Human Papillomavirus Genotyping Using Archival Vulval Dysplastic or Neoplastic Biopsy Tissues: Comparison between the INNO-LiPA and Linear Array Assays

ABSTRACT The Roche Linear Array (LA) and Innogenetics INNO-LiPA human papillomavirus (HPV) genotyping assays were compared for paraffin-embedded vulval tissues. The LA detected amplifiable DNA in 28 (57%) out of 49 biopsy specimens, 20 (40%) being HPV genotyped, compared to 49 (100%) and 41 (83%), respectively, detected by the INNO-LiPA. The INNO-LiPA provides greater sensitivity for HPV genotyping in archival tissue.

Persistent infection with oncogenic human papillomavirus (HPV) is a prerequisite for the development of HPV-related neoplasic precancerous changes in the anogenital region (13). Cervical and vulval biopsy specimens embedded in paraffin for histological diagnoses can also be utilized for detection of HPV genotypes. However, sensitive and accurate detection of HPV genotypes in such archival tissues could be affected, as DNA is often degraded as a result of long and poor storage conditions. Such damage to DNA includes chemical modification, crosslinking, and fragmentation, all of which can reduce the efficiency of PCR amplification (1,6).
This study compared two commercial HPV genotyping assays: the Linear Array assay (LA) (Roche Molecular Systems, Alameda, CA) (2) and the INNO-LiPA Genotyping Extra assay (LiPA) (Innogenetics, Ghent, Belgium) (5) to detect and type HPV in paraffin-embedded vulval tissue biopsy specimens. Although both methods have been applied previously to paraffin-embedded tissues (5, 7), a comparison of the two assays, to our knowledge, has not been reported to date.
The paraffin-embedded biopsy specimens were from a total of 49 histologically diagnosed cases of vulval disease, of which 30 were high-grade vulval intraepithelial neoplasia (VIN grade 2/3) and 19 were invasive vulval cancer, evaluated and diagnosed between January 1996 and December 2005 in women residing in the Northern Territory of Australia.
A sandwich sectioning method was used, with the outer sections stained with hematoxylin and eosin to confirm the histological diagnosis of the inner sections used for HPV detection (4). To minimize cross contamination, the microtome blade was changed and the microtome surface was cleaned after each sample was sectioned. Gloves were changed regu-larly, and the sectioning of control HPV-negative tissue specimens was carried out in a random order to ensure no possibility of HPV DNA carryover between sectioning of samples.
Each specimen was sectioned and deparaffinized according to the manufacturer's instructions for a Roche DNA Isolation tissue kit (Roche Molecular Systems). Briefly, a 7-m section was mixed with a 1.2-ml mixture of histolene (800 l) and ethanol (400 l). The tissue was pelleted at 14,000 ϫ g for 10 min, and the resultant pellet washed twice with 500 l of 70% ethanol. The tissue was air dried to ensure that no residual ethanol was present and was subsequently incubated with 160 l of tissue lysis buffer (Roche Molecular Systems) and 40 l of proteinase K (Roche Molecular Systems) on a 55°C heat block until fully digested. The digested tissue was then transferred onto an automated Roche MagNA Pure LC isolation and purification system, using a DNA Isolation kit 1. The final isolated DNA volume in elution buffer was 100 l.
The Roche Linear Array HPV genotyping test (Roche Molecular Systems) directs the amplification of a 450-bp region of the HPV L1 gene and allows the identification of 37 anogenital HPV genotypes: HPV  , and CP6108, as well as the amplification of a 265-bp region of the human ␤-globin gene that serves as an internal control. Samples were denatured and assayed using a modified method as previously described (8,9).
The INNO-LiPA HPV genotyping test (Innogenetics) directs the amplification of a 65-bp region of the HPV L1 gene and allows the identification of 28 anogenital HPV genotypes: HPV  Comparisons of the detection of internal controls and of single and multiple HPV genotypes by each assay are shown in Tables 1, 2, and 3. All strips were interpreted visually.
Amplification of the respective internal controls demonstrated that 28 (57%) biopsy specimens had adequate amplification by the LA, whereas all 49 (100%) biopsy specimens had adequate amplification of the internal control in the LiPA (P Ͻ 0.0001) ( Table 1). In examining the numbers of biopsy specimens with single infections, the LiPA detected 34, of which 28 (82%) were HPV 16, and the LA found 18, of which 17 (94%) were HPV 16, a statistically significant difference (P ϭ 0.0258) ( Table 2). Overall, when taking into account samples with multiple infections, 41 (83%) samples were positive for HPV by the LiPA compared with 20 (40%) detected by the LA (Table 3).
Sensitive methods of HPV DNA detection are important for accurate identification of HPV genotypes in neoplastic tissue. The differences in sensitivity between these assays can be attributed to the length of amplicon targeted, with the amplicon in LA being 450 bp and that in the LiPA 65 bp. Longer target amplicons can have reduced amplification efficiency due to fragmentation of the DNA template, particularly in archival tissue (3). In this study, all HPV genotypes detected by the LA were also detected by the LiPA, except in one sample where the LiPA detected HPV 16 and the LA detected HPV 16 and 6. Competition between primers may have contributed to this discrepant result, as decreased sensitivity for amplification of genotypes can occur when multiple HPV infections are present with one genotype in excess concentration (10,11).
Certain challenges arise when paraffin-embedded biopsy specimens are utilized for the detection of DNA targets. Utilization of assays targeting longer amplicons, like the LA, could result in decreased sensitivity on paraffin-embedded tissue. A previous study demonstrated the use of the LA on archival cervical cancer tissue (7); however, this method involved screening samples with primers targeting the HPV L1 gene and utilized a nested PCR approach for application with the LA. A nested approach is time consuming and more prone to contamination and hence is not practical for large epidemiological studies. The LiPA has a greater sensitivity for detection, given the smaller amplicon target, and so a greater number of samples will produce assessable results, as found in this study. Similar to other PCR assays, it also has the limitation of detecting not only the agent causing a particular lesion (one genotype per lesion, since they are clonal) but other colonizing HPVs, related but not causing the lesion (12).
As tissue biopsy specimens are routinely embedded in paraffin for histological diagnoses and then stored over time, there is an important need for accurate detection of HPV genotypes in archival tissue. Although a limitation of this study is the small number of samples examined, the LiPA assay had greater sensitivity than did the LA in the detection of HPV genotypes in paraffin-embedded tissue. Although the LA is able to provide accurate results for liquid-based cytology (2, 11) and frozen tissue (14), we found that the LiPA offers a more-sensitive detection method for HPV genotypes in paraffin-embedded biopsy specimens archived for up to 10 years.  The total numbers of samples with multiple infections found by the assays were 7 for the LiPA and 3 for the LA. Of a total of 49 samples, the LiPA found 41 to be HPV positive and the LA found 20 to be HPV positive.
b HPV 84 is only detected by the LA.