Detection of High-Risk Papillomavirus DNA with Commercial Invader-Technology-Based Analyte-Specific Reagents following Automated Extraction of DNA from Cervical Brushings in ThinPrep Media

Abstract

We compared the performances of the Third Wave Technology Invader method and the Digene Hybrid Capture 2 assay to detect high-risk human papillomaviruses in 87 cervical brushing specimens submitted in Cytyc ThinPrep media. Two different methods for the extraction of DNA from squamous epithelial cells were also evaluated.

The human papillomavirus (HPV) family consists of more than 100 types. Those shown to be causative agents of cervical cancer are referred to as “high-risk” strains. Current guidelines recommend the testing of all women with a cytological diagnosis of atypical squamous cells of undetermined significance (ASCUS) for infection by high-risk types of HPV (16, 23). HPV testing has been found to have a higher sensitivity than standard or liquid-based Pap testing for detecting high-grade cervical intraepithelial neoplasia (1, 17-20). Only the Digene Hybrid Capture 2 (HC2) assay (Digene, Gaithersburg, MD) has been approved by the FDA for the detection of high-risk strains of HPV as an ASCUS triage for referral to colposcopy, and, in conjunction with cervical cytology, as a primary screen for cervical cancer. The purpose of this study was to compare a laboratory-developed test utilizing analyte-specific reagents (ASRs) developed by Third Wave Technologies Inc. (TWT, Madison, WI) with the Digene HC2 assay for the detection of high-risk HPV DNA in cervical brushings. The TWT method used in this study, based on proprietary Invader technology, employed isothermal signal amplification to detect 13 high-risk HPV types, utilizing three probe pools (A5/A6 [types 51 and 56], A7 [types 18, 39, 45, 59, and 68], and A9 [types 16, 31, 33, 35, 52, and 58]) based on phylogenetic relatedness. The assay also incorporated an internal control for human alpha-actin to control for DNA quality and quantity in each reaction.

The 87 specimens used in this study were cervical brushings collected in ThinPrep fluid (Cytyc Corp., Marlborough, MA) submitted to the cytology laboratory at CompuNet Clinical Laboratories for routine Pap testing. Those samples that received a cytological diagnosis of ASCUS were tested for high-risk HPV by using the Digene HC2 high-risk HPV test according to the manufacturer's instructions. Cells were collected from 2 ml of ThinPrep fluid by centrifugation at 14,000 × g for 10 min and suspended in 0.4 ml of phosphate-buffered saline. DNA was extracted using the Qiagen BioRobot M48 and the MagAttract virus DNA minikit (Qiagen Inc., Valencia, CA) and eluted into 100 μl of RNase-free distilled water. Duplicate samples were prepared from 2 ml of each cytology sample by using Gentra Systems Puregene spin columns (Qiagen Inc., Valencia, CA) by following the manufacturer's instructions. All extracted samples were stored at −20°C until tested. The extracted DNA was tested for the presence of high-risk HPV by a laboratory-developed assay utilizing Invader HPV ASRs. The TWT ASRs consisted of three separate oligonucleotide pools with 6-carboxyfluorescein (FAM)-labeled probes, one Redmond red (RED)-labeled oligonucleotide set for the internal control, Cleavase X enzyme, and MgCl₂ solution. Each of three 10-μl aliquots from each sample was added to separate wells of a 96-well reaction plate and layered with 20 μl of mineral oil. The plates were incubated at 95°C for 5 min in a PTC-100 Peltier thermal cycler (MJ Research, Waltham, MA) and cooled to 63°C. Ten microliters of master mix containing one of the three oligonucleotide sets plus the internal control oligonucleotide pool was added to each of the triplicate aliquots. Each master mix consisted of 4 μl of MgCl₂, 4 μl of one FAM oligonucleotide mixture, 1 μl of RED oligonucleotide mixture, and 1 μl of Cleavase X enzyme per reaction. The plates were incubated at 63°C for an additional 4 h and then cooled to ambient temperature for 5 min. Readings were taken immediately after cooling, using a Tecan GENios microplate fluorometer (Tecan Inc., Durham, NC). Excitation/emission settings were 485 nm/535 nm and 560 nm/612 nm for FAM and RED, respectively. Data collected from the fluorometer were imported into a generic Microsoft Excel spreadsheet programmed with user-defined cutoffs to determine the positivity/negativity of the reactions.

For the manually extracted samples, 50 tested negative and the remaining 37 were positive for HPV DNA. Results obtained using the automated extractor were nearly identical, with 49 and 38 samples testing negative and positive, respectively, for HPV DNA. The one discordant positive specimen was determined to be HPV type 51 by PCR amplification and sequencing analysis. The Invader internal controls for all DNA extracts, irrespective of extraction methodologies, were positive, indicating that sufficient DNA was extracted from all of the specimens tested. It is not clear as to why the one sample that tested positive with the M48-extracted DNA gave a negative HPV result when extracted with the manual Puregene method. The major advantage of the M48 BioRobot was that it totally automated the extraction process, freeing laboratory personnel to perform other tasks. One technician could comfortably extract 12 ThinPrep specimens per hour by using the Puregene method. Total hands-on time using the M48 BioRobot was also approximately 1 h, including instrument setup and centrifugation steps; however, the Qiagen instrument has the capability of extracting 48 specimens per run.

A comparison of the results obtained by Digene HC2 testing versus those obtained by TWT Invader is shown in Table 1. The agreement between the Digene HC2 and TWT Invader tests was 81.6%. Of the 16 discordant samples, 15 were positive by Digene HC2 and negative by the TWT Invader method, while one was negative by the former and positive by the latter. Therefore, prior to the resolution of the discordant results, the sensitivity and specificity of the TWT Invader method compared to Digene were 77.6% and 97.22%, respectively. Of the 16 discordant specimens, 12 were analyzed by PCR using two different sets of HPV consensus primers (8, 25). The remaining four discordant specimens did not have sufficient ThinPrep material for further testing. Amplified DNA resulting from each positive PCR was cloned into the pGEM-T Easy vector. Two clones from each PCR were sequenced, with a total of four sequencing results being obtained for each discordant sample (two for each of the two PCR primer sets). The results of the PCR and sequencing analysis are shown in Table 2. In each case with negative TWT Invader/positive Digene HC2 results, either there was no HPV DNA detected by PCR (4/11) or the HPV sequence was determined to belong to a low-risk genotype (7/11). Sequence analysis of the sole TWT Invader-positive/Digene HC2-negative sample determined the sample to be high risk, type 51. As stated above, the Invader internal control demonstrated that sufficient DNA had been extracted for this method. There is no internal control in the Digene assay; therefore, it is not possible to determine whether results classified as negative by the Digene assay resulted from insufficient DNA after the specimen processing procedure used by this method. The Digene method detected one type 53 HPV; the TWT assay is not designed to detect this genotype. HPV type 53 was classified by the International Agency for Research on Cancer (IARC) as a probable high-risk type, but the IARC has not recommended its inclusion into high-risk HPV diagnostics (3, 11).

TABLE 1.

Comparison of TWT Invader and Digene HPV testing results

Digene result	No. of samples with indicated TWT result
	Positive	Negative
Positive	37	15
Negative	1	34

TABLE 2.

Resolution of TWT Invader and Digene discordant results

Sample no.	TWT Invader result	Digene result	HPV type
6Q	Negative	Positive	No HPV
13Q	Negative	Positive	No HPV
17Q	Negative	Positive	81
21Q	Negative	Positive	84
30Q	Negative	Positive	No HPV
49Q	Negative	Positive	30
57Q	Positive	Negative	51
62Q	Negative	Positive	70
65Q	Negative	Positive	42
76Q	Negative	Positive	No HPV
80Q	Negative	Positive	53
86Q	Negative	Positive	42

In their review of the clinical relevance of HPV testing, Snijders et al. discuss the relationship between analytical and clinical sensitivity and specificity (17). Analytical sensitivity refers to the number of HPV-infected women who are positive for HPV by a given method, while analytical specificity is the number of HPV-negative women correctly identified by that test. On the other hand, clinical sensitivity refers to the proportion of women who are determined to be infected with HPV and who also demonstrate cytological findings of high-grade cervical intraepithelial neoplasia (≥3 cervical lesions). The findings of several studies have clearly demonstrated that HPV testing is more sensitive for detecting high-grade cervical intraepithelial neoplasia than Pap smears alone (2, 4-7, 10, 12-14, 22). Based on these studies, it has been proposed that HPV testing be performed in addition to Pap testing on all women over the age of 30 (7). Colposcopic examination is the standard follow-up for women with abnormal cytological findings, including ASCUS. Based on the results from Digene HC2 testing, colposcopy would have been performed inappropriately on 11 women. Similar results were obtained in a recent study comparing the Amplicor HPV PCR assay (Roche Molecular Systems, Branchburg, NJ) to the Digene HC2 method (9). The PCR assay also demonstrated considerable cross-reactivity with low-risk types of HPV since the consensus primers were designed to amplify all HPV types and not just high-risk types (21). With any screening assay, specificity is typically sacrificed to ensure a very high degree of sensitivity. However, with most screening assays, (e.g., human immunodeficiency virus and hepatitis C virus enzyme immunoassays), positive results are confirmed by a secondary method with high specificity. Unfortunately, no such confirmatory assays are currently available for HPV testing. Therefore, if HPV DNA testing is to be performed as an adjunct to Pap testing, the analytical specificity of the HPV test becomes of critical importance to ensure that women are not subjected to a needless invasive procedure. Our results clearly demonstrate that the TWT Invader HPV ASR methodology is suitable for use as a companion to Pap testing. It should be pointed out that this study is based on a relatively small number of samples. However, the disparity between the performances of the Invader and Digene HPV assays, especially regarding the cross-reactivity of the latter with low-risk HPV types, is very apparent. It is expected that these results will be confirmed using a larger data set; such studies are now in progress. In the limited number of patients tested utilizing the Invader assay, there were no false-positive results and possibly only one false-negative result (if HPV type 53 is to be placed in the high-risk category).