The VisionArray® HPV Chip was successfully implemented in our laboratory with good analytical capabilities illustrated by the high sensitivity and specificity in comparison to our previously used standard reference assay, pan-HPV DNA PCR which had a sensitivity of 86.7% and specificity of 92% as shown previously [1]. Another advantage of the VisionArray® HPV Chip assay is that it provides simultaneous HPV genotyping as a same-step procedure detecting 41 clinically relevant genotypes classified as low risk, probably high risk, and high risk. In the current validation cohort, the most frequently detected HPV genotype was HPV16 which reflects the demographic patient group from Eastern Denmark [7, 8]. Prior to this study, we had to acquire the genotype as a secondary analysis using next-generation sequencing which was more time consuming and expensive.

To our knowledge, HPV genotyping does not have clinical implications in terms of treatment or prognostication but it may provide important information for research purposes in epidemiologic studies and in the etiopathogenesis of OPSCC. In future, it may add value in relation to liquid biopsies with cell-free HPV DNA in the follow-up surveillance after treatment [10]. A recent study has investigated the role of high-risk HPV genotypes on survival in patients with oropharyngeal cancer, and found no differences when comparing HPV16 with non-HPV16 genotypes [11]. The subgroup analysis indicated that the group of patients with genotypes HPV33 and HPV35 has a significantly better 5-year overall survival than other non-HPV16 genotypes.

Advanced HPV testing including genotyping can be adopted by the pathologist in other head and neck entities. It has an important role in the diagnostics of sinonasal malignancies, in particular the HPV-related Multiphenotypic Sinonasal Carcinoma, which requires the presence of certain HPV genotypes [12]. The 5th edition of the world health organization classification of head and neck tumors recommends that high-risk HPV must be demonstrated by in situ hybridization or PCR-based techniques, specifically to include type 33. p16 alone is not sufficiently specific to make the diagnosis [13].

HPV genotyping is often requested by otolaryngologists after excision of respiratory papillomas where approximately 30% of head and neck papillomas are related to HPV [14]. Establishing low-risk HPV genotypes such as HPV6 and HPV11 plays an important role during the patient consultation.

The VisionArray® HPV Chip assay has proven to be highly suitable in clinical practice, as the HPV testing can be performed within 24 h compared with the more time-consuming pan-HPV DNA PCR where the turnaround time is up to four days. This is especially useful if applied on cytology specimens, and if HPV-positive, it facilitates the diagnostic work-up and treatment for patients who present with a metastatic squamous cell carcinoma of the neck and unknown primary. HPV DNA testing is feasible on previously stained cytology smears as previously shown [1, 5]. The use of p16 staining on cytology specimens is not recommended as there is no validated cutoff value and it requires the preparation of a cell block, with a risk of not having viable tumor cells [15].

This study was strengthened by the large validation cohort including both HPV-positive and negative OPSCC, and benign tumors previously tested for HPV DNA. Among the HPV-negative patient samples, we experienced a few samples that had either poor DNA quality or insufficient DNA concentration for HPV testing which could be explained by several factors. Firstly, we utilized 4- to 5-year-old archived FFPE material where the DNA quality can be varying, and secondly, some samples consisted of very small biopsies with none or limited tumor tissue left after being used for p16 staining and the reference pan-HPV DNA PCR assay. As part of every implementation process, we must highlight the importance of the pre-analytical steps in the laboratory to avoid contamination and cross-contamination between patient samples that could lead to inaccurate results.

The HPV testing algorithms and choice of assay vary among laboratories across the world and are dependent on the resources, capacity, and staff at the different institutions. A full review of all HPV testing assays is beyond the scope of this paper; nonetheless, the most commonly used targets are HPV DNA, HPV RNA, viral oncoproteins, cellular proteins, and HPV-specific serum antibodies [16, 17]. In the CAP guidelines, p16 IHC is recommended for biopsy/resection specimens and is the most widely used surrogate marker for HPV [4]. The guidelines further state that additional HPV-specific testing may be done based on the decision of the pathologist and/or treating clinician, or in the context of a clinical trial. p16 IHC is suitable as a stand-alone assay in most clinical settings for histologic specimens deriving from oropharyngeal tumors as it provides acceptable sensitivity and specificity, is much more cost-effective than molecular tests, has a short turnaround time, and is easy to analyze. However, the additional use of specific HPV DNA testing should be strongly considered for patients that are p16-positive and potential candidates for clinical trials that offer either de-escalation or intensification of treatment. This is based on a recently published multinational study which investigated discordant p16/HPV oropharyngeal cancers and its prognostic implications [3]. The authors argue that an exception could be made in some geographical regions which is associated with a high p16/HPV concordance (i.e., North America). Interestingly, the study showed that if p16 IHC alone is used to determine HPV status, 8.1% of p16-positive patients worldwide and up to almost 26% in regions of low HPV-attributable fractions such as southern Europe would be incorrectly classified as having HPV-related tumors.

Therefore, at our institution, HR-HPV testing is currently being applied on either a resection specimen of a metastasis or on the primary tumor (biopsy/resection) along with p16 IHC as the majority of our patients participate in clinical trials. In the few discordant p16-positive and HPV DNA-negative cases, we request HPV genotyping at a different laboratory as a quality assessment, to ensure that we do not fail to detect a genotype that is not detected by the VisionArray® HPV Chip assay which has not been an issue to date.

The recommended gold standard method for HR-HPV testing is the mRNA ISH which detects transcriptionally active HPV E6/E7 oncogenes. This assay has previously been validated against p16 and HPV DNA by several institutions who agree on its excellent analytical capabilities, technical feasibility, and fast turnaround [6, 18,19,20,21]. Furthermore, it provides direct visualization of the HPV-positive staining which further strengthens the sensitivity and specificity. However, the assay is limited by its use on research platforms only; it is less cost-effective than p16 IHC and PCR-based assays and currently only allows for the detection of mRNA of up to 18 HR-HPV genotypes in a single cocktail probe. Based on our validation cohort, it would have failed to detect HPV-type 67 in a single patient which is classified as a “probably high risk” HPV type using the VisionArray® HPV Chip. Another limitation is that the ISH method requires a morphologically well-preserved specimen where the cell nuclei are well visualized. On the contrary, the PCR-based analysis works well on crushed tumor cells like the defrosted frozen section tissue specimen or laser coagulated tissue specimens. In the future, it is necessary to clarify when to perform HPV mRNA ISH in routine clinical use in relation to the existing assays.

In conclusion, we found that the VisionArray® HPV Chip assay can be recommended for HR-HPV testing in FFPE tissue samples from OPSCC providing both a fast and simultaneous genotyping for 41 clinically relevant HPV types.