Danish Pathology Register. Scand J Public Health 2011, 39:72e74 Szarewski A, Mesher D, Cadman L, Austin J, Ashdown-Barr L, Ho L, Terry G, Liddle S, Young M, Stoler M, McCarthy J, Wright C, Bergeron C, Soutter WP, Lyons D, Cuzick J: Comparison of seven tests for high-grade cervical intraepithelial neoplasia in women with abnormal smears: the predictors 2 study. J Clin Microbiol 2012, 50:1867e1873 Castle PE: Screening: hPV testing for cervical cancer: the good, the bad, and the ugly. Nat Rev Clin Oncol 2010, 7:364e365 Cuzick J, Arbyn M, Sankaranarayanan R, Tsu V, Ronco G, Mayrand MH, Dillner J, Meijer CJ: Overview of human papillomavirus-based and other novel options for cervical cancer screening in developed and developing countries. Vaccine 2008, 26(Suppl 10):K29eK41
677
jmd.amjpathol.org
Prevalence of Human Papillomavirus Infection in Unselected SurePath Samples Using the APTIMA HPV mRNA Assay Matejka Rebolj,* Sarah Preisler,yz Ditte M. Ejegod,yz Jesper Bonde,yz Carsten Rygaard,y and Elsebeth Lynge* From the Department of Public Health,* University of Copenhagen, Copenhagen; and the Department of Pathologyy and the Clinical Research Center,z Hvidovre University Hospital, Hvidovre, Denmark Accepted for publication April 10, 2013. Address correspondence to Matejka Rebolj, Ph.D., Department of Public Health, University of Copenhagen, Øster Farimagsgade 5, DK 1014 Copenhagen K, Denmark. E-mail: [email protected].
The APTIMA Human Papillomavirus (HPV) Assay detects E6/E7 mRNA from 14 human papillomavirus genotypes. Horizon was a population-based split-sample study among well-screened women, with an aim to compare APTIMA, Hybrid Capture 2 (HC2), and liquid-based cytology (LBC) using SurePath samples. APTIMA testing on the PANTHER platform, and HC2 testing on the Rapid Capture System were performed in accordance with protocols agreed on with the manufacturers before the study, on 5070 consecutive, routine, cervical cytology samples from Copenhagen, Denmark. In this high-risk population, 17% of all samples tested positive on APTIMA, 20% of samples tested positive on HC2, and 7% of samples had abnormal cytology. Among the 4411 samples without recent abnormalities, 15% tested positive on APTIMA, 19% tested positive on HC2, and 5% had abnormal cytology. The k coefficient of 0.75 suggested substantial agreement between APTIMA and HC2. This is the first APTIMA study using SurePath samples on the PANTHER platform. The trends in positivity rates on SurePath samples for APTIMA, HC2, and LBC were consistent with studies based on PreservCyt samples, and the agreement between the two HPV assays was substantial. The high proportions of women testing positive suggest that in countries with a high HPV prevalence, caution will be needed if HPV tests, including mRNA-based tests, are to replace LBC. (J Mol Diagn 2013, 15: 670e677; http://dx.doi.org/10.1016/ j.jmoldx.2013.04.002)
Screening with human papillomavirus (HPV) DNA tests offers excellent protection from cervical cancer,1,2 but many women with transient HPV infections test positive.3 Integration of HPV into the host cell genome leads to expression of E6 and E7 oncogenes, causing disruption in p53 and pRb cell-cycle control pathways.4 Therefore, detection of E6/E7
mRNA (mRNA) from HPV may be more specific for the detection of high-grade cervical intraepithelial neoplasia (CIN) than HPV DNA testing.5 The Food and Drug Administrationeapproved APTIMA HPV Assay (Hologic/Gen-Probe, San Diego, CA) is an in vitro nucleic acid amplification test for qualitative
Supported by the Danish Strategic Research Council (10-092793 to M.R. and S.P.). None of the funders had any role in the study design; in the collection, analysis, or interpretation of the data; or in the writing of the report or the decision to submit the article for publication. Disclosures: Hologic/Gen-Probe, the manufacturer of the APTIMA HPV Assay; Qiagen, the manufacturer of the Hybrid Capture 2 assay; and Genomica, the manufacturer of the CLART HPV2 assay, provided assays, instrumentation, and limited co-funding for laboratory material for the study. None of the authors was compensated for their work on this project, hold stock in, or received bonuses from any of the manufacturers. According to the contract between the manufacturers, Hvidovre University
Hospital, and the University of Copenhagen, Hologic/Gen-Probe, Qiagen, and Genomica had the right to comment on a draft version of this manuscript but had no editorial rights. The authors alone retained the right to submit the manuscript. All authors have attended meetings with manufacturers of HPV assays. S.P. received honoraria from Hologic/Gen-Probe for lectures; D.M.E. received honoraria from Genomica for lectures; J.B. has served as a paid advisor to Roche and received honoraria from Hologic/Gen-Probe, Roche, Qiagen, and Genomica for lectures; C.R. has served as an unpaid advisor to Roche; and E.L. has served as an unpaid advisor to Hologic/Gen-Probe and Norchip. Hvidovre Hospital holds a recompense agreement with Genomica on a KRAS/BRAF diagnostic system.
Copyright ª 2013 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2013.04.002
APTIMA mRNA HPV on SurePath Samples detection of E6/E7 viral mRNA from 14 HPV genotypes (genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68; Hologic/Gen-Probe, San Diego, CA) detected in combination. Of these, 13 are known to be carcinogenic in humans, and 1 (genotype 66) is possibly carcinogenic in humans.7 The assay has an internal control spiked in each reaction, and these controls monitor each step of the process. APTIMA’s sensitivity for high-grade CIN appears to be similar to that of HPV DNA tests, and much better than that of cytology, which historically has been the standard cervical screening test. At the same time, studies indicate that APTIMA has better clinical specificity than HPV DNA tests.7e16 APTIMA has been validated for use only on samples stored in PreservCyt (Hologic, Boxborough, MA). Arbyn et al16 have called for evaluation of APTIMA using other storage media. SurePath (BD DiagnosticseTriPath, Burlington, NC) is the other widely used medium; however, the formaldehydebased fixative in SurePath represents a special challenge for molecular-based testing because it leads to extensive crosslinking of cellular proteins and nucleic acids, rendering the genomic material less easily accessible for extraction and subsequent analysis.17,18 To some extent, this cross-linking can be reversed, and the amount of amplifiable RNA increased, by proteinase K treatment,18,19 which is part of the pre-analytical treatment for APTIMA (Schröder AI, Guzenski LI, Dockter J: Poster handout. Eurogin, February 17e20, 2010, Monte Carlo). Previous studies performed the testing using either the TIGRIS platform8,10,20 or a semiautomated Direct Tube Sampling system.9,10,12 A new, smaller, fully automated and Food and Drug Administrationeapproved platform PANTHER now also is available (all Hologic/Gen-Probe). An aim of Horizon, a population-based split-sample study, was to compare APTIMA with the Hybrid Capture 2 High-Risk HPV assay (HC2; Qiagen, Hilden, Germany) and liquid-based cytology (LBC) on unselected samples from Copenhagen, Denmark. Danish women are well screened but have a high prevalence of HPV infection and a high background risk of cervical cancer.21,22 In Denmark, SurePath has an 80% market share, and gradual implementation of HPVbased screening therefore has to be based on SurePath samples. Unlike previous studies evaluating the APTIMA assay, the Horizon study was performed on samples stored in SurePath, and testing was performed on the PANTHER platform.
Materials and Methods Setting The Department of Pathology at Hvidovre University Hospital in Copenhagen is accredited with Joint Commission International, and is the largest cervical cytology laboratory in Denmark. While the present study was ongoing, the Department of Pathology at Hvidovre University Hospital in Copenhagen evaluated approximately 66,000 cervical
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
SurePath samples per year. It handles all cervical cytology from women living in the municipalities of Copenhagen and Frederiksberg (denoted as Copenhagen in the remainder of the article), regardless of the reason for sample collection. Since the 1960s, Copenhagen has been covered by an organized cervical screening program. At present, women aged 23 to 49 years are targeted for screening every 3 years, and women aged 50 to 65 years are targeted for screening every 5 years. In 2011, 76% of women in the Capital Region, which includes Copenhagen, were screened at least once within the recommended interval.23
Sample Collection Horizon was nested into routine laboratory practice. Samples were identified by the internal laboratory specimen identifier and a barcode before they were divided sequentially into racks of 48 samples. For this study, samples were collected from June 10, 2011, to August 25, 2011, equally from Monday to Friday. Based on capacity and processing considerations at the molecular pathology laboratory, the target number of samples was set to 5000. After completion of routine LBC evaluation [which includes HC2 triage of women aged 30 years and older with atypical squamous cells of undetermined significance (ASCUS)], approximately 2 mL of residual cellular suspension of SurePath per sample was collected for Horizon from each of the tubes in, at most, the first four racks processed on the sampling days. This method mimicked collection of unselected consecutive LBC samples, assuming that the time of day that the sample arrived in the laboratory was not associated with any specific characteristics. In accordance with the protocol agreed on with the assay manufacturers before the study, collected samples were diluted with 2 mL of SurePath (dilution factor, approximately 1:1) to obtain enough volume for additional testing, which will be reported separately (unpublished data). All testing was performed in the same laboratory and by the same staff.
Cytology Cytologic evaluation of SurePath samples was performed by cytoscreeners, in collaboration with a pathologist in cases of abnormal findings. Reading of the specimens was assisted by the FocalPoint GS Imaging System (SlideWizard; BD, Burlington, NC). Cytologic findings were reported using the Bethesda 2001 system, and abnormalities were classified as ASCUS, low-grade squamous intraepithelial lesions, or highgrade squamous intraepithelial lesions or worse. The last category includes atypical squamous cells, which cannot exclude high-grade squamous intraepithelial lesions, atypical glandular cells, adenocarcinoma in situ, and carcinoma. Samples with ASCUS or worse (ASCUS) were considered to have abnormal cytology. Cytoscreeners and pathologists were blinded to the results of HPV testing in Horizon.
671
Rebolj et al
APTIMA HPV mRNA Testing For handling of SurePath samples, we followed a protocol provided for this study by Hologic/Gen-Probe. Diluted sample (1 mL) was aliquoted into an APTIMA Specimen Transfer Tube containing 2.9 mL of buffered solution (Hologic/Gen-Probe). Before APTIMA testing, proteinase K treatment was performed using the Pace 2 Fast Expression Kit containing 1 mL diluent and lyophilized reagent (all from Hologic/Gen-Probe). Reconstituted proteinase K (100 mL) was added to each Specimen Transfer Tube and incubated at 65 C for 2 hours. The treated specimen tube was stored at 2 C to 8 C until testing. Testing was performed on the PANTHER platform. After lysis of the cells, mRNA was captured using oligonucleotide probes attached to magnetic beads. The released target mRNA molecules were amplified using transcription-mediated amplification, in which two different primers, nucleotides, the T7 RNA polymerase and the Moloney murine leukemia virus reverse transcriptase with RNase H activity produced mRNA amplicons in an isothermal reaction. The amplicons were hybridized to DNA probes labeled with an acridinium ester detector molecule, leading to photon emission. The light signal (relative light units) was detected on a luminometer and a signal to cut-off ratio (S/CO) was calculated from the positive and negative calibrators loaded onto the machine. An internal control was included in each sample as supplied in the kit.6 Positive APTIMA samples were defined in accordance with the manufacturer’s specifications as an S/CO of 0.5 or greater. We also evaluated APTIMA’s intralaboratory reproducibility in a routine setting, based on a selection of 361 negative and 375 positive APTIMA samples from different Horizon batches. These samples were retested once within an average of 8 days (range, 1 to 50 days). For samples with nonreproducible APTIMA results (disagreement between initial and retest outcomes), viral DNA was extracted from separate 1-mL aliquots using a LC Total Nucleic Assay kit on the automated LC96 MagnaPure platform (Roche Diagnostics, Rotkreuz, Switzerland) and subsequently analyzed using the CLART HPV2 genotyping assay (Genomica, Madrid, Spain) according to the manufacturer’s specifications. CLART targets L1 genomic sequences of the DNA from 35 HPV genotypes.
Hybrid Capture 2 HPV DNA Testing HC2 is widely considered the standard HPV DNA assay.24 It is a hybridization assay detecting 13 high-risk genotypes (genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) in combination. The assay has no internal control for sufficiency of test material. Sample DNA was either denatured before testing by pretreating manually according to the manufacturer’s protocol (n Z 5043), or DNA was isolated and purified using the DSP AXpH DNA kit on the QIASymphony SP platform (Qiagen, Hilden, Germany) (n Z
672
1086). Testing of these samples was performed on the automated Rapid Capture System (Qiagen, Gaithersburg, MD) using different scripts depending on pretreatment. A minority of samples used for routine triage of women aged 30 years and older with ASCUS were denatured and tested manually (n Z 106). The remaining 23 samples were not tested with HC2. Positive HC2 samples were defined as those having a relative light unit/cut-off value of 1 or greater.
Screening History The screening history of women from January 1, 2000, onward was retrieved from the Danish Pathology Data Bank.25 Horizon samples with an earlier diagnosis of cervical cancer, a CIN diagnosis up to 3 years earlier, ASCUS in the previous 15 months, or more severe cytologic abnormalities or a positive HPV test in the past 12 months were considered follow-up samples. Horizon samples with no recent abnormality were considered primary samples. Reflecting routine practice, primary samples included screening samples and a small proportion of samples taken by indication.
Statistical Analysis Differences in the distributions of age, screening history, cytology, and HC2 results between the included and excluded samples were tested with the c2 test. Trends in HPV positivity by age were tested with the Mantel-Haenszel c2 test for trend. Agreement between the APTIMA and HC2 assays was measured by the k coefficient. The 95% CIs for the differences in the relative proportions (RPs) of women testing positive on APTIMA or HC2 by cytology grade were calculated by assuming that the logarithms of RPs were approximately normally distributed.
Ethical Considerations Horizon was designed as a quality development study, using only residual material that otherwise would have been discarded. According to Danish regulations of biomedical research, published on May 5, 2011, in the Guidelines about Notification of a Biomedical Research Project to the Committee System on Biomedical Research Ethics no. 9154 section 2.5, quality development studies do not require ethical approval.
Results Study Sample Of the 12,138 samples processed by the laboratory during the sample collection period, 6258 (52%) were collected and processed for the Horizon study. All collected samples were tested with LBC, and all but 23 samples (0.4%, for technical reasons) were tested with HC2. In total, 1188 (19.0%) of the 6258 samples were not tested with APTIMA owing to insufficient volume for the complete study (n Z 1165) or because of technical reasons
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples (n Z 23), and were excluded from the analysis. There was no significant difference in the distribution of age (P Z 0.08), cytology (P Z 0.46), or HC2 result (P Z 0.31) between the 5070 included and the 1188 excluded samples. There was a small but significant difference in women’s screening history, with 87.0% of the included and 89.9% of the excluded samples being primary (P Z 0.01). Among the 5070 samples, 5011 (98.8%) women had one sample collected during the course of the study, whereas 59 (1.2%) samples were from 29 women, reflecting consecutive sampling from routine practice. The mean age of the women was 37.3 years (SD, 12.3 years; range, 16 to 89 years), and 64% of the samples were obtained from women younger than age 40 years (Table 1), reflecting a predominantly young source population (Statistics Denmark, http://www.statbank.dk/folk1; last accessed, April 11, 2013). The mean number of days between the receipt of the specimen in the laboratory and storage in the APTIMA Specimen Transport Tubes was 2 (range, 1 to 5 days). The mean number of days between storing and proteinase K treatment was 81 (range, 45 to 112 days); and the mean number of days between proteinase K treatment and testing with APTIMA was 12 (range, 0 to 25 days). All of these times were in accordance with the protocol agreed on with the manufacturer before the study.
seven of these, whereas for the remaining four samples there were errors in the testing process (data analysis error, failure in sample dispense verification, and/or incorrect sample volume). One sample (<0.1%) remained invalid after the testing had been repeated. APTIMA was positive in 849 (16.7%) samples. For women targeted by the screening program (age, 23 to 65 years), the proportion testing positive was 16.1%. This proportion decreased from 27.8% at ages 23 to 29 years to 3.8% at ages 60 to 65 years (P < 0.0001 for trend across age categories) (Table 1). In women ages 16 to 22 and older than age 65, the proportions testing positive on APTIMA were 44.4% and 6.2%, respectively. Women in this age range are not invited routinely for screening, but may have presented for medical conditions or may have sought screening themselves. In total, 38 (0.7%) samples tested positive on APTIMA at S/CO values of 0.50 or greater and less than 1.00. From the 5064 samples that were tested with HC2, 1035 (20.4%) tested positive. Although the proportion testing positive was higher overall than with APTIMA, it showed a similar decrease with increasing age. Among women ages 16 to 22 years, 56.8% of samples tested positive on HC2, 33.0% of samples tested positive at ages 23 to 29 years, 17.4% of samples tested positive at ages 30 to 39 years, 11.3% of samples tested positive at ages 40 to 49 years, 7.3% of samples tested positive at ages 50 to 59 years, 6.0% of samples tested positive at ages 60 to 65 years, and 6.2% of samples tested positive at age older than 65 years (data not shown). Of the 846 APTIMA-positive samples for which HC2 results were
HPV Testing Overall, APTIMA testing had to be repeated on 11 (0.2%) samples to obtain valid results. The internal control failed in Table 1
APTIMA Results for 5070 Primary and Follow-Up Samples by Age, Screening History, Cytology, and HC2 Results Results on APTIMA, N (%)
Characteristics
Positive
Negative
Invalid
Total
Total Age, years 16e22 23e29 30e39 40e49 50e59 60e65 >65 Screening history Primary sample Follow-up sample Cytology Normal ASCUS LSIL HSIL ASCUS Inadequate HC2* Positive Negative
849 (16.7%)
4220 (83.2%)
1 (0.0%)
5070 (100%)
72 427 225 82 27 9 7
90 1108 1301 911 479 225 106
0 0 0 0 1 0 0
162 1535 1526 993 507 234 113
(44.4%) (27.8%) (14.7%) (8.3%) (5.3%) (3.8%) (6.2%)
(55.6%) (72.2%) (85.3%) (91.7%) (94.5%) (96.2%) (93.8%)
(0.0%) (0.0%) (0.0%) (0.0%) (0.2%) (0.0%) (0.0%)
(100%) (100%) (100%) (100%) (100%) (100%) (100%)
676 (15.3%) 173 (26.3%)
3734 (84.7%) 486 (73.7%)
1 (0.0%) 0 (0.0%)
4411 (100%) 659 (100%)
587 58 110 93 261 1
4083 65 33 14 112 25
1 0 0 0 0 0
4671 123 143 107 373 26
(12.6%) (47.2%) (76.9%) (86.9%) (70.0%) (3.8%)
747 (72.2%) 99 (2.5%)
(87.4%) (52.8%) (23.1%) (13.1%) (30.0%) (96.2%)
288 (27.8%) 3929 (97.5%)
(0.0%) (0.0%) (0.0%) (0.0%) (0.0%) (0.0%)
0 (0.0%) 1 (0.0%)
(100%) (100%) (100%) (100%) (100%) (100%)
1035 (100%) 4029 (100%)
*Six samples were not tested with HC2. HSIL, high-grade intraepithelial lesions or worse; LSIL, low-grade squamous intraepithelial lesions.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
673
Rebolj et al available, 747 (88.3%) were also HC2 positive, and of the 4217 APTIMA-negative samples with HC2 results, 3929 (93.2%) were also HC2 negative. The k coefficient of 0.75 indicated substantial agreement in these samples between the APTIMA and HC2 assays. Cytology was abnormal in 373 (7.4%) of 5070 samples. The proportion of samples testing positive on APTIMA and HC2 increased with cytologic grade. Among samples with normal cytology, 12.6% tested positive on APTIMA and 15.6% tested positive on HC2 (RP, 0.81; 95% CI, 0.73 to 0.89); among samples showing ASCUS 47.2% and 64.2% were positive, respectively (RP, 0.73; 95% CI, 0.58 to 0.92); among samples showing low-grade squamous intraepithelial lesions, 76.9% and 89.5% were positive, respectively (RP, 0.86; 95% CI, 0.77 to 0.96); and among samples with highgrade squamous intraepithelial lesions or worse, 86.9% and 90.7% were positive, respectively (RP, 0.96; 95% CI, 0.87 to 1.05) (for both assays, P < 0.0001 for trend). These results were fairly similar in the subset of 4411 primary samples (Table 2). In the primary samples, 676 (15.3%) tested positive on APTIMA, 822 (18.6%) tested positive on HC2 (k coefficient for agreement between APTIMA and HC2, 0.75), and 242 (5.5%) samples had ASCUS. Among women ages 23 to 29 years, 26.8% tested positive on APTIMA, 31.7% tested positive on HC2, and 7.4% tested positive on LBC (Table 3). Among women ages 30 to 65 years, these proportions were 9.4%, 11.7%, and 4.4%, respectively. Because women from Copenhagen are on average younger than Danish women in general, the proportion of women with positive test results would be somewhat lower if testing were performed for all women in Denmark. Table 2
APTIMA and LBC were both positive in 166 (3.8%) primary samples. The proportions of APTIMA-positive, cytology-normal primary samples strongly decreased with age (Figure 1): from 20.5% at ages 23 to 29 years, 9.7% at ages 30 to 39 years, 5.9% at ages 40 to 49 years, to 3.1% at ages 50 to 65 years (P < 0.0001 for trend).
Intralaboratory Reproducibility of APTIMA Among the 375 APTIMA-positive samples in the initial run, 318 (84.8%) had reproducible results on retesting (positive reproducibility). The average initial S/CO value of the 57 nonreproducible samples was 4.99 (range, 0.56 to 11.55), compared with 11.17 (range, 0.63 to 40.89) for reproducible samples. By using the CLART genotyping assay, 37 (64.9%) of the 57 samples contained at least 1 of the 14 HPV genotypes targeted by APTIMA, 5 (8.8%) samples contained at least 1 of the other 21 genotypes detectable by CLART but not included among the 14 APTIMA-targeted genotypes, whereas none of the 35 CLART HPV genotypes was detected in 15 (26.3%) of the 57 samples. Among the 361 negative samples in the initial run, 355 (98.3%) had reproducible results (negative reproducibility). The average S/CO value of the six nonreproducible samples was 5.01 (range, 0.58 to 10.59) in the second run.
Discussion General Findings This is the first study to report APTIMA results using SurePath samples tested on the PANTHER platform. Whether this combination affects APTIMA’s sensitivity and/or specificity
APTIMA Results for 4411 Primary Samples by Age, Screening History, Cytology, and HC2 Results Results on APTIMA, N (%)
Characteristics
Positive
Negative
Invalid
Total
Total Age, years 16e22 23e29 30e39 40e49 50e59 60e65 >65 Cytology Normal ASCUS LSIL HSIL ASCUS Inadequate HC2* Positive Negative
676 (15.3%)
3734 (84.7%)
1 (0.0%)
4411 (100%)
53 345 172 71 23 6 6
(39.3%) (26.8%) (13.2%) (7.9%) (4.9%) (2.8%) (5.7%)
82 942 1128 833 441 209 99
(60.7%) (73.2%) (86.8%) (92.1%) (94.8%) (97.2%) (94.3%)
0 0 0 0 1 0 0
(0.0%) (0.0%) (0.0%) (0.0%) (0.2%) (0.0%) (0.0%)
135 1287 1300 904 465 215 105
(100%) (100%) (100%) (100%) (100%) (100%) (100%)
510 40 65 61 166 0
(12.3%) (47.1%) (74.7%) (87.1%) (68.6%) (0.0%)
3637 45 22 9 76 21
(87.7%) (52.9%) (25.3%) (12.9%) (31.4%) (100.0%)
1 0 0 0 0 0
(0.0%) (0.0%) (0.0%) (0.0%) (0.0%) (0.0%)
4148 85 87 70 242 21
(100%) (100%) (100%) (100%) (100%) (100%)
592 (72.0%) 82 (2.3%)
230 (28.0%) 3502 (97.7%)
0 (0.0%) 1 (0.0%)
822 (100%) 3585 (100%)
*Four samples were not tested with HC2. HSIL, high-grade intraepithelial lesions or worse; LSIL, low-grade squamous intraepithelial lesions.
674
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples Table 3
APTIMA in Primary Screening: Comparison of the Horizon Data With Previously Published Studies
Age group 30 years Horizon Horizon Monsonego et al11 Ratnam et al12 Wu et al14 <30 years Horizon Monsonego et al11 Ratnam et al12
Area
N
Age group, years
Copenhagen* Denmark* Parisy Canadax Shenzhenk
2884 3320 845 2000
30e65 30e65 30e65 30e81 25e59**
Copenhagen* Parisy Canadax
1287 1109 521
23e29 20e29 16e29
Positive on APTIMA
Positive on HC2
Positive on LBC
9.4% 7.7% 8.5% >6.1%{ 10.1%
11.7% 9.7% 13.0% >7.9%{ 16.5%
4.4% 4.3% 8.3%z NR 5.4%
26.8% 15.6% >20.5%{
31.7% 23.5% >25.9%{
7.4% 13.5%z NR
*Inclusion criteria were as follows: routine primary samples, see Materials and Methods. Copenhagen: the results refer to samples included in Horizon. Denmark: the results refer to proportions from Horizon, age-standardized to all Danish women. y FASE trial. Inclusion criteria were as follows: women seen for their annual examination in 17 private gynecology practices, not having undergone total hysterectomy, nonpregnant, no cytologic abnormality in the past 6 months. Women with unsatisfactory cytology results (10.5%) were excluded from analyses. z Adjudicated cytology results. x Inclusion criteria were as follows: women undergoing routine Papanicolaou screening at two screening centers. { Lower bound, approximated as 100% specificity for CIN1. jj Inclusion criteria were as follows: women eligible for the Shenzhen Cervical Cancer Screening Trial I (nonpregnant, no cervical cancer screening for at least 3 years, no prior hysterectomy, no prior pelvic irradiation). Of the 2098 enrolled women, 3 did not have complete data for APTIMA, HC2, or LBC, and 95 women with at least one positive test result did not return for colposcopy. These 98 women were excluded from the study. Women were medically underserved and predominantly urban. **Reported as 24 to 59 years in tables from the original publication. Data were not reported separately by age group. NR, not reported.
16−22 Yrs N=135 Cytology − Cytology + 23−29 Yrs N=1287
16−89 Yrs N=4411
30−39 Yrs N=1300 40−49 Yrs N=904 50−59 Yrs 60−65 Yrs N=465 N=215
>65 Yrs N=105
30−65 Yrs N=2884
-
jmd.amjpathol.org
10% 20%
Although 19% of the collected samples had to be discarded, there was no significant difference between the included and the excluded samples in terms of women’s age, cytology, or the HC2 result. All samples were tested in one laboratory by the same staff that is involved in routine screening. In a 2004 to
% Cytology Positive
0
Strengths and Weaknesses
The Journal of Molecular Diagnostics
20%
% HPV−Positive
30%
40%
2005 study, a total of 11,617 (primary) consecutive SurePath samples from the same area evaluated in the same laboratory were tested with HC2.22 The median age of the women in that study was 36.4 years (range, 15 to 93 years), and 6% had abnormal cytology. The proportion of women aged 25 to 64 years testing positive on HC2 was approximately 17%, which is similar to the 16% in Horizon. Horizon samples therefore can be considered representative for the source population. Following routine recommendations from the Danish cervical screening program, women with abnormal cytology were referred for colposcopy or for repeat testing. As a follow-up to Horizon, we will invite women with positive HPV tests and normal cytology for repeat testing in about a year. The standard measures of sensitivity and specificity of the screening tests will be reported once this histologic follow-up evaluation of the study population has been completed.
10%
for detecting high-grade CIN has not yet been described in the literature. In Horizon, the overall prevalence of HPV infection measured by APTIMA was 17%, compared with 20% on HC2, whereas 7% of all samples had abnormal cytology. The prevalence of HPV infection on APTIMA decreased with age and increased with cytology grade. These findings all conform to expectations based on prior studies using PreservCyt samples. The results suggest that APTIMA performed well on SurePath samples, but this remains to be confirmed with histologic outcomes collected in a planned follow-up study to Horizon. The positive reproducibility observed for APTIMA was only 85%, which might signal a need to further optimize the protocol for storing and handling SurePath samples with respect to molecular target stability. It also is noteworthy that the nonreproducible positive samples had on average a lower S/CO value to start. A lack of positive reproducibility may pose problems for laboratories, for example, in terms of their ability to verify the initial results on issues of patient safety, and to show reliable quality assurance and control procedures. However, a low positive reproducibility to some degree might be balanced by a low frequency of positive tests. Furthermore, in routine settings, action is taken only on a single test, therefore low positive reproducibility would not constitute a problem for an assay’s sensitivity for CIN.
HPV + HPV −
Figure 1 Proportion of women with a positive primary APTIMA test and abnormal primary cytology by age.
675
Rebolj et al
Comparison With Other Studies and Implications for Screening A majority of studies published to date on APTIMA testing was performed in women referred for colposcopy because of prior abnormal cytology (ASCUS). In these disease-enriched populations, the proportions of women whose PreservCyt samples tested positive on APTIMA were typically on the order of 50% to 80%.10,12,13,15,20,26 This was comparable with our study, in which 69% of primary samples with ASCUS tested positive (Table 2). In women with ASCUS aged 30 years and older from Jutland, Denmark, 31.7% tested positive on APTIMA.8 Similarly, in a Norwegian study using 6- to 12month follow-up samples of women with ASCUS or lowgrade squamous intraepithelial lesions, of whom 80% were aged older than 30 years, 32.0% tested positive on APTIMA.9 Among the 86 women with ASCUS aged 30 years and older in Horizon, 39.5% tested positive on APTIMA. The findings of the Horizon study using SurePath samples thus correspond well with the findings of previous studies based on PreservCyt samples. Although HPV DNA screening has been strongly discouraged for women younger than age 30 years because of its lack of specificity for high-grade CIN,1,27 conclusions from the French APTIMA Screening Evaluation study (FASE) supported the use of APTIMA in women younger than age 30 years.11 In this age group, FASE researchers estimated APTIMA’s specificity for CIN3 at 84.9% (95% CI, 82.8 to 87.0), almost the same as that of LBC (86.9%; 95% CI, 84.9 to 88.9), and better than that of HC2 (76.9%; 95% CI, 74.4 to 79.4). In Horizon, the proportion of women aged 23 to 29 years testing positive on LBC was about half of that in FASE, whereas the proportion testing positive on APTIMA was about twice as high (Table 3). This large difference between the proportions testing positive on APTIMA and on LBC in Horizon probably will lead to a larger difference in the specificity for the two screening tests than was observed in FASE. In FASE, APTIMA’s good specificity compared with LBC may reflect, among other factors, a relatively high proportion of women with ASCUS. Proportions of women aged 30 years and older testing positive on APTIMA were similar in all four reported screening studies, including Horizon (range, >6% to 10%).11,12,14 The proportion of women with a positive LBC was about 4% to 5% in the Chinese Shenzhen and Horizon studies, but again was higher, at 8%, in FASE. Nonetheless, all available screening trials observed a lower proportion of positive tests by APTIMA than by HC2. The only two studies to date that used a manufacturerrevised cut-off value for positive APTIMA resultsd0.50 or greater S/CO instead of 1.00 or greater S/COdwere a Canadian screening study12 and Horizon. A wider definition of a positive result could help improve the assay’s sensitivity for high-grade CIN, but also could decrease its specificity. In Horizon, the proportion of samples with S/CO values between 0.50 and less than 1.00 was only 0.7% (1.0% in women aged 23 to 29 years, and 0.6% in women aged 30 to 65 years).
676
Based on these baseline data it does not appear that the lowered cut-off value for a positive APTIMA test substantially decreased the assay’s specificity. In light of the high cost and limited capacity for colposcopy in many countries, including Denmark, the high proportion of women with positive HPV DNA tests calls for triage testing. It appears that triage also will be necessary for women with positive mRNA tests. Cytology may be a reasonable candidate triage test. In Horizon, 6.3% of women with a primary test at age 23 to 29 years tested positive on both LBC and APTIMA (Figure 1). Among women aged 30 to 65 years, this was 2.5%. The remaining 20.5% (26.8% 6.3%) among young women, and 6.9% (9.4% 2.5%) among older women, who tested positive on APTIMA and negative on LBC, would wait several months for repeat testing. As stated earlier, histologic verification of positive screening tests in Horizon is underway. Nevertheless, some preliminary estimates of the frequency of false-positive test results, defined as positive tests without histologic CIN3, can be calculated based on recent historical data for Copenhagen. These data, retrieved from the Pathology Data Bank, showed that 1.3% of all screened women had CIN3 detected after abnormal cytology (ASCUS). If HPV testing is 32% more sensitive for CIN3 than cytology,28 about 1.7% of women are expected to have CIN3 detected after a positive HPV test. Therefore, 7.7% (9.4% 1.7%) of women aged 30 to 65 years would have a false-positive APTIMA test result with respect to the clinical outcome. This would be 10.0% (11.7% 1.7%) for HC2, and 3.1% (4.4% 1.3%) for LBC. Among women aged 23 to 29 years, the proportions would be 25.5% (27.2% 1.7%) for APTIMA, 30.5% (32.2% 1.7%) for HC2, and 6.2% (7.5% 1.3%) for LBC. Because of a high expected frequency of false-positive test results in countries with a high prevalence of HPV infections, caution therefore is needed when introducing HPV-based screening, even in the case of highly sensitive26 mRNA assays. Optimization of the LBC media with molecular HPV testing in mind, or vice versa, will be a critical issue for future implementation of primary HPV mRNA or DNA testing with cytology triage, and studies using SurePath are highly relevant given its substantial market share. An alternative could be to introduce a two-tier sampling process, requiring either that two samples are taken per examination (one in molecular collection media, and one in LBC media), or that initially one sample is taken from all women but that those who require triage are recalled for an additional test. However, a two-tier sampling process would be associated with extra costs; the need for a recall also might induce unnecessary anxiety, and reduce the effectiveness of screening as a result of loss to follow-up evaluation.
Conclusions Based on data from Copenhagen, an area with a high prevalence of HPV infections, fewer women tested positive on
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
APTIMA mRNA HPV on SurePath Samples APTIMA than on HC2. Baseline data furthermore suggest that APTIMA performed well using SurePath samples on the PANTHER platform.
13.
Acknowledgments We thank Miki Hansen and Maria Louise Deistler (Hvidovre University Hospital) for excellent laboratory assistance; Bryan Howard Goldman (University of Copenhagen) for visual representation of the data and help with statistical analysis; the Danish Hologic/Gen-Probe, Qiagen, and Genomica representatives for enabling the collaboration between their parent companies; Hvidovre University Hospital; and the University of Copenhagen.
14.
15.
16.
References 1. Ronco G, Giorgi-Rossi P, Carozzi F, Confortini M, Dalla Palma P, Del Mistro A, Ghiringhello B, Girlando S, Gillio-Tos A, De Marco L, Naldoni C, Pierotti P, Rizzolo R, Schincaglia P, Zorzi M, Zappa M, Segnan N, Cuzick J: Efficacy of human papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial neoplasia: a randomised controlled trial. Lancet Oncol 2010, 11:249e257 2. Rijkaart DC, Berkhof J, Rozendaal L, van Kemenade FJ, Bulkmans NW, Heideman DA, Kenter GG, Cuzick J, Snijders PJ, Meijer CJ: 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, 13:78e88 3. Rebolj M, Pribac I, Lynge E: False-positive human papillomavirus DNA tests in cervical screening: it is all in a definition. Eur J Cancer 2011, 47:255e261 4. Doorbar J: Molecular biology of human papillomavirus infection and cervical cancer. Clin Sci (Lond) 2006, 110:525e541 5. Cuschieri K, Wentzensen N: Human papillomavirus mRNA and p16 detection as biomarkers for the improved diagnosis of cervical neoplasia. Cancer Epidemiol Biomarkers Prev 2008, 17:2536e2545 6. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, IbrahimTallaa L, Guha N, Freeman C, Galichet L, Cogliano V: A review of human carcinogensepart B: biological agents. Lancet Oncol 2009, 10:321e322 7. Szarewski A, Ambroisine L, Cadman L, Austin J, Ho L, Terry G, Liddle S, Dina R, McCarthy J, Buckley H, Bergeron C, Soutter P, Lyons D, Cuzick J: Comparison of predictors for high-grade cervical intraepithelial neoplasia in women with abnormal smears. Cancer Epidemiol Biomarkers Prev 2008, 17:3033e3042 8. Waldstrom M, Ornskov D: Comparison of the clinical performance of an HPV mRNA test and an HPV DNA test in triage of atypical squamous cells of undetermined significance (ASC-US). Cytopathology 2012, 23:389e395 9. Ovestad IT, Vennestrom U, Andersen L, Gudlaugsson E, Munk AC, Malpica A, Feng W, Voorhorst F, Janssen EA, Baak JP: Comparison of different commercial methods for HPV detection in follow-up cytology after ASCUS/LSIL, prediction of CIN2-3 in follow up biopsies and spontaneous regression of CIN2-3. Gynecol Oncol 2011, 123:278e283 10. Dockter J, Schroder A, Hill C, Guzenski L, Monsonego J, Giachetti C: Clinical performance of the APTIMA HPV Assay for the detection of high-risk HPV and high-grade cervical lesions. J Clin Virol 2009, 45(Suppl 1):S55eS61 11. Monsonego J, Hudgens MG, Zerat L, Zerat JC, Syrjanen K, Halfon P, Ruiz F, Smith JS: 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, 129:691e701 12. Ratnam S, Coutlee F, Fontaine D, Bentley J, Escott N, Ghatage P, Gadag V, Holloway G, Bartellas E, Kum N, Giede C, Lear A: Aptima
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
17.
18.
19.
20.
21. 22.
23.
24.
25. 26.
27. 28.
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, 49:557e564 Clad A, Reuschenbach M, Weinschenk J, Grote R, Rahmsdorf J, Freudenberg N: Performance of the APTIMA high-risk HPV mRNA assay in a referral population in comparison with Hybrid Capture 2 and cytology. J Clin Microbiol 2011, 49:1071e1076 Wu R, Belinson SE, Du H, Na W, Qu X, Wu R, Liu Y, Wang C, Zhou Y, Zhang L, Belinson JL: Human papillomavirus messenger RNA assay for cervical cancer screening: the Shenzhen Cervical Cancer Screening Trial I. Int J Gynecol Cancer 2010, 20:1411e1414 Reuschenbach M, Clad A, von Knebel DC, Wentzensen N, Rahmsdorf J, Schaffrath F, Griesser H, Freudenberg N, von Knebel DM: Performance of p16INK4a-cytology. HPV mRNA, and HPV DNA testing to identify high grade cervical dysplasia in women with abnormal screening results. Gynecol Oncol 2010, 119:98e105 Arbyn M, Roelens J, Cuschieri K, Cuzick J, Szarewski A, Ratnam S, Reuschenbach M, Belinson S, Belinson JL, Monsonego J: The APTIMA HPV assay versus the hybrid capture 2 test in triage of women with ASC-USor LSIL cervical cytology: a meta-analysisof the diagnostic accuracy. Int J Cancer 2013, 132:101e108 Horvath CA, Boulet G, Sahebali S, Depuydt C, Vermeulen T, van den Broeck D, Vereecken A, Bogers J: Effects of fixation on RNA integrity in a liquid-based cervical cytology setting. J Clin Pathol 2008, 61: 132e137 Powell N, Smith K, Fiander A: Recovery of human papillomavirus nucleic acids from liquid-based cytology media. J Virol Methods 2006, 137:58e62 Murphy PG, Henderson DT, Adams MD, Horlick EA, Dixon EP, King LM, Avissar PL, Brown CA, Fischer TJ, Malinowski DP: Isolation of RNA from cell lines and cervical cytology specimens stored in BD SurePath preservative fluid and downstream detection of housekeeping gene and HPV E6 expression using real time RT-PCR. J Virol Methods 2009, 156:138e144 Waldstrom M, Ornskov D: Clinical performance of a human papillomavirus messenger RNA test (Aptima HPV Assay) on residual material from archived 3-year-old PreservCyt samples with low-grade squamous intraepithelial lesion. Arch Pathol Lab Med 2011, 135:1052e1056 Clemmesen J, Nielsen A: The social distribution of cancer in Copenhagen, 1943 to 1947. Br J Cancer 1951, 5:159e171 Kjaer SK, Breugelmans G, Munk C, Junge J, Watson M, Iftner T: Population-based prevalence, type- and age-specific distribution of HPV in women before introduction of an HPV-vaccination program in Denmark. Int J Cancer 2008, 123:1864e1870 Styregruppen for DKLS: Årsrapport DKLS 2011: Dansk Kvalitetsdatabase for Livmoderhalskræftscreening [annual report 2011: Danish Quality Assurance database for cervical cancer screening]. Hvidovre, DKLS, 2012, pp 1e122 Meijer CJ, Berkhof J, Castle PE, Hesselink AT, Franco EL, Ronco G, Arbyn M, Bosch FX, Cuzick J, Dillner J, Heideman DA, Snijders PJ: Guidelines for human papillomavirus DNA test requirements for primary cervical cancer screening in women 30 years and older. Int J Cancer 2009, 124:516e520 Bjerregaard B, Larsen OB: The Danish Pathology Register. Scand J Public Health 2011, 39:72e74 Szarewski A, Mesher D, Cadman L, Austin J, Ashdown-Barr L, Ho L, Terry G, Liddle S, Young M, Stoler M, McCarthy J, Wright C, Bergeron C, Soutter WP, Lyons D, Cuzick J: Comparison of seven tests for high-grade cervical intraepithelial neoplasia in women with abnormal smears: the predictors 2 study. J Clin Microbiol 2012, 50:1867e1873 Castle PE: Screening: hPV testing for cervical cancer: the good, the bad, and the ugly. Nat Rev Clin Oncol 2010, 7:364e365 Cuzick J, Arbyn M, Sankaranarayanan R, Tsu V, Ronco G, Mayrand MH, Dillner J, Meijer CJ: Overview of human papillomavirus-based and other novel options for cervical cancer screening in developed and developing countries. Vaccine 2008, 26(Suppl 10):K29eK41
677