Quantitative multiplex PCR assay for the detection of the seven clinically most relevant high-risk HPV types

Journal of Clinical Virology 44 (2009) 302–307

Contents lists available at ScienceDirect

Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv

Quantitative multiplex PCR assay for the detection of the seven clinically most relevant high-risk HPV types Martina Schmitz a , Cornelia Scheungraber a , Jörg Herrmann a , Karin Teller a , Mieczyslaw Gajda b , Ingo B. Runnebaum a , Matthias Dürst a,∗ a b

Klinik für Geburtshilfe und Frauenheilkunde, Universitätsklinikum Jena, Germany Institut für Pathologie, Universitätsklinikum Jena, Germany

a r t i c l e

i n f o

Article history: Received 10 October 2008 Received in revised form 29 December 2008 Accepted 14 January 2009 Keywords: Multiplex real-time PCR HPV Genotyping Viral load

a b s t r a c t Background: High-risk HPV DNA detection has become a valuable tool for the triage of borderline, questionable and abnormal cytologic findings in cervical carcinoma screening programs. This knowledge is largely based on studies which could only discriminate between low-risk (LR-) and high-risk (HR-) HPV groups. However, it is becoming increasingly clear that HPV genotyping may allow further risk stratification and may offer different treatment options in the future. Objectives: To establish a fast and cost-effective system not only for genotyping but also for quantification of viral DNA. Study design: Development and validation of a 5 exonuclease fluorescent probe multiplex real-time PCR assay (TaqMan format) for the detection and quantification of the 7 most frequent HR-HPV types (16, 18, 31, 33, 45, 52 and 58) which account for over 87% of cervical carcinomas world-wide. Two PCR reactions are required to detect the designated HPV types. Results: Experiments with plasmid constructs of all 18 HR-HPV DNA showed that the multiplex real-time PCR assay was highly sensitive and specific. Evaluation of DNA extracted from archived cell pellets of cervical scrapes by the multiplex assay and the GP5+/6+-EIA showed identical genotyping for 234 of 261 (89.6%) samples and an almost perfect agreement when considering all typing results (kappa 0.901). Viral load did not correlate with disease progression within the CIN spectrum but significant differences were evident when comparing all CIN with the group lacking CIN (p = 0.0028) or with the cancer group (p = 0.0001). Conclusion: Our multiplex assay will be useful to address questions related to viral persistence at the genotype level, the kinetics of viral load and disease recurrence. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Persistent infection with high-risk (HR-) human papillomaviruses (HPV) is known to be a necessary cause for developing high grade cervical lesions and cervical carcinomas.1 Altogether 18 HR-HPV types infect the anogenital mucosa. Of these HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59 are defined as carcinogenic and HPV 26, 53, 66, 68, 73 and 82 as probably carcinogenic.2,3 Numerous studies have shown that testing for HR-HPV, either as an independent screening tool or as an adjunct to Pap smear cytology

Abbreviations: HPV, human papillomavirus; HR, high risk; LR, low risk; PCR, polymerase chain reaction; CIN, cervical intraepithelial neoplasia. ∗ Corresponding author at: Abteilung Frauenheilkunde, Klinik für Geburtshilfe und Frauenheilkunde, Universitätsklinikum Jena, Bachstr. 18, 07743 Jena, Germany. Tel.: +49 3641 933720; fax: +49 3641 934272. E-mail address: [email protected] (M. Dürst). 1386-6532/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2009.01.006

significantly improves the efficacy of screening programs particularly for women above 30 years of age.4,5 To date most of the assays performed are not at the genotype level but rather detect the entire HR-HPV group or a subset of the HR-HPV group. Various techniques are available: liquid hybridization followed by signal amplification (Hybrid capture II)6 and target amplification by a variety of PCRbased assays (AMPLICOR,7 GP5+/6+;8 MY09/119 ). However, it is becoming increasingly clear that HPV genotyping may allow further risk stratification. HPV 16 and 18 have a higher oncogenic potential in comparison to the other HR-HPV types.10,11 This may open the possibility to triage women for treatment of dysplasia on the basis of the HPV type involved. Moreover, HPV genotyping for disease monitoring after conisation permits distinction between re-infection and sustained disease of different HPV types.12 Another important argument for genotyping is that efficient assays for HPV genotyping are required to analyse vaccine efficacy. Thus far only a handful of commercial assays for HPV genotyping are available. The best known are the Roche Diagnostics Linear array (LA) HPV genotyping assay13

M. Schmitz et al. / Journal of Clinical Virology 44 (2009) 302–307

and the Innogenetics INNO-LiPA line probe assay.14 Both tests are based on reverse line blot technology (RLB) in which the PCR product is hybridized to HPV type-specific targets on a membrane. The main drawback of these tests and other non-commercial reverse line blot assays15 is that they are expensive and time-consuming. The costs for commercial tests which do not include DNA extraction range between D 20 and D 40. In addition, the read-out is visual, in borderline cases subjective and it does not allow quantification. Other assays for HPV genotyping make use of the real-time PCR technology. Lindh and colleagues have established a real-time PCR which is able to detect and genotype 12 HR-HPV and 2 LR-HPV types.16 However, 14 individual reactions are required to genotype one sample. Another recently published HPV genotype assay is based on one-step multiplex PCR which allows the simultaneous detection of 16 HR- and LR-HPV types. The amplicons for each type are of unique size and need to be visually identified after capillary electrophoresis.17 However, quantification of the viral DNA in the sample is not possible by this technique. Although theoretically feasible a robust multiplex real-time PCR assay for the individual detection of all HR-HPV types has so far not been reported on. The aim of this study was to develop a fast, reliable and cost-effective assay not only for HPV genotyping and but also for HPV DNA quantification. It is designed as a multiplex real-time PCR assay in the TaqMan format18 and costs less than D 2 excluding DNA extraction. We focused on 7 of the clinically most relevant HR-HPV types (HPV 16, 18, 31, 33, 45, 52 and 58) which cause 87.4% of all carcinomas world-wide.19 Support for selected genotyping is also provided by a prospective population-based screening study and a triage study for ASCUS which showed that the addition of HPV types other than HPV 16, 18, 31, 33, 35, 45, 52 and 58 to an HPV test decreased specificity more than it increased sensitivity.20 2. Materials and methods 2.1. HPV plasmid DNA To determine the sensitivity, specificity and reproducibility of the quantitative multiplex PCR assay cloned plasmid DNA of all 18 HR-HPV types and the four most frequent LR-HPV types (HPV 6, 11, 42 and 44) were used. HPV 6, 11, 18 and 51 were provided by E.-M. deVilliers, Deutsches Krebsforschungszentrum, Heidelberg, Germany; HPV 31, 35, 44 and 56 by A. Lörincz, Digene Corporation Gaithersburg, MD; HPV 33, 42, 66 and 68 by G. Orth, Institut Pasteur, Paris, France; HPV 58 by T. Matsukura, National Institute of Health, Tokyo, Japan. HPV 16 was cloned by Dürst and colleagues formerly at the Deutsches Krebsforschungszentrum, Heidelberg, Germany.21 The LCR/E6/E7 region of HPV 26, 39, 45, 52, 53, 59, 73 and 82 each was cloned from cervical biopsies genotyped in our laboratory in Jena, Germany. Cloning was performed in pCRII or pCR4TOPO (Invitrogen). HPV type was confirmed by sequence analysis. 2.2. Cell lines SiHa (2 genome copies of HPV 16), C4-1 (single copy of HPV 18) and C33A (HPV negative) were used for control purposes. The cell lines were grown in Dulbecco’s modified Eagle medium supplemented with 10% fetal calf serum, 100 units/ml penicillin and 100 ␮g/ml streptomycin in a humidified incubator at 37 ◦ C with 5% CO2 . 2.3. Clinical samples HPV DNA testing is done routinely for all patients who attend our dysplasia unit. For each patient cervical cells are scraped from the ecto- and endocervix with a cytobrush. Material from the brush

303

is used to prepare a conventional Pap smear. The remaining cells are then collected in 3 ml PBS (phosphate-buffered saline, pH 7.4). Samples are vortexed, divided into two reaction tubes and centrifuged at 7000 × g for 5 min. One of the cell pellets is stored at −80 ◦ C for later studies, the second pellet is resuspended in 500 ␮l Tris-buffer (10 mM Tris–HCl pH 8.3, 50 mM KCl) and incubated at 95 ◦ C for 10 min. Ten microliters of this crude extract is then used to screen for HPV DNA by GP5+/6+-EIA according to the protocol of Jacobs et al.22 However, crude cell extracts are not ideal for obtaining consistent results by multiplex real-time PCR. Therefore archived cell pellets of all patients (17–78 years with a median of 34 years) who tested HR-HPV positive by GP5+/6+-EIA (without genotyping) under routine conditions between January 2006 and September 2006 were taken for DNA extraction and re-evaluation by both assays. Moreover, a random number of HPV-negative cases were also included. Re-evaluation of these samples by GP5+/6+-EIA showed 187 samples to be HR-HPV positive; 74 samples were HPV negative. All patients in this retrospective study had colposcopy done at the time the cervical scrapes were taken. If indicated by colposcopy biopsies were taken (n = 118 patients). 2.4. DNA extraction DNA was extracted from archived cell pellets (see above) or cell lines by use of the QIAmp DNA Mini Kit from Qiagen according to the manufactures recommendation. Briefly, the cell pellets were resuspended in 500 ␮l Tris-buffer (10 mM Tris–HCl pH 8.3, 50 ml KCl). 100 ␮l were used for DNA extraction. Instead of 20 ␮l of Proteinase K, 35 ␮l were used and incubation was limited to 30 min at 56 ◦ C with intermitted vortexing every 5 min. The DNA was eluted from the columns with 80 ␮l water. 2.5. Multiplex real-time PCR PCR primers and corresponding TaqMan probes were designed for HPV 16, 18, 31, 33, 45, 52 and 58. To ensure optimum multiplexing the Tm’s of all primers were required to be similar. Moreover, the Tm’s of the probes were about 10 ◦ C higher than that of the primers. For target amplification the LCR/E6/E7 regions of HPV genome were chosen. All primers and probes were verified by BLAST analysis. For closely related HPV types, probes showed at least 9 mismatches. To control for DNA quality and to determine relative viral copy numbers in the samples ␤-globin (HBB) was amplified in one of the multiplex reactions. All data related to the primers and probes are shown in Table 1. PCR was performed in an ABI 7300 cycler (Applied Biosystems) which can detect 4 different fluorescent dyes. Thus, two PCR reactions are required to detect the designated HPV types: one to detect HPV 16, 18, 31 and 45 and a second to detect HPV 33, 52, 58 and ␤-globin. The PCR reaction comprises 9.5 ␮l DNA (up to 50 ng using the QIAamp kit), 12.5 ␮l Platinum Quantitative PCR SuperMix-UDG (Invitrogen), 10 pmol of each primer and 1–5 pmol of each probe in an end volume of 25 ␮l. To prevent amplification of carry-over PCR products UDG (Uracil-DNA-Glycosylase) digestion is performed for 2 min at 50 ◦ C first. This is followed by an initial denaturation step at 94 ◦ C for 10 min which also inactivates UDG but activates the DNA polymerase. Forty-five PCR cycles are then run at 94 ◦ C for 15 s, 50 ◦ C for 20 s and 60 ◦ C for 40 s each. 2.6. GP5+/6+-EIA The assay was performed according to the protocol of Jacobs et al.22 In brief, the PCR reaction comprised 10 ␮l DNA (up to 50 ng), 1× reaction buffer, 3.5 mM MgCl2 , 200 ␮M of each dNTP, 1 U of thermostable DNA polymerase (AmpliTaq, Roche) and 50 pmol each of the GP5+ and GP6+-bio primers in an end volume of 50 ␮l. 37 dif-

304

M. Schmitz et al. / Journal of Clinical Virology 44 (2009) 302–307

Table 1 Primer and probes for multiplex real-time PCR. Reaction tube

1

2

a b

Primer namea

Sequence primer

Sequence probeb

HPV 16 for HPV 16 rev HPV 18 for HPV 18 rev HPV 31 for HPV 31 rev HPV 45 for HPV 45 rev

5 5 5 5 5 5 5 5

gaa ccg aaa ccg gtt agt ata a 3 atg tat agt tgt ttg cag ctc tgt 3 gga ccg aaa acg gtg tat ata a 3 cag tga agt gtt cag ttc ggt gaa ccg aaa acg gtt ggt ata ta 3 atc gta ggg tat ttc caa tgc 3 cag tgt aat aca tgt tgt gac cag 3 aca gga tct aat tca ttc tga ggt 3

5 Fam – cat ttt atg cac caa aag aga act gca atg ttt c – BHQ1 3

HPV 33 for HPV 33 rev HPV 52 for HPV 52 rev HPV 58 for HPV 58 rev bG for bG rev

5 5 5 5 5 5 5 5

gca tga ttt gtg cca agc at 3 ctc aga tcg ttg caa agg ttt 3 aaa cgg tca gac cga aac c 3 cag cac ctc aca caa ttc gt 3 cac gga cat tgc atg att tgt 3 tca gat cgc tgc aaa gtc ttt 3 aca caa ctg tgt tca cta gc 3 caa ctt cat cca cgt tca cc 3

5 Tamra – act ata cac aac att gaa cta cag tgc gtg gaa tgc – BHQ2 3

5 Tamra – atg tga gaa aca cac cac aat act atg gcg cg – BHQ2 3 5 Y.Y. – cat agt att ttg tgc aaa cct aca gac gcc atg t – BHQ1 3 5 Rox – caa gaa aga ctt cgc aga cgt agg gaa aca c – BHQ2 3



Probe finalconcentration

Product size

40 nM

128 bp

200 nM

124 bp

40 nM

144 bp

40 nM

88 bp

200 nM

148 bp

40 nM

116 bp

40 nM

97 bp

40 nM

110 bp



5 Rox – aac aca gtg tag cta acg cac ggc cat gt – BHQ2 3 5 Fam – ttt caa ttc gat ttc atg cac – MGBNFQ 3

5 Y.Y. – tca aac aga cac cat ggt gca tct gac tcc – BHQ1 3

for: forward, rev: reverse. BHQ: Black Hole Quencher; Y.Y.: Yakima Yellow; MGBNFQ: Minor Groove Binder. Non Fluorescence Quencher; Fam: 6-Fam.

ferent HPV types can be amplified under these conditions. The PCR products were detected by an enzyme-immunoassay (EIA) which can detect HR- or LR-HPV groups (digoxygenin probes as a cocktail). For HPV genotyping individual probes were used. 2.7. Statistical analysis Cohen’s kappa values were calculated to assess the degree of agreement between two assays. Kappa values of 0–0.2, 0.21–0.4, 0.41–0.6, 0.61–0.8, 0.81–0.99, and 1.0 indicate poor, slight, fair, moderate, substantial, almost perfect and perfect agreement, respectively. 3. Results 3.1. Sensitivity and specificity of the multiplex real-time PCR assay Our multiplex real-time PCR assay is designed for the individual detection of the 7 most prevalent HR-HPV types in cervical cancer. Two reactions are required: One detects and differentiates between HPV 16, 18, 31 and 45, whereas the second detects and differentiates between HPV 33, 52, 58 and ␤-globin. Using plasmid dilutions of each HPV type the dynamic range was determined to be at least 6-orders of magnitude. A correlation coefficient of ≥0.99 was achieved for HPV 16, 31 and 58 from 106 to 10 genome copies and for HPV 18, 33, 45 and 52 from 106 to 100 genome copies clearly indicating that the PCR conditions allow for reliable quantification (Fig. 1). The lowest sensitivity was obtained for HPV 18

and 33. This is due to the fluorophore Tamra which has the weakest emission of all 4 fluorophores used in this assay. A direct comparison of the HPV 18 probe labelled either with 5 -Fam/BHQ1-3 or 5 -Tamra/BHQ2-3 showed differences of 1.5 to ∼3 cycles in a plasmid dilution series ranging from 0.1 fg to 10 pg (10–106 genome copies) (Table 2). In order to determine the specificity of the multiplex real-time PCR assay plasmid dilutions of all 18 HR-HPV types and 4 LR-HPV types were tested in individual reactions. Primer pairs and probes for each of the 7 HR-HPV types were highly specific and did not show cross reactivity even at target concentrations of up to 106 genomes (data not shown). Double or multiple HPV infections of the cervical epithelium are common. However, the viral load of the coexisting HPV types may differ considerably. We therefore determined the threshold levels of all 7 HR-HPV types in a situation in which these types are diluted in the presence of 10 pg (106 genome copies) of a another HPV type amplified in the same PCR reaction. With exception of HPV 18, all HPV types could be detected at concentrations 100–1000-fold below that of the predominant type (Table 3). However, it should be pointed out that matrix effects characteristic of clinical samples were not accounted for in the above experiments. Finally we tested the multiplex PCR assay with DNA isolated from cervical carcinoma cell lines. 10 ng to 100 pg (corresponding to 103 to 10 cells) of genomic DNA from C4-1 (containing one copy of HPV 18 per cell) and SiHa (containing 2 copies of HPV 16 per cell) could be detected in a background of 10 ng (corresponding to 103 cells) of genomic C33A DNA. The correlation coefficient for this dilution range was 0.918 for HPV 18 and 0.974 for HPV 16, respectively.

Table 2 Effect of different dyes on sensitivity.

Fig. 1. Multiplex real-time PCR for a dilution range of HPV 16 plasmid DNA. Cycle threshold-values (Ct values) are blotted against log10 of target DNA. A correlation coefficient (R2 ) ≥0.99 was obtained.

HPV 18 plasmid

Ct value Tamra dye

Ct value Fam dye

10 pg 1 pg 100 fg 10 fg 1 fg 0.1 fg

23.6 25.9 29.1 32.4 34.0 38.4

21.1 23.6 26.2 29.7 32.5 36.2

M. Schmitz et al. / Journal of Clinical Virology 44 (2009) 302–307

305

Table 3 Detection levels of different HPV types diluted in a background of another HPV type, performed with plasmid DNA.

HPV 16 HPV 18 HPV 31 HPV 45 HPV 33 HPV 52 HPV 58

Target DNA amount

Detected target amount

Dilution factor detectable

Background DNA

Reaction tube

10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg

10 pg to 100 fg 10 pg to 1 pg 10 pg to 100 fg 10 pg to 100 fg 10 pg to 10 fg 10 pg to 10 fg 10 pg to 10 fg

100 10 100 100 1000 1000 1000

10 pg HPV 45 10 pg HPV 16 10 pg HPV 16 10 pg HPV 16 10 pg HPV 52 10 pg HPV 58 10 pg HPV 52

1 1 1 1 2 2 2

Table 4 HPV genotyping of 261 cervical scrapes by GP5+/6+-EIA followed by multiplex real-time PCR. Typing reaction

HPV typea

Total

16

18

31

33

45

52

58

Both GP5+/6+-EIA Multiplex assay

104 4 5

5 4 –

23 4 1

4 2 –

9 – –

2 3 2

4 4 1

151 21 9

Total

113

9

28

6

9

7

9

181

a

In 68 cases additional HR-HPV types were detected by GP5+/6+-EIA.

3.2. Validation of the multiplex real-time PCR assay by comparison with the GP5+/6+-EIA system To assess the performance of the multiplex real-time PCR assay in a clinical setting we examined 261 pre-selected cervical samples archived as cell pellets at −80 ◦ C. DNA was extracted and analysed by GP5+/6+-EIA. This assay is routinely used in our laboratory for HR- and LR-HPV detection and for HPV genotyping. In interlaboratory comparisons the assay was shown to provide reliable results.23 Of 261 cervical samples, 187 were HR-HPV positive and 74 were HR-HPV-negative by the GP5+/6+-EIA. All samples were genotyped using type-specific probes in individual EIA reactions. 30 samples contained HPV types which cannot be detected by the multiplex real-time PCR assay. Considering multiplex HPV types only (HPV 16, 18, 31, 33, 45, 52, 58) 145 samples represented single infections and 12 represented double or triple infections. Multiplex PCR was performed twice and identical results were obtained with one exception. In one sample the first analysis was negative but the second analysis was HPV 16 positive. The GP5+/6+-EIA was only repeated for samples which showed discordant results. ␤-Globin could be amplified in both assays for all cases. The results for both tests are shown in Table 4. By considering multiplex HPV types only, 151 typing results were concordantly positive (8.3%), 1646 were concordantly negative (90.1%) and 30 were discordant (1.6%) yielding a kappa value of 0.901. This resulted in identical genotyping for 234 of 261 (89.6%) clinical samples. For single HPV types kappa values were perfect or almost perfect in

case of HPV 45 (kappa 1.0) and HPV 16 (kappa 0.929) respectively. Even for HPV 52 which showed the least agreement a kappa value of 0.435 was obtained. Twelve cervical scrapes were positive for two or more HPV types. For these samples 16 were concordantly positive (19.0%), 56 were concordantly negative (66.7%) and 12 were discordant (14.3%) yielding a kappa value of 0.630. This resulted in identical typing results for only 2 of 12 clinical samples with multiple infections. 3.3. Relative quantification of HPV DNA levels and relationship to histomorphological findings All cervical scrapes were taken in the course of colposcopic examination. Moreover, biopsies were taken if the morphological appearance was suspicious of CIN or worse. Altogether 118 patients fulfilled this criterion. Of these 90 were positive for one or several HPV types detectable by the multiplex real-time PCR assay. Forty did not reveal CIN, 16 were scored as CIN1, 14 as CIN2, 12 as CIN3 and 8 were diagnosed as invasive cervical carcinomas. For the corresponding cervical scrapes we calculated the relative levels of HPV DNA (delta Ct values of ␤-globin and HPV) and correlated these values with the histopathological findings (Fig. 2). The median Ct values for CIN1, CIN2 and CIN3 are 2.42, 4.24 and 3.69 respectively which translate to less than a 4-fold difference. Moreover, Mann–Whitney analyses (˛ = 0.05) revealed no significant differences between these 3 groups. However, highly significant differences were evident when comparing all grades of CIN with

Fig. 2. Correlation between the histological findings and the normalised Ct values of the prevalent HPV type detected by the multiplex real-time PCR assay. It represents the measurement of the quantity of HPV DNA in the sample and according to the Mann–Whitney U-test, the highlighted cases show significantly different amounts of HPV DNA.

306

M. Schmitz et al. / Journal of Clinical Virology 44 (2009) 302–307

either the group presenting no CIN (p = 0.0028) or with the group of cervical carcinomas (p = 0.0001). 3.4. Clinical sensitivity By screening for 7 high-risk types only, the chances of missing CIN2+ may be critical. We therefore compared the clinical sensitivity of both assays. Of all patients with conspicuous colposcopy (n = 118) biopsies were taken. All cases of CIN and cervical carcinomas were detected by the GP5+/6+-EIA. The multiplex real-time PCR detected all CIN3+ (n = 12) but missed one CIN2 and two CIN1 which were positive for HPV 51, HPV 39 and HPV 52, respectively. 4. Discussion Evidence is accumulating that HPV genotyping may be useful for patient management in the future. In order to further evaluate this possibility, practical and cost-effective assays are required. We have designed an assay for genotyping and quantification of the 7 most frequent cancer related HR-HPV types (16, 18, 31, 33, 45, 52 and 58). In a recent cross-sectional study these types were also shown to be the predominant HPV types in women with high grade lesions.24 Our multiplex real-time PCR assay proved to be highly specific for the detection of HPV 16, 18, 31, 33, 45, 52 and 58. No crossreactivity with any of the other HR-HPV types that infect the genital tract was evident. The dynamic range observed was similar for each target and was characteristic for real-time PCR. Evaluation of DNA extracted from archived cell pellets of cervical scrapes by the multiplex assay and the GP5+/6+-EIA showed identical genotyping for 234 of 261 (89.6%) samples and an almost perfect agreement when considering all typing results (kappa 0.901). This is of particular note since both assays differ considerably in their design: The GP5+/6+-EIA is based on consensus primer in a highly conserved region of the L1 gene and the use of type-specific probes for the detection of the PCR product, whereas the multiplex real-time PCR uses type-specific primers within the LCR/E6 or E6/E7 region and type-specific TaqMan probes. In numerous previous screening studies the GP5+/6+-EIA was shown to be a highly reproducible assay with an excellent clinical sensitivity for CIN2+23,25,26 and was thus considered to be adequate for comparison. Most of the discordant results were seen in cases of multiple infections. Only 2 of 12 samples showed identical genotypes. For 4 cases this could be attributed to the HPV 58 specific probe of the GP5+/6+-EIA which showed cross-reactivity with the PCR product of HPV 33 and vice versa. Similarly, the GP5+/6+ probe for HPV 45 cross-reacted with the PCR product of HPV 18 which accounts for another 3 discrepancies. Finally, sequence analyses confirmed that in case of 3 double infections the multiplex assay missed HPV 31, HPV 33 and HPV 52 respectively. Generally, agreement between various assays for genotyping multiple infections is poor. A recent study which evaluated different RLB assays for the identification of low prevalent HPV types revealed only moderate inter-assay agreement in cases of single HPV infections and poor agreement in cases with multiple HPV infections.27 Our multiplex real-time PCR assay detected all CIN3+ (8 CxCa and 12 CIN3) but missed one CIN2 and two CIN1. In terms of sensitivity for CIN2+ the multiplex assay was only marginally less sensitive than the GP5+/6+-EIA but considerably more specific. Of 187 HRHPV positive samples, 30 cases were positive for HPV types which cannot be detected by the multiplex real-time PCR assay. However, among these 30 cases only one CIN2 (HPV 51) was identified. Future studies will have to show whether this trade off is acceptable. From an economic point of view the multiplex assay is cheap (approximately D 2) and requires less than 24 h to complete including DNA

extraction. Moreover, the real-time PCR approach enables reliable quantification of the target DNA. In this study we could show that the median relative levels of HPV DNA do not vary by more than 4-fold irrespective of the severity of CIN. There is also no statistical significance between these groups by Mann–Whitney analyses. Thus although the number of cases in each group is small the data indicate that viral load in this cross-sectional study does not correlate with disease progression within the CIN spectrum. However, highly significant differences are evident when comparing the group comprising all CIN with the group lacking CIN (p = 0.0028) or with the cancer group (p = 0.0001). A significant decrease in viral load in liquid based cytology samples from cervical carcinoma patients compared to CIN was also reported by Yoshida et al.28 Our quantitative data also shows that the relative number of HPV genome copies within a histological entity can vary up to 32-fold within the upper and lower quartile of the box plot (Fig. 2). The differences are several log-fold when considering the extreme values of each group. Similar observations were made in a recent study in which the variation in HPV 16 viral load within different histological grades of cervical neoplasia was evaluated.29 It is therefore likely that changes in viral load over time, rather than a single measurement, might be predictive for disease progression or clearance.30 This possibility can now be addressed by the use of our multiplex real-time PCR, not just for HPV 16 but also for the other clinically relevant types. Moreover, the assay will be useful to evaluate the clinical relevance of viral persistence at the genotype level, to monitor disease recurrence and to examine the effect of widespread vaccination on HPV type prevalence in the future. Conflict of interest statement All authors have agreed to the content of the manuscript and its submission to your journal. There are no relationships that could be construed as resulting in an actual, potential, or apparent conflict of interest with regard to the manuscript submitted for review. All authors declare that the content of the manuscript is original and has not been published or accepted for publication, either in whole or in part, in any form. Moreover, no part of the manuscript is currently under consideration for publication elsewhere. References 1. 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. 2. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–27. 3. Munoz N, Castellsague X, de Gonzalez AB, Gissmann L. HPV in the etiology of human cancer. Vaccine 2006;24(Suppl. 3):S3/1–/10 [Chapter 1]. 4. Cuzick J, Clavel C, Petry KU, Meijer CJ, Hoyer H, Ratnam S, et al. Overview of the European and North American studies on HPV testing in primary cervical cancer screening. Int J Cancer 2006;119:1095–101. 5. Grce M, Davies P. Human papillomavirus testing for primary cervical cancer screening. Expert Rev Mol Diagn 2008;8:599–605. 6. Wright Jr TC, Schiffman M, Solomon D, Cox JT, Garcia F, Goldie S, et al. Interim guidance for the use of human papillomavirus DNA testing as an adjunct to cervical cytology for screening. Obstet Gynecol 2004;103:304–9. 7. van Ham MA, Bakkers JM, Harbers GK, Quint WG, Massuger LF, Melchers WJ. comparison of two commercial assays for detection of human papillomavirus (HPV) in cervical scrape specimens: validation of the Roche AMPLICOR HPV test as a means to screen for HPV genotypes associated with a higher risk of cervical disorders. J Clin Microbiol 2005;43:2662–7. 8. Jacobs MV, de Roda Husman AM, van den Brule AJ, Snijders PJ, Meijer CJ, Walboomers JM. 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. 9. Manos M, Ting Y, Wright D, Lewis A, Broker T, Wolinsky S. Use of polymerase chain reaction amplification for the detection of genital human papillomaviruses. Cancer Cells 1989;7:209–14. 10. Castle PE, Solomon D, Schiffman M, Wheeler CM. Human papillomavirus type 16 infections and 2-year absolute risk of cervical precancer in women with equivocal or mild cytologic abnormalities. J Natl Cancer Inst 2005;97:1066–71.

M. Schmitz et al. / Journal of Clinical Virology 44 (2009) 302–307 11. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, et al. 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;97:1072–9. 12. Venturoli S, Ambretti S, Cricca M, Leo E, Costa S, Musiani M, et al. Correlation of high-risk human papillomavirus genotypes persistence and risk of residual or recurrent cervical disease after surgical treatment. J Med Virol 2008;80:1434–40. 13. Brummer O, Hollwitz B, Bohmer G, Kuhnle H, Petry KU. Human papillomavirustype persistence patterns predict the clinical outcome of cervical intraepithelial neoplasia. Gynecol Oncol 2006;102:517–22. 14. Kleter B, van Doorn LJ, Schrauwen L, Molijn A, Sastrowijoto S, ter Schegget J, et al. Development and clinical evaluation of a highly sensitive PCR-reverse hybridization line probe assay for detection and identification of anogenital human papillomavirus. J Clin Microbiol 1999;37:2508–17. 15. van den Brule AJ, Pol R, Fransen-Daalmeijer N, Schouls LM, Meijer CJ, Snijders PJ. GP5+/6+ PCR followed by reverse line blot analysis enables rapid and highthroughput identification of human papillomavirus genotypes. J Clin Microbiol 2002;40:779–87. 16. Lindh M, Gorander S, Andersson E, Horal P, Mattsby-Balzer I, Ryd W. Realtime Taqman PCR targeting 14 human papilloma virus types. J Clin Virol 2007;40:321–4. 17. Nishiwaki M, Yamamoto T, Tone S, Murai T, Ohkawara T, Matsunami T, et al. Genotyping of human papillomaviruses by a novel one-step typing method with multiplex PCR and clinical applications. J Clin Microbiol 2008;46:1161–8. 18. Holland PM, Abramson RD, Watson R, Gelfand DH. Detection of specific polymerase chain reaction product by utilizing the 5 –3 exonuclease activity of Thermus aquaticus DNA polymerase. Proc Natl Acad Sci USA 1991;88:7276–80. 19. Munoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, et al. Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–85. 20. Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hildesheim A, et al. A study of the impact of adding HPV types to cervical cancer screening and triage tests. J Natl Cancer Inst 2005;97:147–50.

307

21. Durst M, Gissmann L, Ikenberg H, zur Hausen H. A papillomavirus DNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proc Natl Acad Sci USA 1983;80:3812–5. 22. Jacobs MV, Snijders PJ, van den Brule AJ, Helmerhorst TJ, Meijer CJ, Walboomers JM. A general primer GP5+/GP6(+)-mediated PCR-enzyme immunoassay method for rapid detection of 14 high-risk and 6 low-risk human papillomavirus genotypes in cervical scrapings. J Clin Microbiol 1997;35:791–5. 23. Jacobs MV, Snijders PJ, Voorhorst FJ, Dillner J, Forslund O, Johansson B, et al. Reliable high risk HPV DNA testing by polymerase chain reaction: an intermethod and intramethod comparison. J Clin Pathol 1999;52:498–503. 24. Klug SJ, Hukelmann M, Hollwitz B, Duzenli N, Schopp B, Petry KU, et al. Prevalence of human papillomavirus types in women screened by cytology in Germany. J Med Virol 2007;79:616–25. 25. Bulkmans NW, Rozendaal L, Snijders PJ, Voorhorst FJ, Boeke AJ, Zandwijken GR, et al. POBASCAM, a population-based randomized controlled trial for implementation of high-risk HPV testing in cervical screening: design, methods and baseline data of 44 102 women. Int J Cancer 2004;110:94–101. 26. Schneider A, Hoyer H, Lotz B, Leistritza S, Kuhne-Heid R, Nindl I, et al. Screening for high-grade cervical intra-epithelial neoplasia and cancer by testing for high-risk HPV, routine cytology or colposcopy. Int J Cancer 2000;89: 529–34. 27. Sabol I, Salakova M, Smahelova J, Pawlita M, Schmitt M, Gasperov NM, et al. Evaluation of different techniques for identification of human papillomavirus types of low prevalence. J Clin Microbiol 2008;46:1606–13. 28. Yoshida T, Sano T, Kanuma T, Owada N, Sakurai S, Fukuda T, et al. Quantitative real-time polymerase chain reaction analysis of the type distribution, viral load, and physical status of human papillomavirus in liquid-based cytology samples from cervical lesions. Int J Gynecol Cancer 2008;18:121–7. 29. Fiander AN, Hart KW, Hibbitts SJ, Rieck GC, Tristram AJ, Beukenholdt RW, et al. Variation in human papillomavirus type-16 viral load within different histological grades of cervical neoplasia. J Med Virol 2007;79:1366–9. 30. Monnier-Benoit S, Dalstein V, Riethmuller D, Lalaoui N, Mougin C, Pretet JL. Dynamics of HPV16 DNA load reflect the natural history of cervical HPVassociated lesions. J Clin Virol 2006;35:270–7.