Analytical Comparison of the cobas HPV Test with Hybrid Capture 2 for the Detection of High-Risk HPV Genotypes

The Journal of Molecular Diagnostics, Vol. 14, No. 1, January 2012 Copyright © 2012 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. DOI: 10.1016/j.jmoldx.2011.09.005

Analytical Comparison of the cobas HPV Test with Hybrid Capture 2 for the Detection of High-Risk HPV Genotypes

Maria Luisa Mateos Lindemann,* Mario Jose Rodriguez Dominguez,* Jesús Chacón de Antonio,* Maria Teresa Sandri,† Alessio Tricca,† Mario Sideri,‡ Hacène Khiri,§ Sophie Ravet,§ Sean Boyle,¶ Carrie Aldrich,¶ and Philippe Halfon§ From the Microbiology Department,* Hospital Ramón y Cajal, Madrid, Spain; the Units of Laboratory Medicine,† and Preventive Gynecology,‡ European Institute of Oncology, Milan, Italy; the Alphabio Laboratory,§ Marseille, France; and Roche Molecular Systems,¶ Pleasanton, California

Human papillomavirus (HPV) is a causal agent of cervical cancer, and persistent HPV16 or HPV18 infection carries a particularly high risk. The cobas HPV Test (cobas) provides individual HPV16/HPV18 genotyping with a simultaneous result for 12 other high-risk HPV (hrHPV) genotypes. Its analytical performance for hrHPV genotype detection was retrospectively evaluated against the digene Hybrid Capture 2 HPV DNA test (HC2), in three European centers, in 1360 cervical samples. Both HPV tests performed similarly, with no significant difference in the number of positive and negative samples identified by each test and good agreement between the tests was observed. Discordant samples were analyzed with the Linear Array HPV genotyping test. More low-risk HPV (lrHPV) genotypes were detected in HC2-positive/cobas-negative samples compared with HC2-negative/cobas-positive samples. Conversely, more hrHPV genotypes were detected in HC2-negative/cobas-positive samples compared with HC2-positive/cobas-negative samples. Eight HC2-negative/cobas-positive samples were positive for HPV16 compared with five HC2-positive/cobas-negative samples; HPV18 was detected in one HC2-negative/cobaspositive sample and one HC2-positive/cobas-negative sample. The cobas HPV Test demonstrates comparable analytical performance to the HC2 test, but with a lower rate of cross-reactivity with lrHPV genotypes, and has the advantage of simultaneously providing HPV16/ HPV18 identification. ( J Mol Diagn 2012, 14:65–70; DOI: 10.1016/j.jmoldx.2011.09.005)

Human papillomavirus (HPV) is recognized as a causal agent of cervical cancer.1–3 Although ⬎100 mucosal HPV genotypes have been identified, only a subset has been associated with cervical lesions. This subset has been further categorized into high-risk HPV (hrHPV) genotypes (associated with precancerous or cancerous cervical lesions) and low-risk HPV (lrHPV) genotypes (rarely associated with precancerous or cancerous cervical lesions, cause of genital warts). Overall, 14 hrHPV genotypes (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) have been detected in ⬎99% of cervical cancers worldwide.3 Because of the causal link between HPV and cervical cancer, the potential clinical value of using HPV DNA detection methods to identify those women most at risk of having high-grade cervical disease (defined as cervical intraepithelial neoplasia [CIN] 2 or greater) is clear. HPV DNA testing can identify ⬎9 of every 10 women with high-grade cervical disease, with a reported clinical sensitivity of HPV DNA testing for the detection of ⱖCIN2 of approximately 95% in a screening population.4,5 This compares with a reported clinical sensitivity of cytology testing for the detection of ⱖCIN2 of approximately 55%, with a single round of testing.4,5 The reduction in specificity observed with HPV DNA testing compared with cytology4,5 and the absence of effective algorithms have restricted its use as a primary cervical cancer screening tool; however, HPV DNA testing is currently recommended in both the United States and Europe for the triage of all women with equivocal cytology,6,7 and in the United States

Supported by Roche Diagnostics, Pleasanton, California. Support for third-party writing assistance for this manuscript was provided by Roche Diagnostics. Accepted for publication September 1, 2011. Disclosures: S.B. is a principal scientist at Roche Molecular Systems, the sponsor of the study. C.A. is an employee of Ventana Medical Systems Inc., a member of the Roche group, and was an employee of Roche Molecular Systems at the time of the study. M.T.S., M.L.M.L., and P.H. are consultants for Roche Molecular Systems. M.S. has performed consultancies with no salary for: GSK, Sanofi Pasteur, MTM Labs, Digene Qiagen, Innogenetics, and Archimed Labs. Current address of C.A., Ventana Medical Systems Inc., Tucson, Arizona. Address reprint requests to Maria Luisa Mateos Lindemann, Ph.D., Microbiology Department, Ramón y Cajal Hospital, Carretera de Colmenar Viejo Km 9.1, 28034 Madrid, Spain. E-mail: mmateos.hrc@salud. madrid.org.

65

66 Mateos Lindemann et al JMD January 2012, Vol. 14, No. 1

for routine cervical cancer screening in conjunction with cytology testing for women ⱖ30 years of age.7 The risk of cervical precancer and cancer varies according to hrHPV genotype. HPV16 and HPV18 are detected in approximately 70% of invasive cervical cancers8,9 and women with persistent HPV16 or HPV18 infection have been reported to have the highest 10-year cumulative incidence rate of high-grade cervical disease and cancer (ⱖCIN3) when compared with women positive for non-HPV16/18 hrHPV types and hrHPV-negative women.10 Accordingly, the potential for further stratifying cervical precancer/cancer risk by individual HPV16/18 genotyping in certain populations of women has been acknowledged.7 The cobas HPV Test (Roche Molecular Systems, Inc., Branchburg, NJ) is an automated qualitative in vitro test for the detection of 14 hrHPV genotypes. The test targets the highly conserved L1 region of the HPV genome using primer pairs directed to amplify 14 hrHPV genotypes; genotype-specific fluorescent oligonucleotide probes bind to polymorphic regions within the sequence amplified by these primers. The cobas HPV Test identifies HPV16 and HPV18 genotypes separately, while simultaneously detecting 12 other hrHPV genotypes (HPV31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) at clinically relevant infection levels. The clinical sensitivity and specificity of the cobas HPV Test has previously been demonstrated to be comparable with the current standard of HPV DNA testing, the digene Hybrid Capture 2 (HC2) high-risk HPV DNA test (Qiagen, Gaithersburg, MD) in women with atypical squamous cells of undetermined significance cytology in a large US trial.11 In the same trial, the clinical utility of HPV16/18 genotyping with the cobas HPV Test was demonstrated in women ⱖ30 years of age who had high-grade cervical disease undetected by cytology screening.12 The cobas HPV Test was Conformité Européenne (European Community) CE-marked in 2009 and received US Food and Drug Administration approval in April 2011. CE-marking is a key indicator of a product’s compliance with European Union legislation and enables the free movement of products within the European market. The cobas Early Evaluator Program is a European multicenter study to compare the analytical and clinical performance of the cobas HPV Test against the current standard of HPV DNA testing, HC2, for the detection of hrHPV genotypes and the identification of high-grade cervical disease in different populations of women attending obstetrics/gynecology clinics. This study reports on the analytical comparison of the cobas HPV Test with the HC2 test.

Materials and Methods Evaluated Samples This study used a convenience sample of 1360 consecutive cervical specimens collected from women participating in routine screening or follow-up of Pap abnormalities and who were willing to enter the study. Specimens were submitted for HC2 testing between May and Octo-

ber 2008 to the three clinical centers participating in the cobas Early Evaluator Program (Hospital Ramón y Cajal, Madrid, Spain; Alphabio Laboratory, Marseille, France; and European Institute of Oncology, Milan, Italy). Specimens were not linked to patient identification data (and so not linked to medical history), but were linked to age of patient, and cytologic and histological results. Cervical specimens were stored in liquid cytology vials containing PreservCyt solution (Cytyc, Marlborough, MA) and kept at 4°C until the start of the study. Institutional Review Board approval was obtained from all participating centers to test residual laboratory samples for each site.

HPV Testing All specimens were tested using HC2 (Qiagen) and the prototype cobas HPV Test (Roche Molecular Systems, Inc.). At the Madrid and Marseille sites, all specimens were also tested with the Linear Array HPV (LA HPV) genotyping test (Roche Molecular Systems, Inc.). For the Milan site, only samples testing positive with the HC2 test were genotyped with the LA HPV Test. Because of this verification bias, the results of the LA HPV test for samples with HC2/cobas HPV Test discordant results are confined to the Madrid and Marseille samples.

HC2 High-Risk HPV DNA Test All cervical specimens were analyzed for the presence or absence of 13 hrHPV genotypes (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) with the HC2 test (providing a pooled result). The HC2 test was performed according to the manufacturer’s instructions, using the HC2 sample conversion kit (Qiagen) for PreservCyt medium. If the relative light unit/cutoff ratio of the sample was ⱖ1.0 and ⬍2.5, the sample was recorded as indeterminate, as per manufacturer’s instructions, and repeat tested.13 Samples repeatedly testing indeterminate were reassigned as an HC2-positive result for subsequent analysis here.

cobas HPV Test The prototype cobas HPV Test was carried out according to the manufacturer’s protocol for the detection of 14 hrHPV genotypes (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68).14 The test specifically identifies HPV16 and HPV18 while concurrently detecting the rest of the high-risk types, as a pooled result, at clinically relevant infection levels. Briefly, DNA was extracted via a fully automated sample preparation process using the cobas ⫻ 480 instrument and samples were transferred to a cobas z 480 analyzer for PCR amplification of hrHPV DNA. Complementary primer pairs were used to amplify the highly conserved L1 region of the HPV genome, and fluorescent oligonucleotide probes specific for HPV16, HPV18, and the 12 other hrHPV types allowed real-time detection of the 12 hrHPV genotypes (pooled result) and individual realtime detection of HPV16 and HPV18, in one sample. Detection of the human ␤-globin gene was performed in the same sample to provide a measure of sample adequacy and of

Comparison of cobas HPV Test and HC2 67 JMD January 2012, Vol. 14, No. 1

the quality of extraction and amplification (internal control), and positive and negative controls were included in each run. Interpretation of the amplification and detection stage was carried out using software supplied with the cobas 4800 platform. The cycle threshold cutoffs were previously validated in a large cohort of women11 and set at 40.0 for channel 1 (12 other hrHPV), 40.5 for channel 2 (HPV16), and 40.0 for channel 3 (HPV18).

LA HPV Genotyping Test The LA HPV test was carried out according to the manufacturer’s instructions for the individual detection of 37 HPV genotypes (HPV6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51–56, 58, 59, 61, 62, 64, 66, 67–73 (MM9), 81, 82 (MM4), 83 (MM7), 84 (MM8), 89 [CP6108], and IS39).15 As with the cobas HPV Test, the human ␤-globin gene was monitored in the same sample to provide a measure of sample adequacy, and positive and negative controls were included in each test.

Statistical Analysis Concordance was calculated assuming a standard 2 ⫻ 2 contingency table. The level of agreement between the cobas HPV Test and the HC2 Test in detecting hrHPV genotypes was assessed using Cohen’s kappa statistic, with values of 0.00 to 0.20 indicating poor agreement, 0.21 to 0.40 indicating fair agreement, 0.41 to 0.60 indicating moderate agreement, 0.61 to 0.80 indicating good agreement, and 0.81 to 1.00 indicating excellent agreement. The 95% CIs were calculated on the Cohen’s kappa values assuming a binomial distribution. Marginal homogeneity of the two tests was assessed by a twosided McNemar test and a P ⬍ 0.05 was considered significant. Discordant samples were analyzed in each LA HPV category (lrHPV only, hrHPV positive, HPV negative) using the Pearson ␹2-test.

Results Population Demographics Cervical samples from 1360 women were included in the study; 421 women were enrolled in Madrid, 212 in Marseille, and 727 in Milan. The median age of women in the study was 37 years (range, 16 to 82 years); ⬎70% of women in the study were ⱖ30 years of age. In the Madrid cohort, approximately 15% of women were attending for routine screening and 85% of women were attending for follow-up of an abnormal Pap smear test result, with a similar situation reported by the Milan site. In the Marseilles cohort, approximately 20% of women were attending for routine screening and 80% of women were attending for follow-up of an abnormal Pap smear test result.

Table 1. Comparison of the HC2 Test and the cobas HPV Test for the Detection of hrHPV Genotypesⴱ† HC2 hrHPV hrHPV positive negative Indeterminate‡ Total cobas HPV Test hrHPV positive hrHPV negative Total

552 86 638

71 638 709

9 4 13

632 728 1360

*Cohen’s kappa coefficient ⫽ 0.7625 (95% CI: 0.7281– 0.7969) (with the inclusion of HC2 indeterminate samples as an hrHPV-positive result). † McNemar test, P ⫽ 0.1588. ‡ Defined as a sample with a relative light unit/cutoff ratio of ⱖ1.0 but ⬍2.5. HC2, hybrid capture 2; hr, high risk.

HPV Testing with the HC2 Test and the cobas HPV Test The HC2 test identified 638 cervical samples as hrHPV positive and 709 cervical samples as hrHPV negative; 13 cervical samples were reported as an indeterminate result (defined as an indeterminate result on repeat testing) (Table 1). These 13 indeterminate cervical samples were reassigned as an HC2-positive result for subsequent analysis. The cobas HPV Test identified 632 cervical samples as hrHPV positive and 728 cervical samples as hrHPV negative; there were no indeterminate results with this test.

Concordance between the HC2 Test and the cobas HPV Test Of the 651 cervical samples reported as hrHPV positive by the HC2 test (including the 13 cervical samples with an indeterminate result), 561 were also reported as hrHPV positive by the cobas HPV Test. Of the 709 cervical samples reported as hrHPV negative by the HC2 test, 638 were also reported as hrHPV negative by the cobas HPV Test. Therefore, both tests performed similarly, with no significant difference in the number of positive and negative samples identified by each test (McNemar test, P ⫽ 0.1588). Concordance between the two HPV tests was 86.57%, with a Cohen’s kappa coefficient of 0.7625 (95% CI: 0.7281 to 0.7969). The percentage of positive agreement between the two tests was 86.18% (95% CI: 83.28 to 88.73) and the percentage of negative agreement between the two tests was 89.99% (95% CI: 87.54 to 92.09).

Resolution of Discordant Samples Using the LA HPV Test Of the 651 cervical samples reported as hrHPV positive by the HC2 test, 90 were reported as hrHPV negative by the cobas HPV Test, and of the 709 cervical samples reported as hrHPV negative by the HC2 test, 71 were reported as hrHPV positive by the cobas HPV Test. LA HPV test results were available for 50 of 71 cervical samples (the remaining 21 samples were from the Milan center, where the LA HPV test was performed only on those samples testing positive by the HC2 test).

68 Mateos Lindemann et al JMD January 2012, Vol. 14, No. 1

Figure 1. Genotype distribution of discordant cervical samples with the LA HPV genotyping test. Significant association between HC2-positive/cobas HPV Test-negative and lrHPV-positive LA HPV test result (␹2 ⫽ 25.26, *P ⬍ 0.0001). Significant association between HC2-negative/cobas HPV Test-positive discordant samples and hrHPV-positive LA HPV test result (␹2 ⫽ 17.11, † P ⬍ 0.0001). HC2, hybrid capture 2; hr, high risk; LA, linear array; lr, low risk.

The results of the LA HPV test for discordant samples are depicted in Figure 1. The LA HPV test detected lrHPV genotypes in 58% (52/90) of HC2-positive/cobas HPV Test-negative samples, and in 14% (7/50) of HC2-negative/cobas HPV Test-positive samples. Distribution of the lrHPV genotypes detected by the LA HPV Test is shown

Table 2.

Low-Risk HPV Genotypes Identified by the LA HPV Genotyping Test in Samples with HC2/cobas HPV Test Discordant Results

lrHPV genotype* 6 42 53 54 55 61 62 67 70 72 73 81 82 84 89 (CP6108) Total number of specimens

HC2⫹/cobas– Total infections (single infections) N ⫽ 73 (N ⫽ 35)†

HC2–/cobas⫹ Total infections (single infections) N ⫽ 9 (N ⫽ 5)‡

3 (0) 12 (7) 19 (15) 2 (0) 3 (1) 5 (1) 5 (1) 6 (4) 6 (1) 1 (0) 3 (3) 2 (1) 1 (0) 2 (1) 3 (0) 52

1 (1) 2 (2) 0 (0) 1 (0) 1 (0) 0 (0) 2 (2) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (0) 7

*HPV11, HPV26, HPV40, HPV64, HPV69, HPV71, HPV83, and IS39 were not detected in these cervical samples. † The total number of lrHPV genotypes present was 73, indicating that some cervical samples had multiple lrHPV genotypes. ‡ The total number of lrHPV genotypes present was nine, indicating that some cervical samples had multiple lrHPV genotypes. HC2, hybrid capture 2; LA, linear array; lr, low risk.

in Table 2. HPV53 was the most commonly found lrHPV genotype in HC2-positive/cobas HPV Test-negative samples, with 37% (19/52) containing this genotype; 15 of 19 cervical samples contained HPV53 as a single infection. Conversely, not one lrHPV genotype was captured in the majority of HC2-negative/cobas HPV Test-positive samples; HPV42 was detected in two samples as was HPV62 and HPV89 (CP6108). HPV6, HPV54, and HPV55 were also detected (some samples with ⱖ1 lrHPV genotype). The LA HPV test detected hrHPV genotypes in 16% (14/86) of HC2-positive/cobas HPV Test-negative samples and in 48% (24/50) of HC2-negative/cobas HPV Test-positive samples. Distribution of the hrHPV genotypes detected by the LA HPV test is shown in Table 3. HPV16 was detected in five HC2-positive/cobas HPV Test-negative samples compared with eight HC2-negative/cobas HPV Test-positive samples. HPV18 was detected in one sample from each of the discordant groups. The LA HPV test was negative for lrHPV and hrHPV types in 27% (24/90) of HC2-positive/cobas HPV Testnegative samples compared with 38% (19/50) of HC2negative/cobas HPV Test-positive samples. Statistical analysis has revealed that the association between HC2-positive/cobas HPV Test-negative samples and lrHPV-positive LA HPV test result was significant (␹2 ⫽ 25.26, P ⬍ 0.0001). The association between HC2-negative/cobas HPV Test-positive samples and hrHPV-positive LA HPV test was also significant (␹2 ⫽ 17.11, P ⬍ 0.0001). There was no association between discordant samples and a HPV-negative LA HPV test result (␹2 ⫽ 1.94, P ⫽ 0.164).

Table 3. High-Risk HPV Genotypes Identified by the LA HPV Genotyping Test in Samples with HC2/cobas HPV Test Discordant Results

hrHPV genotype* 16 18 31 33 35 39 45 51 52 58 59 66§ 68 Total number of samples

HC2⫹/cobas– Total infections (single infections) N ⫽ 14 (N ⫽ 8)†

HC2–/cobas⫹ Total infections (single infections) N ⫽ 30 (N ⫽ 14)‡

5 (3) 1 (1) 0 0 0 1 (1) 0 1 (0) 3 (0) 2 (2) 1 (1) 0 0 14

8 (4) 1 (1) 2 (0) 2 (2) 1 (1) 1 (1) 5 (2) 6 (3) 0 1 (0) 1 (0) 1 (0) 1 (0) 24

*HPV31, HPV33, HPV35, HPV45, HPV56, and HPV68 were not detected in these cervical samples. † The total number of hrHPV genotypes present was 14, indicating that multiple hrHPV genotypes were not found in any of these samples; lrHPV genotypes might have been detected. ‡ The total number of hrHPV genotypes present was 30, indicating that some cervical samples had multiple hrHPV genotypes; lrHPV genotypes might also have been detected. § HC2 does not detect HPV66. HC2, hybrid capture 2; hr, high risk; lr, low risk.

Comparison of cobas HPV Test and HC2 69 JMD January 2012, Vol. 14, No. 1

Discussion The aim of this European multicenter study was to compare the analytical performance of the cobas HPV Test versus the HC2 test for the detection of hrHPV genotypes in cervical specimens from women attending for routine cervical cancer screening or follow-up of Pap smear abnormalities. This study formed part of a larger program designed to compare both the analytical and clinical performance of the prototype cobas HPV Test with the HC2 test and with the LA HPV test for the detection of hrHPV genotypes and the identification of women with high-grade cervical disease. The cobas HPV Test and the HC2 test detected hrHPV genotypes in a similar number of cervical samples [632/ 1360 (46.5%) by the cobas HPV Test and 651/1360 (47.9%) by the HC2 test; McNemar test, P ⫽ 0.1588] and the level of agreement between the two tests was good, with 86.57% concordance and a Cohen’s kappa coefficient of 0.76. The HC2 test detects 13 hrHPV genotypes, whereas the cobas HPV Test detects the same 13 hrHPV genotypes plus HPV66, but this alone does not explain the discordant samples observed in this study because only one HC2-negative/cobas HPV Test-positive sample was found to contain HPV66 by the LA HPV test. Of the 90 discordant samples that were hrHPV positive by the HC2 test but hrHPV negative by the cobas HPV Test, 52 (58%) were found to contain only lrHPV genotypes with the LA HPV test and statistical analysis demonstrated that HC2positive/cobas HPV Test-negative samples were significantly associated with lrHPV-positive LA HPV test results. The distribution of lrHPV genotypes identified in these discordant samples was narrow; more than one third of samples contained HPV53 and more than one fifth contained HPV42. Cross-reactivity of the HC2 test with lrHPV genotypes, including HPV53 and HPV42, has been reported previously16 –20 and probably explains the discordance between the tests for these particular samples. In comparison, of those samples that were hrHPV negative with the HC2 test but hrHPV positive with the cobas HPV Test, only 7 of 50 (14%) were found to contain lrHPV genotypes with the LA HPV test, suggesting that the cobas HPV Test demonstrates less cross-reactivity than the HC2 test. Furthermore, these samples contained a diverse range of HPV genotypes. A previous study has already demonstrated that the cobas HPV Test has excellent concordance with the LA HPV test.21 The clinical implications of cross-reactivity with lrHPV genotypes include the over-investigation of women at low risk of having cervical precancer; indeed, the potential for sending women infected only with lrHPV genotypes for follow-up by colposcopy is acknowledged in the HC2 package insert.13 A study evaluating the impact of HPV53 inclusion in HPV DNA testing as part of a population-based cervical cancer screening cohort, and also for the triage of women with equivocal cytology, found that the inclusion of HPV53 into the hrHPV genotype testing pool would most probably decrease the specificity of the HPV DNA test, given that HPV53 was one of the most common HPV genotypes found in women without cervical precancer or cancer in the two populations studied.22

Of the 90 HC2-positive/cobas HPV Test-negative samples, only 14 (16%) were hrHPV positive with the LA HPV test. Conversely, nearly half (24/50 [48%]) of the HC2negative/cobas HPV Test-positive samples were hrHPV positive with the LA HPV test. Of 24 samples, eight were HPV16 positive and one was HPV18 positive; HPV16 and 18 are the two HPV genotypes detected in approximately 70% of invasive cervical cancers.8,9 Statistical analysis demonstrated a significant association between HC2negative/cobas HPV Test-positive samples and a hrHPV LA HPV test result, suggesting that the HC2 test missed more hrHPV genotypes than the cobas HPV Test. The cobas HPV Test utilizes the human cellular gene ␤-globin as an internal control, and an HPV-negative sample without a positive ␤-globin signal cannot be considered a true negative. The HC2 test does not use an internal control and so a negative result may be due to the absence of any hrHPV genotype at the level of test detection, but may also be due to sample inadequacy and test failure. This could explain some or all of the HC2-negative/cobas HPV Test-positive/LA HPV test-positive samples found in this study and suggests the potential for the HC2 test to miss hrHPV genotypes, including HPV16 and HPV18. The comparison of clinically validated HPV test results can be affected by the reproducibility of these assays. To ensure robust and highly reliable performance of HPV tests in clinical practice, Meijer et al23 suggested that new HPV tests display intralaboratory reproducibility (ie, agreement in test result when the same specimens are tested more than once) and interlaboratory agreement with a lower confidence bound of not ⬍87%. The cobas HPV Test has been shown to meet these stringent reproducibility requirements with an intralaboratory reproducibility of 98.3% (95% CI: 96.8 to 99.1) and an interlaboratory agreement of 94.6% (95% CI: 92.4 to 96.2).24 The discordant results observed in this study could be due to a number of other factors, including differing limits of detection between the two tests. In conclusion, the high degree of concordance observed between the cobas HPV Test and the HC2 test in this European study population indicates that both tests perform similarly for the detection of hrHPV genotypes. For the purpose of this comparison, the hrHPV genotype status of each sample was provided as a pooled result by both tests, but the cobas HPV Test is also able to distinguish individual HPV16 and HPV18 genotypes if required. HPV testing that distinguishes HPV16 and HPV18 from other hrHPV types may identify women at the greatest risk of high-grade cervical disease and cancer.10 Therefore, the additional HPV16/18 genotyping information offered by the cobas HPV Test may further help guide patient management in conjunction with the physician’s assessment of cytology history, other risk factors, and professional guidelines.

Acknowledgments We thank Guillaume Penaranda (Alphabio Laboratory) for data management and statistical analysis and Alison

70 Mateos Lindemann et al JMD January 2012, Vol. 14, No. 1

Blackburn (Health Interactions, United Kingdom) for editorial assistance.

12.

References 1. Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, Schiffman MH, Moreno V, Kurman R, Shah KV: Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. International biological study on cervical cancer (IBSCC) Study Group. J Natl Cancer Inst 1995, 87:796 – 802 2. Schiffman MH, Bauer HM, Hoover RN, Glass AG, Cadell DM, Rush BB, Scott DR, Sherman ME, Kurman RJ, Wacholder S, Stanton CK, Manos MM: Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993, 85:958 –964 3. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999, 189:12–19 4. Cuzick J, Clavel C, Petry KU, Meijer CJ, Hoyer H, Ratnam S, Szarewski A, Birembaut P, Kulasingam S, Sasieni P, Iftner T: Overview of the European and North American studies on HPV testing in primary cervical cancer screening. Int J Cancer 2006, 119:1095–1101 5. Mayrand MH, Duarte-Franco E, Rodrigues I, Walter SD, Hanley J, Ferenczy A, Ratnam S, Coutlee F, Franco EL: Human papillomavirus DNA versus Papanicolaou screening tests for cervical cancer. N Engl J Med 2007, 357:1579 –1588 6. Jordan J, Arbyn M, Martin-Hirsch P, Schenck U, Baldauf JJ, Da Silva D, Anttila A, Nieminen P, Prendiville W: European guidelines for quality assurance in cervical cancer screening: recommendations for clinical management of abnormal cervical cytology, part 1. Cytopathology 2008, 19:342–354 7. Wright TC Jr, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D: 2006 consensus guidelines for the management of women with abnormal cervical cancer screening tests. Am J Obstet Gynecol 2007, 197:346 –355 8. Bosch FX, and de Sanjose S: Chapter 1: Human papillomavirus and cervical cancer — burden and assessment of causality. J Natl Cancer Inst Monogr 2003, 3–13 9. Munoz NF, Bosch X, de, R. Herrero SS, Castellsague X, Shah KV, Snijders PJ, Meijer CJ: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003, 348:518 –527 10. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, Rush BB, Glass AG, Schiffman M: The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice. J Natl Cancer Inst 2005, 97:1072–1079 11. Stoler MH, Wright TC Jr, Sharma A, Apple R, Gutekunst K, Wright TL; ATHENA (Addressing THE Need for Advanced HPV Diagnostics) HPV Study Group: High-risk human papillomavirus testing in women

13. 14. 15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

with ASC-US cytology. Results from the ATHENA HPV Study. Am J Clin Pathol 2011, 135:468 – 475 Wright TC Jr, Stoler MH, Sharma A, Zhang G, Behrens C, Wright TL; ATHENA (Addressing THE Need for Advanced HPV Diagnostics) HPV Study Group: Evaluation of HPV-16 and HPV-18 genotyping for the triage of high-risk HPV⫹ cytology-negative women. Am J Clin Path 2011, 136:578 –586 HC2 high-risk HPV DNA test [package insert]. Gaithersburg, MD: Qiagen; 2008 cobas HPV Test [package insert]. Branchburg, NJ: Roche Molecular Systems Inc.; 2011 Linear array HPV genotyping test [package insert]. Branchburg, NJ: Roche Molecular Systems Inc.; 2008 Castle PE, Schiffman M, Burk RD, Wacholder S, Hildesheim A, Herrero R, Bratti MC, Sherman ME, Lorincz A: Restricted cross-reactivity of hybrid capture 2 with nononcogenic human papillomavirus types. Cancer Epidemiol Biomarkers Prev 2002, 11:1394 –1399 Castle PE, Solomon D, Wheeler CM, Gravitt PE, Wacholder S, Schiffman M: Human papillomavirus genotype specificity of hybrid capture 2. J Clin Microbiol 2008, 46:2595–2604 Poljak M, Kovanda A, Kocjan BJ, Seme K, Jancar N, Vrtacnik-Bokal E: The Abbott real time high risk HPV test: comparative evaluation of analytical specificity and clinical sensitivity for cervical carcinoma and CIN 3 lesions with the hybrid capture 2 HPV DNA test. Acta Dermatovenerol Alp Panonica Adriat 2009, 18:94 –103 Poljak M, Marin IJ, Seme K, Vince A: Hybrid capture II HPV test detects at least 15 human papillomavirus genotypes not included in its current high-risk probe cocktail. J Clin Virol 2002, 25(Suppl 3): S89 –S97 Sandri MT, Lentati P, Benini E, Dell’Orto P, Zorzino L, Carozzi FM, Maisonneuve P, Passerini R, Salvatici M, Casadio C, Boveri S, Sideri M: Comparison of the Digene HC2 assay and the Roche AMPLICOR human papillomavirus (HPV) test for detection of high-risk HPV genotypes in cervical samples. J Clin Microbiol 2006, 44:2141–2146 Castle PE, Sadorra M, Lau T, Aldrich C, Garcia FA, Kornegay J: Evaluation of a prototype real-time PCR assay for carcinogenic human papillomavirus (HPV) detection and simultaneous HPV genotype 16 (HPV16) and HPV18 genotyping. J Clin Microbiol 2009, 47:3344 – 3347 Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hildesheim A, Rodriguez AC, Bratti MC, Wheeler CM, Burk RD: A study of the impact of adding HPV types to cervical cancer screening and triage tests. J Natl Cancer Inst 2005, 97:147–150 Meijer CJ, Berkhoff 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, 124:516 –520 Heideman DAM, Hesselink AT, Berkhof J, van Kemenade F, Melchers WJG, Fransen Daalmeijer N, Verkuijten M, Meijer CJLM, Snijders PJF. Clinical validation of the cobas 4800 HPV Test for cervical screening purposes. J Clin Microbiol 2011, doi:10.1128/JCM.05552–11