Anogenital warts is the most common sexually transmitted disease in the world, and is also a marker of transmission of high-risk HPV types, such as HPV 16 and HPV 18, due to similar infection routes. Diagnosis with genital warts is associated with a long-term increased risk of anogenital and head and neck cancers . The monitoring of clinically obvious disease, such as anogenital warts, therefore has an important bearing upon overall HPV disease in the community at large. It is accepted that HPV 6 and HPV 11 are the dominant types in the pathogenesis of anogenital warts and in the upper respiratory tract equivalent, respiratory papillomatosis . However, it is also known that other benign or low-risk types have been implicated in the causation of anogenital warts [5, 10, 18–20]. One potential confounding factor in these studies is that surface swabbing of lesions for HPV DNA, a commonly used method to define the HPV types involved, is susceptible to contamination with unrelated HPV types.
In this study, laser-capture microdissection was used to facilitate the delineation of HPV types likely to be surface contaminants from those likely to be involved in the aetiology of the wart lesion. Subsequent real-time PCR analyses for both E6 DNA and E6 mRNA allowed their suitability for monitoring purposes to be determined. The use of simple equations, based on the premise that HPV types identified from within the deeper epithelium of the wart are more likely to be causal, allowed “infection candidates” to be identified. Messenger RNA detection for five common HPV types found in this cohort was then used to identify the most likely candidate to be involved in actual lesion pathogenesis. HPV 6 was a more frequent cause of lesions than HPV 11. Interestingly, the majority of HPV types identified as surface contaminants were high-risk types (75%), with the remainder being identified as low-risk types (25%). The majority of contaminant HPV types featured only once or twice in the samples (incidence <8%). HPV 16 was the most common surface contaminant (25% of wart lesions). Interestingly, many of the surface contaminants detected via the LCM and typing analyses were high-risk types and we speculate that perhaps high-risk types have an innate survival advantage over many low-risk types (with the exception of HPV 6 and 11, which are very successful wart developers) and this aids their transmission rates.
The incidence of multiple infections detected in wart studies ranges from 33.8% to 40.1% when using only swabbing techniques [10, 21], as well as 21.7% when using fresh wart tissue specimens . These infections were a mixture of high-risk and low-risk types (range: 2–5 types) but usually contained HPV 6 or 11, or both. In our cohort, 28.6% anogenital warts were co-infected with active (E6 mRNA+) HPV types, similar to the published estimates discussed above. However, it is important to note that this statistic is two-fold less than the co-infection levels observed for the DNA detection methods and without exclusion of surface contamination. Therefore existing data concerning multiple infections may be an overestimate. On the other hand, it is possible that the suspected multiple infections are in fact two different lesions growing together and appearing as a single lesion, as has been shown for the cervix . Quint et al. demonstrated that for the cervix, a single lesion is most often associated with a single HPV type. Where multiple types were found, careful laser-capture microdissection of histopathological sections found that different types were associated with several abutting lesions, but with the over-riding conclusion that each type caused a single lesion. There are no data to suggest that two or more types of HPV can infect a single cell and this study does not address that question.
Moderate concordance exists between whole wart tissue and corresponding swabs in terms of HPV detection (κ = 0.574), with high concordance levels for HPV types 6, 11, 31, 59 and 81. This illustrates that anogenital swabs can be used with reasonable reliability to determine HPV types detected in whole wart tissue. However, this study has also shown that HPV DNA viral load estimates from swab samples do not reflect that which is observed in paired wart tissue. Furthermore, the elucidation of surface contamination highlights potential limitations with non-adjusted HPV prevalence data obtained from swab samples [5, 10, 18, 23] and also disagrees with the claim that HPV identification in swabs can be used for analysing clinical changes in a specific lesion. It would thus seem apparent that careful consideration may be needed concerning the existing literature on HPV prevalence in anogenital warts. However, the use of swabs for monitoring the effects of vaccination and treatment  is likely to remain valid.
Interestingly, the lack of correlation between DNA viral loads and mRNA expression suggests that measuring viral DNA levels is not representative of viral activity within the wart. It is possible that the efficiency of the viral transcription machinery differs between individual viruses. Alternatively, it is possible that either the detection sensitivities for the viral DNA/mRNA components using the methods employed vary or there is a differing level of viral E6 gene expression during various stages of the virus replication lifecycle.
The importance of HPV 6 and HPV 11 in the pathogenesis of classic condylomata acuminata has been demonstrated here, and has been deemed to be causal in over 90% of this small cohort of 23 patients. These data therefore support published studies in the literature. There has however, been little RNA work performed in wart disease which could serve to either support or refute this finding, only previous publications which quote the 90% level of causation for these two dominant types [24, 25]. It may be that the total amount of clinical wart disease prevented via the L1 prophylactic vaccine will be similarly high or even higher. Furthermore, if herd immunity of human populations around the world can be achieved utilising currently available L1 prophylactic vaccines, rare diseases such as recurrent respiratory papillomatosis (RRP) will also be reduced but this will require long-term population surveillance.
The detailed and thorough molecular work involved in this small study has revealed that, although multiple infections are apparently a frequent occurrence when determined by swabbing wart lesions, the actual numbers of types potentially involved in lesion pathogenesis is low. This is an important and potentially confounding factor in all studies that involve only the surface swabbing of wart lesions. We also show for the first time that laser capture microdissection can be utilised successfully to assist in objective analyses of the aetiology of anogenital warts, but there was poor correlation with detected RNA levels. This technique is expensive and requires specialised skills and is likely to remain a research tool only.
In this study, samples 4, 9, 11 and 12 demonstrated evidence of dual infection. Interestingly, HPV 16 was shown be elevated for both DNA and mRNA in sample 12, which is evidence of active replication in that lesion. In four cases (samples 3, 4, 9 and 11), HPV 11 was not found by the Linear Array system. The reason for this is not known but may be related to sensitivity levels. The manufacturer has found that their product is able to detect HPV 11 plasmids at a level of 900 copies/ml and it is possible that the virus is present with a DNA quantity lower than this detection threshold. The data shows that HPV 6 or HPV 11 appear to be involved causally in a majority of lesions clinically diagnosed as typical condylomata acuminata. The data also suggests a potential role for HPV 16 or other types in isolated cases of wart disease and this hypothesis needs to be explored in further studies.