6+ polymerase chain reaction reverse line blotting and with commercial type-specific PCR kits

Journal of Clinical Virology 36 (2006) 126–132

Human papillomavirus typing with GP5+/6+ polymerase chain reaction reverse line blotting and with commercial type-specific PCR kits Anna Gillio-Tos a,∗ , Laura De Marco a , Valeria Ghisetti b , Peter J.F. Snijders c , Nereo Segnan d , Guglielmo Ronco d , Franco Merletti a,d a

Unit of Cancer Epidemiology, C.E.R.M.S., University of Turin, via Santena 7, 10126 Turin, Italy Unit of Microbiology, Ospedale S.Giovanni Battista, corso Bramante 88/90, 10126 Turin, Italy Department of Pathology, Vrije Universiteit, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands d CPO Piemonte, via S. Francesco da Paola 31, 10123 Turin, Italy b

c

Received 14 December 2005; received in revised form 22 February 2006; accepted 12 March 2006

Abstract Background: Infection with human papillomavirus (HPV) is a necessary step in the progression to cervical cancer. Many methods for HPV testing are currently available, most developed to detect pools of HPV types. Objectives: To evaluate the HPV typing by molecular methods and to compare commercial kits with an established laboratory method. Study design: Eighty-four cervical samples found to be positive for HPV DNA by GP5+/6+-polymerase chain reaction-enzyme immunoassayreverse line blotting (PCR-EIA-RLB) were re-tested with two commercial methods, INNO-LiPA and Amplisense HPV typing, able to identify the HPV type predicted by PCR-EIA-RLB in 76 and 67 samples, respectively. Results: The INNO-LiPA assay revealed HPV DNA in 75/76 samples (98.7%; 95% CI, 0.93–0.99) that would contain HPV types identifiable by this assay. The Amplisense HPV assay revealed HPV DNA in 58/67 samples (86.6%; 95% CI, 0.76–0.93) containing HPV types detectable by this assay. For samples with a single infection, the unweighted kappa for concordance of HPV typing was 0.87 (95% CI, 0.78–0.97) for PCR-EIA-RLB versus INNO-LiPA, 0.94 (95% CI, 0.87–0.99) for INNO-LiPA versus Amplisense HPV, and 0.82 (95% CI, 0.70–0.94) for PCR-EIA-RLB versus Amplisense HPV typing. PCR-EIA-RLB revealed 12 multiple infections, INNO-LiPA revealed 14, and Amplisense HPV revealed 5. The agreement among tests for samples with multiple infections was lower, giving kappa values of 0.44 (95% CI, 0.18–0.70) for PCR-EIA-RLB versus INNO-LiPA, 0.52 (95% CI, 0.19–0.85) for PCR-EIA-RLB versus Amplisense HPV and 0.43 (95% CI, 0.12–0.74) for INNO-LiPA versus Amplisense HPV. Conclusions: In HPV-positive samples, the agreement among tests for HPV typing was high for single infections but markedly lower for infections with multiple HPV types. © 2006 Elsevier B.V. All rights reserved. Keywords: Papillomavirus (HPV); Typing; Commercial assays; Cervical cancer

1. Introduction Infection with human papillomavirus (HPV) is strongly associated with the development of cervical cancer and is recognized as a necessary step in progression to neoplastic disease (Walboomers et al., 1999). Almost all invasive cervical cancers are HPV-positive. New techniques have become ∗

Corresponding author. Tel.: +39 011 633 6863; fax: +39 011 633 4664. E-mail address: gilliotos [email protected] (A. Gillio-Tos).

1386-6532/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2006.03.002

available for cervical cancer screening by HPV detection, based mainly on molecular detection of viral DNA. Molecular tools allow the identification of many different HPV types, including those considered to confer a high risk for oncogenicity (Munoz et al., 2003). Detection of high-risk HPV DNA might permit selection of women at increased risk for high-grade cervical intraepithelial neoplasia, with implications for prevention and clinical approaches. In the management of women with atypical squamous cells of undetermined significance (ASCUS), HPV DNA analysis can

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

result in fewer women being referred for colposcopy (Manos et al., 1999; ALTS, 2000; Arbyn et al., 2004). Moreover, HPV genotyping can improve understanding of individual risk stratification, prevalence, co-infection, persistence or reinfection, and vaccine development. Many molecular methods for HPV testing are currently available, including commercial [e.g. Hybrid Capture II (Obiso and Lorincz, 2004), Roche Amplicor HPV (Gibson et al., 2004; van Ham et al., 2005)] and laboratory methods [polymerase chain reaction (PCR) with consensus primers targeting the capsid gene L1, such as the MY09/MY11 PCR (Bosch et al., 1995) and GP5+/GP6+-PCR (de Roda Husman et al., 1995; Jacobs et al., 1995) assays]. Most were developed to detect pools of HPV types conferring high and low risks for cancer development. At present, further analysis by expensive, time-consuming, type-specific PCR, hybridisation with type-specific probes, restriction fragment length polymorphism or sequencing is needed for typing positive samples. Reliable, fast, easy assays for HPV genotyping are becoming essential for epidemiological studies that generate many samples. When HPV typing with laboratory assays is not feasible, use of commercial kits is an appropriate alternative if the results are representative. The potential of different assays for detecting HPV types varies because of differences in analytical sensitivity, some cross-hybridisation with related types, and failure to detect specific variants. The performance of commercial kits for typing HPV-positive cervical samples has therefore been compared with that of the established laboratory GP5+/6+-PCR-enzyme immunoassay-reverse line blotting (PCR-EIA-RLB) method (van den Brule et al., 2002). In this study, we evaluated the HPV typing in samples positive for HPV DNA. Two genotyping kits were chosen, INNO-LiPA (Innogenetics) (van Doorn et al., 2002) and Amplisense HPV typing (Nuclear Laser Medicine; Carcheri et al., 2003), both based on PCR and capable of detecting large numbers of high-risk HPV types (15 and 12, respectively). Both tests allow typing of more than 20 samples at the same time, with only a conventional DNA extraction and purification step. The procedures in both tests are easy, do not require high technology equipments to be performed, and allow typing results to be obtained within 2 days.

127

histologically confirmed as cervical intraepithelial neoplasia III and three as cervical intraepithelial neoplasia I; the diagnosis was not histologically confirmed in six samples. Seventy-three samples were normal or had benign cellular changes. The samples of cervical cells were obtained with a Cervex brush and placed in Cytorich® medium (AutoCyte, Inc., Elon College, NC, USA). After preparation of a thin-layer slide, the residual material was pelletted, re-suspended in 0.01 mol/l Tris–HCl (pH 7.5) and frozen at −80 ◦ C for DNA extraction. 2.2. DNA extraction As described previously (Ronco et al., 2005; van den Brule et al., 2002), 100 ␮l of frozen cervical cells were thawed, boiled for 10 min at 100 ◦ C and immediately cooled on ice. The extracted DNA was purified on silica gel (Qiagen, Milan, Italy) and stored at −80 ◦ C. 2.3. GP5+/6+-PCR and genotyping assays Beta-globin gene fragments of 268 base pairs (bp) from the clinical samples were amplified by PCR in order to check DNA quality. The set of primers was chosen according to published sequences (Dong et al., 2002). Samples analysed for HPV DNA by GP5+/6+-PCR-EIARLB as described previously (Ronco et al., 2005), stored at −80 ◦ C and checked for DNA quality by beta-globin PCR amplification, were tested by INNO-LiPA HPV Genotyping (Innogenetics, Ghent, Belgium) and Amplisens HPV typing (Nuclear Laser Medicine, Settala, Italy). High- and low-risk HPV types detectable by the three methods are listed in Table 1. GP5+/6+-PCR-EIA-RLB revealed four HPV types (HPV54, 61, 72, 81) that are not identifiable with the INNOLiPA assay and nine (HPV42, 43, 51, 54, 61, 68, 70, 72, 81) that are not identifiable with the Amplisense HPV typing assay (Table 1). For each amplification reaction, 10 ␮l of extracted DNA were used, and appropriate negative and positive controls were included in each PCR run to monitor the performance of the methods. 2.4. INNO-LiPA assay

2. Materials and methods 2.1. Study population Samples were obtained from women aged 25–70 years who were participating in an organized screening programme for cervical cancer in Turin, Italy (Ronco et al., 2005). Eighty-four DNA samples found to be positive by GP5+/GP6+-PCR were tested a second time with two HPV typing methods. The samples included 81 that have been described previously (Ronco et al., 2005). Eleven samples showed cytological abnormalities, two of which were

The INNO-LiPA HPV genotyping assay is based on amplification of a broad spectrum of HPV genotypes with biotinylated SPF10 primers, targeting a 65-bp fragment in the viral L1 region (Kleter et al., 1998, 1999; van Doorn et al., 2002). PCR was performed according to the manufacturer’s instructions, in a final reaction volume of 50 ␮l, containing 10 ␮l of isolated and purified DNA, 10× GeneAmp PCR buffer II, 2.0 mmol/l MgCl2 1.5 U of AmpliTaq Gold DNA polymerase (Applera, Foster City, CA, USA), 200 ␮mol/l of each deoxynucleoside triphosphate and 10 ␮l of the primer mix included in the kit. The hot-start PCR conditions were

128

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

Table 1 HPV types detectable by the three methods Method

High-risk types

Low-risk types

GP5+/6+-PCR-EIA-RLB

16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68

INNO-LiPA Amplisense HPV typing

16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68 16, 18, 31, 33, 35, 39, 45, 52, 56, 58, 59, 66

6, 11, 26, 34, 40, 42, 43, 44, 53, 54, 55, 57, 61, 70, 71, 72, 73, 81, 82, 83, 84 6, 11, 34, 40, 42, 43, 44, 53, (54a ), 70, 74 6, 11

a

INNO-LiPA does not permit discrimination of type 54 from co-infection with types 31 and 33.

as follows: 9 min at 94 ◦ C; followed by 40 cycles of denaturation for 30 s at 94 ◦ C, annealing for 45 s at 52 ◦ C, extension for 45 s at 72 ◦ C and final extension for 5 min at 72 ◦ C. The presence of amplified products was checked by electrophoresis on a 3% agarose gel stained with ethidium bromide. Ten microlitres of PCR biotinylated products were denaturated and hybridised with type-specific oligonucleotides probes immobilized as parallel lines on nitrocellulose membrane strips. The hybrids were detected with alkaline phosphatase–streptavidin conjugate and substrate (5-bromo-4-chloro-3-indolylphosphate and nitroblue tetrazolium), resulting in a purple precipitate at positive probe lines. After drying, the strips were analysed visually from an interpretation grid supplied in the kit; the presence of a clearly visible line was considered to be a positive reaction. A biotinylated poly(dT) control for conjugate reaction was applied to each strip to ensure good performance of the test and proper alignment of the strips on the interpretation sheet. This version of the assay contains probes for 25 HPV genotypes (Table 1). 2.5. Amplisense HPV typing assay This method reveals 14 HPV genotypes (Table 1). Ten microlitres of DNA were amplified according to the manufacturer’s protocol by multiplex PCR containing a mix of two or four different primer copies (Carcheri et al., 2003). Primers for the beta-globin gene were also added to the mix to detect a fragment of 723 bp as the internal control. PCR was performed in a final reaction volume of 25 ␮l containing 10 ␮l DNA and one of the following four cocktails of type-specific primers: ‘low-risk’ primers for HPV6 and 11 [expected sizes 260 and 425 bp]; ‘16–35’ for HPV16, 31, 33, 35 [expected sizes 325, 520, 227, and 280 bp]; ‘18–59’ for HPV18, 39, 45, 59 [expected sizes 425, 340, 475, and 395 bp]; and ‘52–66’ for HPV52, 56, 58, 66 [expected sizes 360, 325, 240, and 304 bp]. Hot-start PCR conditions were as follows: 5 min at 95 ◦ C; then 40 cycles of denaturation for 40 s at 95 ◦ C, annealing for 40 s at 63 ◦ C, extension for 50 s at 72 ◦ C, and a final extension of 1 min at 72 ◦ C. PCR products were analysed on 3% agarose gel stained with ethidium bromide and visualized by ultraviolet transillumination. HPV genotypes were identified on the basis of the expected size of amplicons, according to the manufacturer’s instructions.

2.6. HPV16 E6 region detection by PCR PCR analysis for the HPV16 E6 region was performed with specific primer pairs according to Yoshinouchi et al. (1999). The PCR conditions were as follows: 5 min at 95 ◦ C; 15 s at 95 ◦ C, 30 s at 55 ◦ C, 60 s at 72 ◦ C for 40 cycles; and a final extension of 5 min at 72 ◦ C. Amplicons (208 bp) were analysed on 2% agarose gel stained with ethidium bromide and visualized by ultraviolet transillumination. 2.7. Statistical analysis Proportions are presented with 95% confidence intervals (95% CIs), estimated by standard methods (Altman et al., 2000). The unweighted kappa statistic was calculated to determine the level of chance-adjusted agreement between pairs of assay methods. In general, kappa values of 0.0–0.2, 0.21–0.40, 0.41–0.60, 0.61–0.80, 0.81–0.99, and 1.0 indicate poor, slight, moderate, substantial, almost excellent, and excellent agreement, respectively (Fleiss, 1981). The twosided McNemar test was used to compare the proportion of multiple infections detected by each method. Statistical analyses were performed with PC-SAS software (version 8.02; SAS Institute, Cary, NC, USA).

3. Results 3.1. PCR positivity HPV detection by PCR was compared with the three methods. GP5+/6+-PCR-EIA-RLB revealed one or more types of HPV in all the samples analysed. Agreement of the two commercial tests with PCR was calculated after excluding the 8 samples with HPV-type infections not detectable or identifiable by INNO-LiPA and the 17 not detectable with Amplisense. Fifty-eight cases were found to be positive for HPV DNA by all methods. The INNO-LiPA assay revealed HPV DNA in 75/76 samples (98.7%; 95% CI, 0.93–0.99) that were predicted by GP5+/6+-PCR-EIA-RLB to contain HPV types identifiable by this assay. Amplisense HPV typing revealed HPV DNA in 58/67 samples (86.6%; 95% CI, 0.76–0.93) in which it could have been detected according to GP5+/6+-PCR-EIA-RLB. All the samples negative for HPV DNA with the kits showed positivity in beta-globin gene PCR analysis.

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

129

Table 2 HPV types detected by PCR-EIA-RLB, INNO-LiPA, and Amplisense HPV typing Sample no.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 a

HPV type

Sample no.

EIA-RLB

INNO-LiPA

Amplisense HPV

6 11 11 16 16 16 16 16 16 16 16,66 16 16 16 16 16 16 16 16 16,31 16 16,45 16 16,31 16 16 18 18 31 31 31 31 31 33 33 33,42,53,58 35 39 39 39 39 42

6,18,53 11 11 16 16 16 16 16,56 31 16 16,66 16 16 16 16 16 16 16 16 31 16 45 16 16,31 16 16 18,6 18 31 31 31 31 16 33 33 33,53 35 39 39 45 39,6 42

0 11 11 16 16 16 16 16 16,31 16 16,66 16 16 16 16 16 16 16 16 16 16 45 16 16,31 16 16 18 0 31 31 31 31 16 33 33 33,56,58 35 39 39 39 0 Nda

HPV16 E6

16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 0 16 16 16 16

16

43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

HPV type EIA-RLB

INNO-LiPA

Amplisense HPV

42 42 42 43 45 45 45 45 45 45 45,56 51,59,66 51 51 52 52 53,31 54 54 56 56 58 58 58,66 61 66 66 66 66 66 66 66,42 68 68 68,52 70 72 72 81 81 81 81,56

42 42 42 43 52 52 45 45,52 45 0 45,56 51 51 51 52 52 53,74 Nd Nd 56 56 58 58 58,66 Nd 66 6,59 66 66 66 66,53 66 68 66 68,52 70 Nd Nd Nd Nd Nd 56

Nd Nd Nd Nd 0 52 45 45 45 0 0 0 Nd Nd 52 52 56 Nd Nd 56 56 58 0 58,66 Nd 66 59 66 66 66 66 66 Nd 66 52 Nd Nd Nd Nd Nd Nd 0

Nd: HPV type not detectable/identifiable by the kit.

3.2. HPV genotyping The HPV type-specific results obtained with each test are shown in Table 2. No infection by HPV type 6 was detected with Amplisense HPV, while one (a single infection, sample 1) was found with PCR-EIA-RLB and four (multiple infections, samples 1, 27, 41, and 69) were found with INNO-LiPA. The Amplisense HPV typing kit confirmed only one of two HPV18 infections detected by PCR-EIA-RLB. INNO-LiPA confirmed both and revealed one more type 18 in a multiple infection. PCR-EIA-RLB detected 19 single infections with HPV16; 17 of these were positive for HPV16 with the INNO-LiPA test, and Amplisense HPV typing detected HPV16 in all 19

samples, as single or multiple infections (Table 2). One sample (No. 9) was found by Amplisense HPV typing to be co-infected with HPV16 and HPV31, whereas INNO-LiPA detected only HPV type 31. INNO-LiPA and Amplisense HPV typing revealed HPV16 in one and two, respectively, of the three cases showing HPV16 in multiple infections by PCR-EIA-RLB. One sample (No. 20) tested positive for HPV16 and HPV31 by PCR-EIA-RLB, whereas INNO-LiPA revealed only HPV31 and Amplisense only HPV16 (Table 2). Another sample (No. 22) was shown to have HPV16 and HPV45 by PCR-EIA-RLB but only HPV45 by the commercial methods. A further check was carried out on all samples that tested positive for HPV16 in at least one assay, by PCR detection of

130

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

the HPV16 E6 region with a specific primer set according to published sequences (Yoshinouchi et al., 1999). In this assay, type 16 positivity was confirmed in all samples in which a single HPV16 infection was detected in all the assays. In addition, HPV16 E6 was found in two samples (Nos. 9 and 20) that were HPV16-negative in INNO-LiPA but positive in the other two assays. Specific PCR for HPV16 revealed HPV in one sample (No. 33) which was found to be negative with PCR-EIA-RLB but positive with the commercial assays. One case of HPV16 positivity (No. 22) detected only by PCREIA-RLB could not be confirmed by E6 type-specific PCR. In two samples (Nos. 36 and 59), Amplisense HPV typing revealed HPV56, whereas PCR-EIA-RLB and INNO-LiPA detected HPV53. In two other samples (Nos. 48 and 76), the HPV type in the reference method differed from that revealed by both commercial kits. 3.2.1. Agreement on typing of single infections PCR-EIA-RLB revealed a total of 72 single infections, INNO-LIPA revealed 61 and Amplisense HPV typing revealed 53. Agreement on typing was strongly conditional on HPV DNA positivity in paired methods. For HPV-positive samples with single infections detected in two tests, there was agreement between PCR-EIA-RLB and INNO-LiPA for 50/56 samples (89.3%; 95% CI, 78.1–96.0%), between PCREIA-RLB and Amplisense HPV typing for 44/48 samples (91.7%; 95% CI, 80.0–97.7%) and between INNO-LiPA and Amplisense HPV for 44/46 samples (95.7%; 95% CI, 85.2–99.5%). The corresponding unweighted kappa values were 0.87 (95% CI, 0.78–0.97), 0.94 (95% CI, 0.87–0.99) and 0.82 (95% CI, 0.70–0.94), respectively. Forty-three samples were HPV DNA-positive for single infection with all three methods. The type was the same by all three methods in 39 cases (90.7%; 95% CI, 77.9–97.4%). 3.2.2. Presence of multiple infections Only samples positive for HPV DNA were included in each comparison (Table 3). PCR-EIA-RLB revealed 12 multiple infections, INNO-LiPA 14 and Amplisense HPV typing 5. The proportion of samples showing multiple infections was significantly higher with INNO-LiPA than with Amplisense HPV typing (p = 0.0339), of borderline significance for PCR-EIA-RLB versus Amplisense HPV typing (p = 0.1025) and not significant for PCR-EIA-RLB versus INNO-LiPA (p = 0.5637) (two-sided McNemar test). The agreement of the tests in detecting the presence of multiple infections was lower than for single infections. The kappa values were 0.44 (95% CI, 0.18–0.70) for PCR-EIARLB versus INNO-LiPA, 0.52 (95% CI, 0.19–0.85) for PCREIA-RLB versus Amplisense HPV typing and 0.43 (95% CI, 0.12–0.74) for INNO-LiPA versus Amplisense HPV typing. In samples with multiple infections, the HPV types were the same in 5/7 samples classified as having multiple HPVs by both PCR-EIA-RLB and INNO-LiPA, in 3/4 classified as having multiple HPVs by both PCR-EIA-RLB and Amplisense HPV typing and in 3/4 classified as having mul-

Table 3 Presence of single and multiple infections: agreement among the three methods Method 1a PCR-EIA-RLB

INNO-LiPA

Single Multiple

56 5

7 7

63 12

Total

61

14

75

Single

Multiple

Total

Single Multiple

48 5

1 4

49 9

Ttotal

53

5

58

Single

Multiple

Total

Method 2b PCR-EIA-RLB

Amplisense HPV

Method 3c INNO-LiPA

Amplisense HPV Single

Multiple

Total

Single Multiple

46 7

1 4

47 11

Total

53

5

58

Only cases positive for HPV DNA by two methods were considered in each sub-table. Among cases with no HPV DNA detectable by Amplisense HPV, one showed multiple infections with both PCR-EIA-RLB and INNO-LiPA, two showed multiple infections with PCR-EIA-RLB only and two showed multiple infections with INNO-LiPA only. a Raw agreement 0.84. b Raw agreement 0.90. c Raw agreement 0.86.

tiple infections by both INNO-LiPA and Amplisense HPV typing. Agreement between all three assays on the HPV types in samples with multiple infections was found in only three samples.

4. Discussion The concordance among the tests was high for the detection of single infections, and comparisons of pairs of methods indicated accurate typing. The kappa values showed almost excellent agreement for the comparisons of INNO-LiPA with PCR-EIA-RLB (0.87) and INNO-LiPA with Amplisense HPV typing (0.94), and substantial agreement for the comparison of PCR-EIA-RLB with Amplisense HPV typing (0.82). The agreement between INNO-LiPA and Amplisense HPV typing was very high, but the analysis excluded HPV types not detectable by the two kits. The Amplisense HPV typing kit includes an internal control for DNA quality and competence (beta-globin gene detection) in the same PCR run, which is not provided in the INNO-LiPA kit. The reliability of the INNO-LiPA assay is, however, based on the short (65 bp) amplicons obtained

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

with the LiPA primers, which allow HPV detection even in slightly fragmented DNA (Kleter et al., 1999) in which few amplicons of more than 200 bp are found. When interpreting PCR results for HPV detection, we included only samples positive in GP5+/6+-PCR-EIARLB, the efficiency of which was implicitly set at 100%. Amplisense HPV typing showed a lower positivity rate (86.6%) than the INNO-LiPA assay (98.7%), which could be due to the larger size (227–520 bp) of the HPV amplicons in the Amplisense test and to an imperfectly preserved target DNA. Nevertheless, as all the samples tested in this study were positive for the beta-globin gene, a more likely explanation is lower sensitivity for the detection of PCR products by gel electrophoresis without a subsequent hybridisation step. In some samples, the methods revealed different viral types. In two samples, HPV56 was recognized by the Amplisense HPV typing system and type 53 by the other two tests. As Amplisense HPV typing does not include a specific primer mix for type 53, a cross-hybridisation event might have occurred with the ‘52–66’ primer mix of this kit. Cross-hybridisation events could also explain the differences in HPV typing of samples 48 and 76 by the two commercial kits and PCR-EIA-RLB. Some authors have reported that the INNO-LiPA system detects more multiple infections than other PCR assays (Perrons et al., 2002; van Doorn et al., 2002). They have suggested that the analytical sensitivity of LiPA is generally higher and that low copy numbers of a co-existing HPV type would be better detected by LiPA than by other assays. In this study, we found 14 multiple infections with INNO-LiPA, five with Amplisense HPV typing and 12 with GP5+/6+-PCREIA-RLB. Some were reproducible with one of the other systems: five were confirmed with PCR-EIA-RLB and three with Amplisense HPV typing, whereas nine were not confirmed in either of the other two tests. Similarly, six multiple infections detected by PCR-EIA-RLB were not confirmed with either of the other two tests. Only three multiple infections were detected with all three methods, indicating poor reproducibility, and analysis of paired methods gave moderate kappa values (0.4–0.5). It is noteworthy, however, that in all cases of multiple types, detection of at least one type was reproduced with at least one of the other methods. Differential sensitivity of the tests for different HPV types, subtypes or variants, lack of specificity or a combination thereof might be implicated in discordant multiple typing. As we studied only samples that tested positive for HPV DNA by PCR-EIA-RLB, the design did not allow a correct estimate of agreement on the presence of HPV DNA. It did allow an unbiased estimate of agreement on HPV typing, computed conditionally on the detection of HPV DNA in pairs of samples. This agreement was high for samples with single infection but limited for those with multiple infections. These results indicate that complementary assays should be used to avoid missing potential high-risk HPV infections. Genotyping methods might become important for improving understanding of progression to cervical cancer, as individual

131

risk stratification, prevalence, persistence and new infections can be detected by genotyping, as can correlations between viral burden and grade or risk of tumour progression. HPV vaccine development depends on typing (Crum and Rivera, 2003), and many studies are being conducted on variants of high-risk HPV types (Zehbe et al., 1998; Burk et al., 2003). Assays with better sensitivity for different viral loads of coexisting types might be useful for investigating the hypothesis that infection of the same cervical sample with multiple HPV types is associated with an increased risk for cervical intraepithelial neoplasia (Morrison et al., 1991; Becker et al., 1994; van der Graaf et al., 2002).

Acknowledgement This study was partly supported by the Special Project ‘Oncology’, Compagnia di San Paolo FIRMS.

References Arbyn M, Buntinx F, Van Ranst M, Paraskevaidis E, Martin-Hirsch P, Dillner J. Virologic versus cytologic triage of women with equivocal Pap smears: a meta-analysis of the accuracy to detect high-grade intraepithelial neoplasia. J Natl Cancer Inst 2004;96:280–93. Altman DG, Mschin D, Bryant TN, Gardner MJ. Statistics with confidence. 2nd ed. Bristol: JW Arrowsmith Ltd.; 2000. ALTS Group. Human papillomavirus testing for triage of women with cytologic evidence of low-grade squamous intraepithelial lesions: baseline data from a randomized trial. J Natl Cancer Inst 2000;92:397–402. Becker TM, Wheeler CM, McGough NS, Parmenter CA, Jordan SW, Stidley CA, et al. Sexually transmitted diseases and other risk factors for cervical dysplasia among south western Hispanic and non-Hispanic white women. JAMA 1994;271:1181–8. Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, et al. 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. Burk RD, Terai M, Gravitt PE, Brinton LA, Kurman RJ, Barnes WA, et al. Distribution of human papillomavirus type 16 and 18 variants in squamous cell carcinomas and adenocarcinomas of the cervix. Cancer Res 2003;63:7215–30. Carcheri M, Lacitignola G, Riva R, Capuzzo R, Mentasti M, Ventura A, et al. HPV-DNA research and viral genotyping in diagnosis and prevention of uterine cervix carcinoma. Microbiol Med 2003;18:39–42. Crum CP, Rivera MN. Vaccines for cervical cancer. Cancer J 2003;9:368–76. Dong SM, Pai SI, Rha SH, Hildesheim A, Kurman RJ, Schwartz PE, et al. Detection and quantitation of human papillomavirus DNA in the plasma of patients with cervical carcinoma. Cancer Epidemiol Biomarkers Prev 2002;11:3–6. de Roda Husman AM, Walboomers JMM, van der Brule AJC, Meijer CJLM, Snijders PJF. The use of general primers GP5 and GP6 elongated at their 3 ends with adjacent highly conserved sequences improves human papillomavirus detection by polymerase chain reaction. J Gen Virol 1995;76:1057–62. Fleiss JL. Statistical methods for rates and proportion. 2nd ed. New York: John Wiley & Sons Inc.; 1981. Gibson J, Vinson D, Flynn D, Garcia R, Jacquin D, Kohler M, et al. The Roche amplicor HPV test as a primary screen for HPV in cervical specimens collected in ThinPrep and SurePath liquid cytology

132

A. Gillio-Tos et al. / Journal of Clinical Virology 36 (2006) 126–132

media. In: 21st International Papillomavirus Conference and Clinical Workshop; 2004. p. 261. Jacobs MV, de Roda Husman AM, van den Brule AJC, Snijders PJF, Meijer CJLM, Walboomers JWW. Group specific differentiation between high- and low-risk human papillomavirus genotypes by general primer-mediated PCR and two cocktails of oligonucleotide probes. J Clin Microbiol 1995;33:901–5. Kleter B, van Doorn LJ, ter Schegget J, Schrawen L, van Krimpen C, Burger MP, et al. A novel short-fragment PCR assay for highly sensitive broad-spectrum detection of anogenital human papillomaviruses. Am J Pathol 1998;153:1731–9. Kleter B, van Doorn LJ, Schrawen L, Molijn A, Sastrowijoto S, ter Schegget J, et al. Development and clinical evaluation of a highly sensitive PCR-reverse hybridisation line probe assay for detection and identification of anogenital human papillomavirus. J Clin Microbiol 1999;37:2508–17. Manos MM, Kinney WK, Hurley LB, Sherman ME, Shieh-Ngai J, Sherman RJ, et al. Identifying women with cervical neoplasia: using papillomavirus DNA testing for equivocal Papanicolaou results. JAMA 1999;281:1605–10. Morrison EA, Ho GY, Vermund SH, Goldberg GL, Kadish AS, Kelley KF, et al. Human papillomavirus infection and other risk factors for cervical neoplasia: a case–control study. Int J Cancer 1991;49:6–13. Munoz N, Bosch FX, De Sanjose S, Herrero R, Castellsague X, Shah KV, et al., International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–27. Obiso R, Lorincz A. Digene corporation. Pharmacogenetics 2004;5: 129–32. Perrons C, Kleter B, Jelley R, Jalal H, Quint W, Tedder R. Detection and genotyping of human papillomavirus DNA by SPF10 and MY09/11

primers in cervical cells taken from women attending a colposcopy clinic. J Med Virol 2002;67:246–52. Ronco G, Ghisetti V, Snijders PJF, Gillio Tos A, Merletti F, Segnan N, et al. Prevalence of papillomavirus infection in women in Turin, Italy. Eur J Cancer 2005;41:297–305. van den Brule AJ, Pol R, Fransen-Daalmeijer N, Schouls LM, Meijer CJ, Snijders PJ. GP5+/6+-PCR followed by reverse line blot analysis enabled rapid and high throughput identification of human papillomavirus genotypes. J Clin Microbiol 2002;40:779–87. van Doorn LJ, Quint W, Kleter B, Molijn A, Colau B, Martin MT, et al. Genotyping of human papillomavirus in liquid cytology cervical specimens by the PGMY line blot assay and the SPF10 line probe assay. J Clin Microbiol 2002;40:979–83. van der Graaf Y, Molijn A, Doornewaard H, Quint W, van Doorn LJ, van den Tweel J. Human papillomavirus and long term risk of cervical neoplasia. Am J Epidemiol 2002;156:158–64. van Ham MAPC, Bakkers JMJE, Harbers GK, Quint WGV, Massuger LFAG, Melchers WJG. Comparison of two commercial assays for detection of human papillomavirus (HPV) in cervical scrape specimen: validation of the Roche AMPLICOR HPV test as a means to screen for HPV genotypes associated with higher risk of cervical disorders. J Clin Microbiol 2005;43:2662–7. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–9. Yoshinouchi M, Iiongo A, Nakamura K, Kodama J, Itoh S, Sakai H, et al. Analysis by multiplex PCR of the physical status of human papillomavirus type 16 DNA in cervical cancer. J Clin Microbiol 1999;37:3514–7. Zehbe I, Wilander E, Delius H, Tommasino M. Human papillomavirus 16 E6 variants are more prevalent in invasive cervical carcinoma than the prototype. Cancer Res 1998;58:829–33.