HPV-negative condylomas often result of error in diagnosis, sampling or testing
Genital warts or Condylomas are the most common viral sexually transmitted disease in the world. Although genital warts should, essentially by definition, be caused by the Human Papillomavirus (HPV), a minority of about 4% of condylomas are testing HPV-negative even when tested with several different “general” primer PCR systems that should detect all possible HPVs. However, using deep sequencing of 40 apparently “HPV-negative” condylomas, we found HPV in 37/40 samples. The HPVs found by sequencing were either already known HPVs (that should have been detected in the first place but for some reason were not) or previously unknown ones that are not detected by general primer PCR. With the proportion of “HPV-negative” condylomas now being down to 0,3%, there is reason to ask if HPV-negative condylomas actually exist as a biological entity or are merely a result of error in either diagnosis, sampling or testing.
The work was inspired by a previous study of a single patient with multiple skin cancers on both hands. The cancers on one hand were positive for high copy numbers of HPV26, but the cancers on the other hand were only positive in the general PCR for extremely low levels of HPV26 (much less than a copy per cell), which seemed unreasonable. By unbiased cloning we found that these cancers contained extremely high levels (107 copies per cell – the virus band was even visible when the tumor sample was simply run on a gel!) of HPV88, but that the general primer PCR just did not pick up this HPV type at all. We reasoned that for reliable virus detection, we really had to go for PCR-free methods that were not biased by primer sequences. Fortunately, deep sequencing developed fast. We basically had to buy a new platform every year to keep up. With the more recent platforms it is straightforward to detect as little as 1 virus copy per cell and voila´ – now most of the seemingly “HPV-negative” condylomas were indeed HPV-positive. At an international HPV conference, we were asked “if it is so straightforward to do virus detection that is not biased by PCR, why is not everybody doing it?”. Historically, it has been expensive and the bioinformatics complicated, but that is not so any longer. When it comes to virus detection, we will certainly no longer take “no” for an answer without having sequenced the sample first.
Bayesian phylogenetic tree based on the L1 part of the complete 198 established HPV types and 3 putative novel HPV types (SE-types.). Alpha, Beta, Gamma, Mu and Nu papillomaviruses are presented in red, green, blue, orange and purple colors, respectively.
About the authors
Sara Arroyo Myhr is a Swedish-Spanish graduate student who coordinated the study. Emilie Hultin is the deep sequencing expert who ensures that we always have the latest technology in-house. Davit Bzhalava is the bioinformatician who single-handedly perfoms bioinformatics analyses that just a few years ago required a whole building with high performance computers. Associate professor Ola Forslund and his coworkers at Lund University collected and HPV-tested the so far largest collection of condylomas in the world. Joakim Dillner is the professor of infectious disease epidemiology of the Karolinska Institutet, Stockholm, Sweden.
About the research
Virology, Volume 485, November 2015, Pages 283–288
Laila Sara Arroyo Mühr, Davit Bzhalava, Camilla Lagheden, Carina Eklund, Hanna Johansson, Ola Forslund, Joakim Dillner, Emilie Hultin