A global epidemic increase of an HPV‐induced tonsil and tongue base cancer – potential benefit from a pan‐gender use of HPV vaccine

In 2007, human papillomavirus (HPV) type 16 was finally recognized as a risk factor, besides smoking and alcohol, for oropharyngeal squamous cell carcinoma (OPSCC), including tonsillar squamous cell carcinoma (TSCC), by the International Agency for Research against Cancer. Just before, in 2006, the Food and Drug Administration had approved Gardasil, the first vaccine against HPV16, 18, 6 and 11, for preventive vaccination women against cervical cancer. Concurrently, some Western countries, where smoking was decreasing, disclosed an epidemic increase in the incidence of OPSCC, especially of TSCC and base of tongue cancer (BOTSCC), together accounting for 80–90% of all OPSCCs, and mainly affecting men. The epidemic was later revealed to be due to a rise in HPV‐positive cases, and scientists in the field suggested HPV vaccination also of boys. Globally, there are roughly 96 000 incident OPSCC cases/year of which 20–24% are caused by HPV, thereby accounting for around 22 000 OPSCC cases annually. Of these cases, 80–90% are due to HPV16 infection and would be prevented with the presently registered HPV vaccines. In Western countries, such as Sweden (with almost 400 TSCC and BOTSCC cases per year) and the United States, HPV prevalence in OPSCC is higher and around 70%. HPV vaccination of girls has been initiated in many countries, and the vaccines have been efficient and their side effects limited. HPV vaccination of boys has, however, been the exception, but should definitely not be delayed any further. It would benefit both girls and boys directly, and result in better and more robust herd immunity. Today, we have the possibility to eliminate several high‐risk HPV types in the younger generations and avoid more than 600 000 cancer cases annually worldwide, and this possibility should be embraced by offering global pan‐gender HPV vaccination.


Historic background
The presence of genital warts was described already by the ancient Greeks, and 1760-1839, in Verona, it was reported that cervical cancer was much more common in prostitutes than in nuns [39]. A hundred years later in 1941, Georgios Papanicoloas showed that cellular changes were present in the cervix before invasive cancer appeared [40]. Together, these findings suggested an infective agent being the culprit. However, in 1983 when HPV16 was detected in cervical cancer, this finding was not only complex, due to a plethora of HPV types concurrently being disclosed, but also controversial, by being a challenge to the existing opinion that Herpes simplex virus 2 (HSV-2) was liable for the development of cervical cancer [41,42]. Finally, in 1995, HPV16 was acknowledged as being carcinogenic by IARC [43]. In 2006, the first vaccine against HPV16, 18, 6 and 11 (Gardasil) was approved by the Food and Drug Administration (FDA) and a second vaccine (Cervarix) against HPV16 and 18 was validated in 2007 [44][45][46][47]. Soon after in 2008, Harald zur Hausen was awarded the Nobel Prize. That HPV was associated with OPSCC and TSCC, mainly affecting men, was reported already in 2000, but it took until 2007, until HPV was (besides the traditional risk factors smoking and alcohol) recognized as a risk factor for OPSCC and TSCC by IARC [1][2][3]. By then, reports on an epidemic increase of these diseases had already been observed in many Western countries [6][7][8][9][10][12][13][14][15]. Since 2014, the nonavalent HPV vaccine (Gardasil 9) (against HPV16, 18,31,33,45,52,58, and 6 and 11) has been available, and HPV vaccination of girls is ongoing, but the vaccination of boys has still not yet been initiated in most countries [44][45][46][47].

HPV and its genome
There are >200 HPV types, where most are found in the skin, and where roughly 40 are detected in the mucosa, and with the latter often defined as low-risk, high-risk or putatively high-risk HPV types [48][49][50]. All HPV types have an~8 kb double-stranded circular DNA genome surrounded by a 52-55 nm viral capsid [50]. The HPV genome is arbitrarily divided into a noncoding regulatory region, and an early and a late coding region (coding for the E1-E2, E4-E7 regulatory proteins and the L1 and L2 structural proteins) (Fig. 1) [48,50]. The E1-E2, E4-E7 regulatory proteins are responsible for gene regulation, replication, pathogenesis and transformation [48][49][50]. In high-risk HPVs, E6 and E7 are regarded as oncogenes and deregulate cell cycle control. E6 binds and degrades p53, permitting cells with mutated or damaged DNA to enter the cell cycle without undergoing DNA repair and allowing the cells to harbour more mutations [49,50]. E7 binds to the retinoblastoma protein (Rb) and disrupts its binding to the E2F transcription factor, resulting in E2F activation and progression of the cell cycle into synthesis phase (S-phase) [50]. This leads to overexpression of the cyclin-dependent kinase inhibitor p16 INK4a (p16), often used as surrogate marker of HPV infection [49][50][51][52][53]. E5 is also regarded as an oncogene and is supportive of transformation [54,55]. Furthermore, both E5 and E7 down-regulate major histocompatibility (MHC) class I antigens, thereby enhancing the virus to avoid immune recognition [54][55][56]. The late region encodes the major L1 and minor L2 viral capsid proteins [48][49][50]. L1 contributes with 360 molecules on the viral capsid surface, whilst L2 contributes with 12-20 molecules on the inside of the capsid [48][49][50]. L1 can under specific circumstances self-assemble into virus-like particles (VLPs), which constitute today's HPV vaccines, but also L2 has been suggested to be useful for vaccine production [44][45][46][47]. skin cancer is not resolved, except in patients with epidermodysplasia verruciformis, an autosomal recessive disease, where some HPV infections are not cleared [50]. Mucosal low-risk HPV types, for example HPV6 and 11 can induce genital warts, condylomas and laryngeal papillomas [50]. High-risk HPV types (e.g. HPV16 and 18) are associated with virtually all cervical cancer cases and a considerable proportion of vulvar cancer, vaginal cancer, penile cancer and anal cancer, as well as OPSCC, more specifically TSCC and BOTSCC [50]. Together, HPVpositive tumours account for more than 600 000 cancer cases annually worldwide (Table 1) [50,57,58]. In TSCC and BOTSCC, the proportion of HPVpositive cases varies depending on geographical region, and the time periods when the investigations were performed [14]. HPV16 dominates and is responsible for 80-90% of all HPV-positive cases, whilst HPV18 frequently observed in cervical cancer is rarely found and less common as compared to, for example, HPV33 [6,14,16,48,50].
Methods to detect HPV in HNSCC, with EMPHASIS on OPSCC, TSCC AND BOTSCC HPV DNA is detected in 0-100% in OPSCC, TSCC and BOTSCC [1, 2, 6-10, 12-17, 50]. This variation can depend on, for example, OPSCC subsite, since HPV DNA is mainly detected in TSCC and BOTSCC as compared to other OPSCC subsites [59,60]. In addition, the time period and the country from which the samples were obtained both influence HPV prevalence, due to that in some countries, there has been an epidemic increase in HPV-positive cases and a varying decrease of smoking [6][7][8][9][10][11][12][13][14][15]. The material and the methodology used also influence HPV prevalence. Fresh-frozen and newly formalin-fixed paraffin-embedded (FFPE) tumours are superior to FFPE material stored for decades, since the latter is degraded and suboptimal to test for HPV DNA and RNA [14,61]. Detection of HPV DNA is generally done using FFPE tumour tissue, but attempts have also been made to test for HPV DNA Fig. 1 The double-stranded DNA HPV16 genome is represented by a grey circle annotated with the nucleotide numbers. The positions of the long control region (LCR) and the early genes (E1-7) and late genes (L1 and L2) are also shown. The early and late promoters, P97 and P670, respectively, are indicated by arrows. The main functions and features of the early and late gene products are listed in the table. Published with permission from the publisher from Tommasino. [50] in fine needle aspirate cytology [FNAC] [61,62]. In the 1980s, Southern blot techniques or in situ hybridization was used to assay for HPV DNA, but their sensitivity was generally lower than that of polymerase chain (PCR)-based technology, where the elaboration of different general primers further  [67][68][69]. Still, the presence of HPV DNA cannot determine biological activity of HPV. Here, assaying for E6 and E7 mRNA expression is regarded as the golden standard and can be done by, for example, RT-PCR [61]. The combined presence of HPV DNA and p16 overexpression by immunohistochemistry (IHC) has, however, been shown to be almost as sensitive as the golden standard, and this approach is now used in many laboratories [60]. p16 overexpression (strong diffuse cytoplasmatic and nuclear staining in >70% tumour cells) alone is nevertheless still frequently used as a surrogate marker for HPV, due to its correlation to the presence of HPV [70,71]. This is unfortunate, since 5-10% of the p16 overexpressing TSCC and BOTSCC cases are not HPV DNA-positive and vice versa, and for OPSCC at other subsites, this discrepancy is higher [52,59,60].
HPV-Positive OPSCC, TSCC AND BOTSCC differ from corresponding HPV-negative cancer and have better clinical outcome

HPV-positive OPSCC, TSCC and BOTSCC, and characteristics of patients and their tumours
In 2000, the association between HPV and OPSCC and TSCC and that patients with HPV-positive cancer were younger and had much better clinical outcome than those with corresponding HPV-negative cancer was reported (Fig. 2a) [1,2,13,14]. In 2004, a similar association of HPV to BOTSCC was found [70]. HPV-positive and HPV-negative OPSCC, TSCC and BOTSCC were suggested to be separate entities and have different characteristics, with the former due to HPV infection (often never smokers) and the latter due to smoking and alcohol [1-3, 13, 14]. Similar to cervical cancer, HPVpositive TSCC and BOTSCC generally possessed normal (and not mutated) p53 and overexpressed 16 INK4a contrary to corresponding HPV-negative cases [1][2][3]. HPV-positive TSCC was more frequently aneuploid, less differentiated and had similar to cervical and vulvar cancer chromosome 3q amplification, and independent of stage, differentiation or ploidy, patients had better prognosis than those with HPV-negative tumours [1][2][3][72][73][74][75]. HPV was often episomal in the tumours, but whether the genome was integrated or episomal did not correlate to prognosis, whilst having a high viral load was a prognostic favourable factor [74,75]. Viral genome integration into cancer cell genomes could, however, affect DNA methylation and viral and host gene expression [76]. Furthermore, when defining HPV-positive status only by the presence of HPV DNA, or only by p16 overexpression, never smokers with HPV-positive cancer had better clinical outcome than smokers [16,52]. Later, defining HPV-positive cancer by the combined presence of HPV DNA and p16 overexpression, the influence on smoking on outcome was not as prominent [19]. Still, similar to cervical cancer, it was later disclosed that smoking also increased the risk of acquiring an HPV-positive OPSCC [77].
There are important differences between OPSCC sites and tumour location matters. In a metaanalysis, covering >50 studies, HPV was more frequently found in TSCC and BOTSCC compared to other OPSCC sites (56% vs. 19%, respectively) [60]. This is also reflected by the epidemic increase of HPV-positive TSCC and BOTSCC, and not of other OPSCC subsites, and that HPV is a prognostic factor for TSCC and BOTSCC, and not for other OPSCC subsites (Fig. 2) [59,78]. Notably, TSCC and BOTSCC arise in the tonsil and the base of tongue, two subsites that share a distinct histological appearance with a reticulated epithelium that invaginates into lymphatic tissue and forms crypts ( Fig. 3) [60]. The importance of lymphatic tissue is also emphasized in that HPV-positive status occurs more frequently in specified, than nonspecified tonsillar tissue, and has better prognostic significance in specified TSCC as compared to that in nonspecified TSCC [79,80].
The Union for International Cancer Control 8th edition (UICC8) staging of OPSCC an improvement, but with some disadvantages p16 overexpressing and p16 non-overexpressing OPSCC are classified separately in the Union for International Cancer Control 8th edition (UICC8) [81,82]. In the earlier UICC 7th edition, all OPSCC were staged together, resulting in that small primary HPV-positive OPSCC with lymph node metastasis were classified as high-stage tumours, although their prognosis was better compared to that of corresponding stages of HPV-negative OPSCC [81,82]. Despite p16-overexpressing OPSCC and p16 non-overexpressing OPSCC are classified separately in UICC8, treatment of these patients in the Nordic countries has not changed, partly due to the ambiguity of OPSCC subsite and the definition of HPV-positive status. The new staging system does namely not use TSCC and BOTSCC location, and optimal determination of HPV-positive status, since it includes all OPSCC and only evaluates p16 overexpression [81,82]. This results in the risk of de-escalating treatment for a considerable number of OPSCC patients with a higher risk of disease relapse [59,60]. In our opinion, the next UICC staging edition should classify only TSCC and BOTSCC, and not all OPSCC, and define HPV-positive status as tumours expressing either HPV E6 and/or E7 mRNA, or displaying presence of HPV DNA combined with p16 overexpression.
HPV in cancer of unknown primary (CUP) of the head and neck region and similarities with HPV-positive TSCC and BOTSCC In the above context, some information on cancer of unknown primary (CUP) of the head and neck region could be relevant. CUP of the head and neck region has better outcome (35-50% 5-year survival) than CUP in general, with a mostly very dismal outcome [83][84][85]. Moreover, location of a squamous cell carcinoma (SCC) metastasis in level I or II of the neck can be due to an OPSCC [86]. The prevalence of HPV in CUP of the head-neck region was reported 18-52% depending on, for example, if tonsillectomy or biopsies were, or were not done, as well as geographical region, since HPV prevalence in TSCC and BOTSCC (OPSCC) can vary considerably between different countries [86][87][88][89][90]. Notably, patients with HPV-positive (HPV DNA and p16 overexpression) CUP of the head-neck region tend to have a better clinical outcome than those with corresponding HPV-negative CUP, with some data reaching statistical significance (80.0% vs. 36.7% 5-year overall survival (OS), respectively, P = 0.004) [87,89,90]. Moreover, assessing HPV status in FNAC in head-neck masses indicated that an HPV-positive lymph node metastasis likely had an HPV-positive OPSCC, TSCC or BOTSCC origin, irrespective if the primary tumour was disclosed or not [62,[91][92][93][94][95][96]. Patients with HPVpositive CUP with SSC cytology should therefore likely receive similar treatment as those with HPVpositive TSCC and BOTSCC and be spared neck dissection and receive more limited radiation fields.
HPV in OPSCC, TSCC and BOTSCC and an epidemic increase of these two latter tumour sub-types An increased incidence of TSCC/BOTSCC and OPSCC and the association with HPV Soon after the first reports of HPV being causative of OPSCC and TSCC, several studies from, for example, Scotland, Sweden, Finland, the Netherlands, the United Kingdom and the United States, reported an increased incidence of OPSCC, especially TSCC [4][5][6][7][8][9][10][11][12][13][14][15]. The reason for this was initially obscure and caused considerable confusion, especially since the rise in incidence was occurring in countries, where smoking was declining. Furthermore, HPV prevalence in OPSCC, TSCC and BOTSCC varied between countries depending on, for example, the proportion of smokers in that country, the quality of the material, the methodology used and the time period at which the analysed samples were acquired [6][7][8][9][10][12][13][14][15]. Eventually, it was revealed that some of the variation especially within countries, where an increased incidence of TSCC and BOTSCC was described, was due to an epidemic rise of HPV-positive cases [6][7][8][9][10][12][13][14][15]. A report from Sweden by us was likely the first in its kind [6]. In 2006, using the Swedish Cancer Registry, which covers virtually all cancer cases in Sweden since 1953, we disclosed an increase in the incidence of TSCC (2.6-fold for men and 3.5-fold for women) for 1970-2002 in the Stockholm region, where around 25% of all Swedish TSCC patients are treated [6]. Moreover, when analysing 237 obtained TSCC samples from the 515 reported patients during the above period, a parallel rise in the percentage (23% to 68%) of HPV-positive TSCC was observed [6]. Similar developments were later reported from, for example, the United States, Greece and Denmark [13,15,79,[97][98][99]. In addition, in 2009 and 2010, studies from Sweden on TSCC and BOTSCC, respectively, and in 2011, on OPSCC from the United States, confirmed the increased incidences of HPV-positive cases and decreased incidences of HPV-negative cases, where the latter was ascribed to a decline in smoking (Fig. 4) [10,12,15].

Reasons for the increased incidence of HPV-positive cases
The reason for the rise in incidence of HPV-positive OPSCC was in due course assigned to changes in sexual habits, since a relation between HPV-positive OPSCC, early sex debut and numbers of oral or vaginal partners was observed [100]. The fact that HPV-positive OPSCC, TSCC and BOTSCC were all more frequently occurring in men than in women was enigmatic and has not entirely been resolved yet. Still, when analysing different factors, men generally had more lifetime sexual partners, which could be one contributing factor, whilst smoking also more frequent in menat least in some countriescould be another [101]. Smoking has namely, as mentioned above, been shown to increase the risk of acquiring an HPV-positive OPSCC [77]. It has also been hypothesized that women develop a better immune response than men, after a corresponding genital HPV infection, and are thereby more resistant to developing an OPSCC [102]. That the increased incidences of TSCC and BOTSCC were due to changes in the numbers of tonsillectomies per decade or to different ways of removing the tonsils has, however, been excluded [103,104]. Irrespective of the reason for the observed epidemic increase of OPSCC, TSCC and BOTSCC, if these trends continue, it has been calculated that the number of cases in these categories will surpass the numbers of cervical cancer cases in, for example, both the United States and Denmark (Fig. 5) [15,99].

Reflections regarding HPV tumour location, the epidemic and impact on survival
In the context of an epidemic due to HPV infection, emphasis on HPV tumour location and whether HPV has impact on outcome is of importance [78][79][80]. It must be highlighted that HPV is mainly found in TSCC and BOTSCC and not in other OPSCC subsites, and the rises in incidence of OPSCC are attributed to increases in TSCC (i.e. specified TSCC) and BOTSCC [10,12,60,[78][79][80]. The influence of HPV on survival is as stated above, limited to TSCC and BOTSCC, and not OPSCC at other subsites (Fig. 2) [59]. Historically, due to an original report on HPV and OPSCC, and because TSCC and BOTSCC account for 80-90% of all OPSCC, the distinction between OPSCC, and TSCC and BOTSCC is not always made [1,79,80]. To cover the scientific field, one has to follow reports on OPSCC, TSCC and BOTSCC. In the future, as mentioned above, being more specific, and using specified TSCC and BOTSCC location, with an improved definition of HPV-positive status, will likely be of benefit for diagnosis and prognostication, patient treatment, survival and quality of life.
New trends in the epidemic and changes in the age of the patients In general, there is still an increase in incidence of HPV-positive OPSCC, TSCC and BOTSCC in many countries, but some variations have been reported, both with regard to the age groups affected, and as to whether one studies urban or rural areas [14,15,79,98,99,[105][106][107][108][109][110][111] Following the incidences of TSCC and BOTSCC in Stockholm, more specifically, for the years 2000-2012, we noted a stabilization of the increased incidence of TSCC in Stockholm for some years, but not for BOTSCC, or for TSCC and BOTSCC in Sweden as a whole [105]. Possible differences in trends of OPSCC, TSCC and BOTSCC depending on whether one is covering rural or urban areas were also recorded by others, but the data were not concordant over time, which is very likely due to the complexity of geographical areas [106,107,110]. In some cases, rural areas had more HPVassociated cancers than urban areas, and in some cases, the nonmetropolitan counties initially had less HPV-associated TSCC and BOTSCC, but encountered increased incidences later in time [106,107,110].
The incidences of TSCC and BOTSCC were followed in more detail for the period 2000-2016 in Stockholm, and we noted they had continued to rise especially between 2008 and 2016 [108]. The less steep increased incidence curves between the earlier time periods 2000-2004 and 2005-2008, we assume may have been due to changes in sexual behaviour in the early 1990s due to the HIV epidemic. This assumption was supported by a documented decrease in the number of chlamydia cases in the Stockholm region from 1988, with a major dip 1992-1998, just before the introduction of HIV medication, a period followed by an increase of Chlamydia cases [112]. This would hypothetically indicate a lag period of roughly 20-30 years between HPV infection and the development of TSCC. Notably, it has also been observed that in both the United States and in Stockholm, the median age for patients with HPV-positive OPSCC, TSCC and BOTSCC has gone up over the years [6,10,105,108]. In our studies from Stockholm, during the period 1970-2002, the median age of patients with HPV-positive cancer was 55 years of age as compared to those with HPV-negative cancer, which was 65 years of age [6]. In a later study in Stockholm, patients with HPV-positive cancer diagnosed 2013-2016 were on average 61.2 years of age and those with HPV-negative cancer were 66.7 years of age [108]. We suspect that we are encountering older patients from a cohort that was infected approximately 30 years ago. However, the numbers of our patients are limited and one should not draw too many conclusions. Nevertheless, in a very recent study on different populations in the United States, the authors suggest that the rise in incidence is increasing in older men over 65 years of age and that in younger men the increase in incidence has ebbed for the time being [109]. The reason for the decrease in younger men was suggested to be due to a decrease of sexual activity in this group, as indicated by a decline in this group of the seroprevalence of HSV-2, a surrogate for high-risk sexual behaviour [113]. This should be important information with regard to the planning of treatment of an older cohort [22,109]. On the other hand, trends in Stockholm and with increases in Chlamydia infections, since 1998, suggest a more recent increase in sexual activity and may point towards the risk for a new sharp rise in incidence of TSCC and BOTSCC in younger men within the next decade or two [108]. It is also likely that the same shifts will be seen also in other countries.

Attempts to screen for OPSCC, TSCC AND BOTSCC
Today, there is no screening for OPSCC, TSCC and BOTSCC, since in contrast to cervical cancer corresponding prestages including cellular abnormalities are not readily observed in tumour patients [114,115]. The presence of HPV DNA in mouthwashes as indicative of HPV infection, or HPV-positive TSCC and BOTSCC was also explored [115][116][117][118][119][120]. HPV prevalence in mouthwashes varied between 1% and 20.7% in healthy individuals, sometimes with and sometimes without gender differences [115][116][117][119][120][121][122]. The quantity of HPV DNA in mouthwashes was generally lower than that obtained in cervical samples, likely due to the daily production of 0.5-1.5 litre of saliva, and subsequently, it is more complicated to identify an HPV infection in the oral cavity as compared to the cervix [115,120]. When examining HPV prevalence in tonsillar swabs and mouthwashes from patients with TSCC, BOTSCC and HNSCC and benign conditions, an HPV-positive mouthwash was indicative of an HPV-positive TSCC or BOTSCC [115]. More specifically, HPV DNA was present in 82% and 50%, respectively, of mouthwashes from respective TSCC and BOTSCC patients, as compared to 14% in those with other HNSCC and benign conditions [115]. The HPV DNA signal was generally stronger in mouthwashes from TSCC and BOTSCC patients compared to that in healthy youth, but still often weaker than that in cervical samples [115,120]. HPV prevalence in mouthwashes was also studied in patients undergoing tonsillectomy, but there was no association between presence of HPV in their oral samples and in the tonsillectomies, where HPV was not found [121][122][123]. Antibodies to HPV16 E6 and E7 have, however, been observed in patient sera 10 years prior to the development of OPSCC, but so far there is no optimal strategy for using this approach for individual screening [124,125].

Treatment OF OPSCC, TSCC AND BOTSCC
Patients with HPV-positive TSCC and BOTSCC often have nodal spread and generally seek help when their tumours are spread, and are therefore mainly given the aggressive treatment offered to other HNSCC patients [13][14][15][16][17][18][19][20][21][22]. Radiotherapy (RT) is delivered at high doses for 6-7 weeks. Patients with advanced stage (III-IV), according to UICC7, are treated with chemotherapy (CT) or chemoradiotherapy (CRT) and sometimes epidermal growth factor receptor (EGFR) blockers, for example cetuximab [20]. Treatment is, however, not only determined by the tumour burden, but also by whether the patients are fit medically [20]. Intensified therapy may result in complications with regard to swallowing and eating due to radiation, as well as nausea, mucositis and local and/or systemic infections and fatigue. Patients with lymph node metastases persisting after RT are managed with neck dissection. CRT and surgery cause more side effects than RT alone, with more fibrosis, stiffness of the neck, less mobility and worsening of the possibility to swallow due to RT. Later adverse effects are, for example, xerostomia, taste alterations, swallowing problems, trismus and diminishing of hearing, and some present radioosteonecrosis needing reconstruction surgery.
Prognostication focusing on quality of life and not only survival is essential for the patients. Notably, in parallel, to therapy intensification, patient numbers with HPV-positive TSCC and BOTSCC are growing, with patients of varying ages, and most do not need aggressive therapy [126]. However, to personalize medicine, better means to identify patients that will respond well clinically or could benefit by targeted therapy are needed [126]. There are clinical trials decreasing RT, or replacing cisplatin by cetuximab, using p16 overexpression as surrogate marker for HPV in OPSCC [21]. These trials allow for suboptimal selection of the patients included in the trials, because 10-15% of their patients with a p16-positive cancer will not have a truly HPV-driven cancer, and the data obtained can be false, since recurrences are rare.
Other studies have investigated RNA expression. MicroRNA expression has been explored in OPSCC, TSCC and BOTSCC with regard to both HPV and outcome, but the obtained data with regard to their prognostic value have been inconsistent between studies [134][135][136][137][138][139][140]153]. RNA expression according to HPV status and also heterogeneity in transcription between HPV-positive tumours and outcome have been described [141][142][143][144][145]. Using transcriptional analysis, CD8a was notably correlated to HPV status and superior clinical outcome [145]. Reverse-phase protein array profiling on a limited set of total and phosphorylated proteins was attempted and the proteomic analysis disclosed differences in, for example, PI3K/serine-threonine protein kinase (AKT)/mammalian target of rapamycin (mTOR) [133]. Likewise, proteomic profiling by mass spectrometry showed differences in pathways in HPV-positive and HPV-negative OPSCC [132].
Research on the microbiome and HPV infection is still limited. Significant variations in microbiota have been reported in oral squamous cell carcinoma (OSCC) patients [154]. However, no specific species were showed consistently to be associated with OSCC. Such data in other sites of HNSCC are very preliminary and need to be pursued further.
The HNSCC-related shifts in the variations of the microbiome may allow for an alternative potential explanation for HNSCC progression as well as a therapeutic potential.

Background and present HPV vaccines
Already in 1978, it was shown that purified major capsid protein VP1 of murine polyomavirus (previously belonging to the papovavirus family, where HPV was also included in before) could, upon renaturation, form pentameric subunits, which could assemble into VLPs [155]. Knowledge regarding the potential possibility to form HPV VLPs was therefore available early on. However, large-scale production was not resolved at the time, despite that it was known that VP1 or L1 pentameric subunits could be produced in mammal and yeast cells, insect cells using baculoviruses, Escherichia coli or tobacco chloroplasts [156][157][158][159][160]. The choice finally depended on the ease of production, safety and economical issues. The present HPV VLP vaccines have been established on an industrial scale in yeast (Gardasil â , Merck, NJ, USA) and baculovirus (Cervarix â , GlaxoSmithKline, UK). It is more than a decade that the tetravalent Gardasil vaccine against HPV 16, 18, 6 and 11 and Cervarix directed against HPV16 and 18 were FDA-approved, and this has in 2014 been followed by approval of the nonavalent Gardasil 9 vaccine against HPV16, 18, 31, 33, 45, 6 and 11 [42, 44-47]. All vaccines induce high titres of neutralizing antibodies that protect against HPV infection with the corresponding HPV types, included in the respective vaccines [44][45][46][47][161][162][163][164][165].
Several studies have since then examined the potential effects of reducing the numbers of doses given, and within which time frames they should be offered in order to prevent prestages of cervical cancer [44][45][46][47]. The World Health Organization (WHO) has presently recommended two doses for either Gardasil, Gardasil 9 or Cervarix for those up to 15 years of age and three doses for women 15 years or older [44]. Different doses are not specified specifically for males according to age, but one should assume that the recommended doses should be similar to those offered to women and depend on the age of the individual. A number of high-income countries have introduced these vaccines quite successfully, but from a global perspective, the vaccination of young girls is still very low and the uptake of HPV vaccinations across the world was in 2017 <2% in females 9-45 years of age [44]. More details regarding the HPV vaccines, mainly from high-income countries, are presented below.
HPV vaccine efficacy against cervical cancer and other HPVassociated tumours All three vaccines have been produced to inhibit infection with HPV types involved in cervical cancer, affecting more than 500 000 women every year, and the end-point of the vaccines, which was prevention of prestages to cervical cancer, was efficiently accomplished [42,[44][45][46][47][161][162][163][164][165]. These three vaccines can thereby inhibit roughly 70-90% of all cervical cancer cases, but they can also potentially also prevent other associated cancers such as other anogenital cancer as well as HPV-positive TSCC and BOTSCC [42,47,[166][167][168][169][170][171][172]. The two Gardasil vaccines can, in addition, also abrogate a substantial number of condylomas and papillomas [44][45][46][47]172]. In the Future I/II study, prestages of vulvar and vaginal cancer as well as anogenital warts were prevented after HPV vaccination [166]. Furthermore, early experimental models, in the related canine oral papillomavirus (COPV), demonstrated that it was possible to protect against the development of oral cancer through COPV L1 VLP vaccination [167]. In addition, the preliminary utilities of HPV vaccination for prevention of oral HPV infection, as well as the acquisition of antibodies in the saliva, have been demonstrated [168][169][170][171]. Notably, since HPV16 is present in roughly 80-90% of all HPV-positive OPSCC, TSCC and BOTSCC, and all today's vaccines are directed against HPV16, they should all be extremely efficient to prevent the development of these tumours [6,14,16,48,50].

Annual global contribution of HPV to cancer and benefits with the introduction of the HPV vaccine
The global attribution of HPV to cancer is 630 000 cases per year according to Globocan 2012 [57]. Of these, 570 000 affect women and 60 000 affect men, and for details of these cancer cases please, see Table 1 [57,58]. Cervical cancer followed by anal cancer dominates in women, whilst OPSCC followed by penile cancer dominates in men. The relative contribution of HPV16/18 and HPV6/11/ 16/18/31/33/45/52/58 was estimated with 73% and 90%, respectively. Global HPV vaccination would within decades result in avoiding most HPV-attributable cancer cases [57,173,174]. In fact, today we have the means within a few generations to wipe out a great fraction of HPV-associated cancer cases worldwide. This would decrease the need of screening, as well as the medical treatments necessary to take care of all these cancer patients and take a considerable burden of the health system. Furthermore, it would save the lives of relatively young people that contribute socio-economically.

Vaccination coverage today and different hurdles
Since 2007, almost 100 countries and territories have introduced HPV vaccination programmes and approximately 270 million doses have been distributed; however, in the vast majority of these countries, vaccination is only offered to girls [172][173][174]. Very few side effects have been documented, and these have been limited to local reactions [173]. Some countries have a high HPV vaccine coverage (≥80%) such as Australia of young girls and boys, and the UK and Sweden for girls [174][175][176]. Others like Denmark and Japan have dropped from a high coverage to <40% and <1%, respectively, due to different unconfirmed worrisome negative reports on, for example, side effects regarding the HPV vaccines [177,178]. A low socioeconomic background is often associated with lower vaccine uptake [174]. In Norway, a lower vaccine initiation was observed in girls, with mothers with lower education and in the lowest income bracket [179]. Likewise, in Denmark, a lower HPV vaccine introduction and completion was found in immigrant girls, where much, but not all, could also be due to socio-economic differences [180]. Ethnicity and cultural norms also affected HPV vaccination [180,181].
In spite of the documented successful effect of these vaccines, their worldwide presentation has been hampered not only due to economic resources, or to concern for potential physical side effects or efficiency of the vaccines, but also due to anxiety including risk for changes in the behaviour of vaccinated youth [182][183][184][185][186][187][188][189]. It was suggested that the vaccines would allow youth to lose their moral and become sexually adventurous, and that HPV vaccination should not be necessary for youth with higher principles [186,187].
In addition, youth have limited knowledge about HPV and the link between sexual behaviour and HPV-associated cancer, and in general boys know less than girls [182]. Boys often believed HPV only affected girls and would accept being vaccinated for the sake of girls, but upon being aware that an HPV infection also could result in a cancer affecting them, their awareness and will to be vaccinated have increased [190]. To summarize, the introduction of HPV vaccination has clearly differed in many respects from the establishment of other vaccines [172][173][174][182][183][184][185][186][187][188][189].

Herd immunity and gender-neutral HPV vaccination
Unfortunately, currently a large part of the human population is still sceptical to many types of vaccines, and for HPV vaccination, there is an even greater reluctance to accept its need from a moral point of view [181,183]. This has together with socio-economic reasons, ethnicity and cultural norms in some respects hindered the efficient introduction of this vaccine to young girls (and boys) in order to reach stable high vaccine coverage and herd immunity that could spill off to also protecting young boys [174,[191][192][193][194][195]. Instead, when scientific reports or other media of dubious quality are spread, immediate serious dips in the compliance to HPV vaccination may occur as mentioned above in, for example, Denmark and Japan [177,178]. This immediately abolishes the possibility to achieve a ≥ 80% HPV vaccine coverage amongst girls and efficient herd immunity for unvaccinated girls and boys [177,178]. The efficacy of HPV vaccines by using gender-neutral vaccination is well documented [191][192][193][194][195]. Furthermore, when setting the horizon to 100 years, this was also the case when having a vaccine coverage ≥80% amongst girls [194]. Gender-neutral HPV vaccination would benefit both girls and boys directly, and also be more robust in that it also can reduce the effects of potential acutely occurring dips in HPV vaccination coverage, which in turn wipe out the prospect of herd immunity [177,178]. As mentioned above, some surveys have been conducted about adolescent boys' attitudes to HPV vaccinations. Many have focused on male college students, or men who have sex with men, but also young boys' knowledge and beliefs regarding HPV, as well as HPV vaccination, have been studied [190,[196][197][198][199]. With increasing information, present knowledge suggests that also boys want to be vaccinated [190].
In some countries, for example Australia and the United States, gender-neutral vaccination has been introduced, and recently, additional some European countries, for example Denmark and the UK, have decided to follow this policy [172].
Including boys in the national vaccination programme is an approach towards which the Public Health Agency of Sweden is also in favour, but unfortunately so far this has not yet been applied [195]. We believe in the long run, it is of great benefit for both health care and government to vaccine both genders. Finally, last but not least, one should recollect, that importantly, it is amongst human rights to be treated equally [200].

Conclusion
Patients with HPV-positive TSCC and BOTSCC have much better prognosis than those with corresponding HPV-negative cancers, and the incidence of HPV-positive cancers has been increasing epidemically the past decades. Tailoring therapy for this growing group of patients is crucial, and efforts in finding markers useful for tapering and targeting therapy are on the way and will hopefully be of benefit for some patients. For the younger generations, however, it is feasible to inhibit most HPV-associated tumours, and for this purpose, the introduction of pan-gender HPV vaccination is urgently encouraged.