Abstract
Human papillomavirus (HPV) is recognized as a risk factor for oropharyngeal squamous cell carcinomas (SCC), especially tonsillar and base of tongue cancer. Furthermore, HPV-positive tonsillar and base of tongue SCC have a significantly better prognosis than their HPV-negative counterparts and head and neck cancer (HNSCC) in general. HPV has recently also been implicated in cancer of unknown primary (CUP) in the head and neck region, where a primary tumour is not found despite extensive workup. Using fine-needle aspiration cytology to determine CUP HPV status in cervical lymph nodes could be of advantage, since it is minimally invasive and it is assumed that an HPV-positive lymph node metastasis likely has an HPV-positive otopharyngeal SCC origin. We review the current knowledge of HPV in HNSCC, with an emphasis on CUP of the head and neck region, its relation to oropharyngeal, tonsillar and base of tongue SCC and implications of HPV status for diagnosis, prognosis and treatment.
- HPV
- cancer of unknown primary
- head and neck cancer
- oropharyngeal cancer
- tonsillar cancer
- base of tongue cancer
- fine-needle aspiration cytology
- review
Infections are estimated to cause around 15-20% of all human cancers worldwide (1). Although certain bacteria (e.g. Helicobacter pylori) have been implicated in tumour development (2), viruses are by far the most frequent pathogens associated with cancer (1). Today the list includes Epstein–Barr virus, hepatitis B and C virus, human herpes virus 8, human T-lymphotropic virus 1 and Merkel cell polyomavirus (1). However, most frequently involved in cancer development is the human papillomavirus (HPV) family, where certain types can cause cervical, vaginal, vulvar, penile and anal cancer, as well as specific subsets of head and neck squamous cell carcinoma (HNSCC) (1, 3, 4). The most common types causing cervical cancer are HPV16 and 18, while HPV16 alone accounts for >90% of HNSCC (3, 4). HNSCC comprises of squamous cell carcinomas (SSC) in the oral cavity, the oropharynx, the nasopharynx, the hypopharynx, the larynx, the salivary glands, the lips and the nasal and sinus cavities, as well as cancer of unknown primaries (CUP) in the area (5). Worldwide, head and neck cancer is the sixth most common type of cancer, with ~600,000 cases/year and a mortality rate of ~50% (5). In 2007, besides smoking and alcohol consumption, established risk factors for HNSCC, HPV16 was also acknowledged as a risk factor for oropharyngeal SCC, by the International Agency for Research on Cancer (3). The most frequent HPV-driven oropharyngeal SCC sites, are the tonsils and base of the tongue (6-8). There is also some indication for a minority of HPV-driven carcinomas in the hypopharynx, but most notably during the past decade there is accumulating evidence for HPV and p16 overexpression in a considerable proportion of CUP in the head and neck region (9-24). We review current knowledge of HPV in HNSCC, with emphasis on CUP of the head and neck region and implications of HPV status for diagnosis, prognosis and treatment (11-30).
The HPV Family
HPVs are small and non-enveloped with a double-stranded, circular DNA genome of approximately 8,000 base pairs (4). The genome encodes the early proteins E1, E2, E4-E7 involved in various functions of the viral life cycle, and the capsid proteins L1 and L2 (4). Viruses can cause cancer in three principal ways: through integration of the viral genome into the host genome, through interaction with host proteins, and through chronic inflammation (1). Although HPV often integrates into the host genome (4), viral protein interactions with host proteins are regarded as very important for HPV-induced carcinogenesis, such as the E6 protein binding to p53, leading to its degradation, while E7 binds to and inactivates the retinoblastoma protein (pRb) (4). This leads to a markedly decreased half-life for two of the most important tumour-suppressor proteins, driving the cells into S-phase and triggering overexpression of the cyclin-dependent kinase inhibitor p16INK4a, that is often used as a surrogate marker for HPV infection (8).
HPV in Oropharyngeal, Tonsillar and Base of Tongue Cancer and Consequences for Treatment
HPV-positive oropharyngeal SCC, especially tonsillar and base of tongue, appear to be different clinical and biological entities compared to their corresponding HPV-negative counterparts. HPV-positive tumours more often affect younger people and non-smokers than do HPV-negative tumours (7, 8, 31-34). Most notable, however, is the fact that patients with HPV-positive tonsillar and base of tongue SCC have significantly better clinical outcome than those with HPV-negative tumours, with e.g. approximately 85% versus 45% 5-year disease-free survival, respectively (7, 8, 31-35).
Furthermore, in the past decades, a sharp rise in the incidence of tonsillar and base of tongue SCC, in parallel with an increase of the proportion of HPV-positive cases has been observed in many countries, although the percentage of HPV-positive cases varies between countries (31-41). This analogous increase had lead to the hypothesis that an HPV epidemic is causing the rise in tonsillar and base of tongue SCC incidence (8, 31, 37, 42). In countries where smoking has decreased in recent decades, e.g. the US and Sweden, the proportion of patients with HPV-positive tonsillar and base of tongue SCC has increased among the total number of HNSCC cases (8).
However, due to the poor prognosis for HNSCC in general, in recent years treatment has become intensified with the addition of chemotherapy and targeted therapy to the previously administered radiotherapy, thus increasing adverse effects for patients (8, 41). Markedly, patients with HPV-positive tonsillar and base of tongue SCC have a considerably better prognosis than HNSCC in general and it has been proposed that a de-escalation of treatment for some patients with HPV-positive tumours might be possible in the future (8, 41). Nevertheless, before this can be tested in clinical trials, additional biomarkers are required to identify patients who would benefit from less treatment, since 20% of those in the HPV-positive group still succumb to their disease (8, 34, 35). Recently, several studies were conducted to find useful biomarkers, using e.g. immunohistochemistry or molecular methods, with CD8+ tumour-infiltrating lymphocyte counts (by immunohistochemistry) and expression of HPV16 E2 mRNA being especially promising markers (43-47). Moreover, interest has been focused on examining for the presence and influence of HPV for prognosis in other HNSCCs, especially, CUP of the head and neck region.
Diagnosis and Treatment of Head and Neck Cancer with Emphasis on CUP
In head and neck cancer, the first sign of disease which brings the patient to the doctor is quite frequently a lump in the neck, i.e. an enlarged lymph node. The subsequent diagnostic workup includes fine-needle aspiration cytology (FNAC), computed tomography and biopsies for a pathological anatomical diagnosis if a primary tumour is evident at first examination (13). When cytology from a cervical lymph node SCC metastasis is HPV-positive, the primary tumour is likely to be an HPV-positive oropharyngeal SCC, which is why HPV analysis of lymph node metastases may be helpful in guiding further diagnostic procedures (12, 13). If no apparent primary is recognised, a panendoscopy with tonsillectomy, and biopsies from the nasopharynx and base of tongue, is applied to detect the primary tumour.
In 3-5% of cases with lymph node metastasis, a primary tumour is not found despite a thorough clinical workup and the diagnosis will be a CUP (13). If the FNAC shows non-SCC, e.g. adenocarcinoma, positron-emission-tomograpy/computed tomography is recommended since the primary tumour is likely situated outside of the head and neck region, while presence of SCC suggests that a tumour of HNSCC origin is more likely (13).
Treatment for CUP used to comprise, and at some centres still includes, a neck dissection to confirm the cytological diagnosis, and as part of the treatment (13). In addition, postoperative full-dose radiotherapy was administered to large areas of the head and neck region, and in advanced cases, sometimes with concomitant chemotherapy, most often cisplatin-based (13, 48). The adverse effects of radiotherapy can be severe and include mucositis, with subsequent severe pain, dysphagia and nutritional difficulties, as well as skin reactions, xerostomia, osteoradionecrosis and hypothyreosis (13, 48). For surgery, the side-effects include risk of lymph oedema, nerve damage, pain, arm and shoulder function deficits, as well as cosmetic issues, especially if the sternocleidomastoid muscle is included. Combined therapy, i.e. both surgery and radiotherapy, with few exceptions, leaves the patient with much more pronounced side-effects than monotherapy does.
It has long been suggested that CUP presenting at different parts of the body share common distinct biological properties leading to rapid progression and early metastasis (27). However, this view has been challenged in recent years, since increasing evidence points to CUP as being different site-specific metastases retaining the properties of the primary tumour and sharing the attribute of arising from a primary tumour that escapes recognition (27).
CUP of the head and neck region appears to be a distinct subtype of CUP, with a clinical picture more resembling advanced HNSCC than CUP in general. For instance, CUP of the head and neck region has a remarkably better prognosis, with a 5-year survival of 35-50% (28), compared to CUP in general, which has a dismal prognosis with a median survival of only 6 to 10 months (29). In Sweden, its incidence rate was approximately one case cases per 200,000 inhabitants per year between 2008 and 2012 (48). It is more common in men (73%) than in women, in line with other types of head and neck cancer, and SCC is the most common form, accounting for 86% of cases (48). Furthermore, location of an SCC metastasis in level I or II of the neck gives rise to suspicion of oropharyngeal SCC, while a metastasis located in the lower part of the neck more likely has a non-head and neck cancer as a primary (48).
Reasons for not finding the primary tumour include that it may be very small and therefore missed both during clinical examination and by the pathologist investigating the surgical tonsillar specimens (13, 14). This is likely true case in many cases of metastases from oropharyngeal SCC, especially HPV-positive tonsillar and base of tongue SCC, which often arise in the miniscule crypts of the palatine and lingual tonsils, respectively (15). Furthermore, it could also be due to spontaneous regression of a primary tumour or that the primary tumour has already been removed, e.g. a skin lesion interpreted as a benign nevus (14).
Today oropharyngeal SSC, also with regional metastasis, at many centres is treated only with radiotherapy with or without chemotherapy, and neck dissection is used only in the case of residual disease or recurrence. A clear cytological SCC diagnosis from a CUP together with a reliable HPV status obtained by FNAC would indicate that the primary tumour is likely an undetectable HPV-positive tonsillar or base of tongue SCC, and treatment could be planned accordingly (12, 13, 32-35, 40). Subsequently, the patient would be spared neck dissection, as well as the adverse effects associated with combined therapy (41, 49).
HPV Prevalence in CUP of the Head and Neck Region
A number of very recent studies have documented the prevalence of HPV DNA, p16, or both in CUP of the head and neck region, although with varying results (11-30). The foremost reason for the discrepant results is the inconsistent definition of CUP and the extent of the clinical work-up before diagnosis. Studies with very high prevalence (59-92%) included patients who had not been subjected to biopsies of the tonsils and base of tongue nor had been tonsillectomised (16-18). Tonsillectomy has particularly been shown to be of great value in detecting the primary tumour. If applying tonsillectomy, which today is routinely used and should be included before the CUP diagnosis can be considered, the prevalence of HPV is lower (28-52%), as reported in studies conducted in the U.S., Germany and Sweden (by our group) (13, 19, 21). In a Danish study where blind biopsies were included in the work-up, but not tonsillectomy, the prevalence of HPV DNA-positive/p16-positive tumours was relatively low, 18% (22). In two other studies, that included patients both with and without having undergone tonsillectomy, the HPV prevalence was in between the above-mentioned studies (23, 25). The variation in HPV prevalence was not surprising, since there are also differences in HPV prevalence between different countries for oropharyngeal SCC, as well as depending on whether patients have undergone tonsillectomy or not (36-40).
HPV in CUP of the Head and Neck Region and Prognosis
Since HPV-positive CUP of the head and neck region likely originates from HPV-positive tonsillar and base of tongue SCC, which have better clinical outcome than their HPV-negative counterparts, it is likely that HPV-positive CUP may also have a better prognosis than HPV-negative CUP. Indeed, in an earlier study by our group, patients with HPV-positive HNSCC CUP had a significantly better 5-year overall survival compared to patients with HPV-negative CUP (80.0% vs. 36.7%, respectively, p=0.004) and this was confirmed in a multivariate analysis (13). Similar numbers were obtained whether HPV DNA, p16 or a combination of both were used to determine HPV status (12). Another report showed very similar, significant, survival benefits from HPV-positive or 16-positive status in the metastases (22). However, other studies have failed to show a significant survival benefit for patients with HPV-positive metastases in CUP of the head and neck region, although a trend towards better prognosis could be seen in all of them (20-22). It is likely that statistical significance would have been reached in these studies if more patients had been included, but this is difficult to achieve for such a rare entity as CUP in the head and neck region.
Other studies including both patients who had, and who had not undergone tonsillectomy before CUP diagnosis showed similar results. In one report, significantly better 4-year overall survival was demonstrated for patients with HPV DNA-positive or p16-positive CUP at a univariate level, while p16 also reached significance at the multivariate level for disease-free survival (23). In another report, it was shown that having a p16-positive lymph node was significantly correlated to improved overall survival at the multivariate level (25). A few other smaller studies have shown similar trends, but failed to reach statistical significance (22, 30)
Consequently, patients with HPV-positive CUP in the head and neck region appear to have a better prognosis than patients with HPV-negative CUP. They could therefore potentially, in analogy with HPV-positive oropharyngeal SCC, be subjected to prospective studies for de-escalation of treatment in the future, although additional biomarkers in conjunction with HPV status are needed in order to select appropriate patients (8, 45). Using the combination of HPV DNA-positive and p16-positive status would be advisable for selecting patients for de-escalation based on HPV status in order to ensure that patients with a poor prognosis do not receive less treatment. The reason for this is that p16 is overexpressed in ~15% of HPV DNA-negative oropharyngeal SCC and these p16-positive/HPV-negative tumours behave in general more like HPV-negative oropharyngeal SCC (49).
Detection of HPV in CUP of the Head and Neck Region by FNAC
There have been a few studies using a variety of approaches for HPV analysis in FNAC from neck masses but unfortunately many of the studies are quite small and have retrospective designs using archival direct or fresh-frozen smears or formalin-fixed paraffin-embedded cell blocks (50-62). Furthermore, some of them did not compare HPV status in the FNAC with HPV status of the primary tumour, but rather investigated whether the HPV status in the FNAC could be used to determine a primary oropharyngeal SCC, showing very good specificity (90-100%), but average sensitivity (50-70%) (50-53, 59, 60).
In earlier reports, in situ hybridization was often used to detect HPV in FNAC (50-53). More recently, commercial methods [e.g. Cervista (Hologic, Marlborough, MA, USA) and Hybrid Capture 2 (Qiagen, Valencia, CA, USA)] already established to detect HPV in gynaecological cytology specimens have been tested (55, 56, 58-60). In these studies, a variety of FNAC preparations (needle rinses or aspirates stored in different solutions and at different temperatures) were used, showing that HPV detection with these methods is feasible, albeit not perfect (51-60). Recently, two prospective polymerase chain reaction-based studies where DNA status was determined using the Roche cobas 4800 system (Roche Molecular System, Pleasanton, CA, USA) were reported (61, 62). One study found perfect correlation between HPV status in the FNAC and the corresponding primary or metastatic carcinomas, as determined by overexpression of p16 (61). Furthermore, they found that all 18 patients with HPV-positive FNAC had an oropharyngeal SCC primary (61). Another study demonstrated a more than 90% concordance between FNAC of the metastasis and the corresponding tumour formalin-fixed paraffin-embedded biopsies (62).
In summary, HPV detection in FNAC is clinically important and shows great potential use, but there is no consensus yet about which methods are optimal with regard to sample preparation and storage, nor for the detection of HPV DNA (50-62). Moreover, in the above studies, mostly patients with malignancies were examined. The accuracy for HPV DNA detection in FNAC needs to be tested in a broader unselected cohort of patients with unknown neck masses. Larger prospective studies are therefore needed to fully conclude how accurate HPV detection in FNAC is and which methods are the most reliable.
Conclusion
HPV is frequently found in CUP of the head and neck region and when so, the CUP has likely an HPV-positive tonsillar and base of tongue SCC as its origin. HPV status is thus useful in guiding the diagnostic work-up and treatment, especially if FNAC is used to determine the HPV status. Patients with HPV-positive CUP with unquestionable SSC cytology could potentially be spared neck dissection, and the radiation fields could possibly be more limited, sparing parts of the head and neck region. Finally, in analogy to tonsillar and base of tongue SCC, patients with HPV-positive CUP have a better clinical outcome than those with HPV-negative CUP, and HPV is thus a valuable prognostic marker in CUP. By combining HPV positivity with additional prognostic biomarkers, selected patients could potentially be included in prospective studies for de-escalated treatment in the future. For these reasons, HPV-testing should be included in the standard clinical work-up of CUP in the head and neck region.
Acknowledgements
This work was supported in part by the Swedish Cancer Foundation, the Stockholm Cancer Society, the Cancer and Allergy Foundation, the Stockholm City Council and the Karolinska Institutet.
Footnotes
This article is freely accessible online.
- Received December 8, 2015.
- Revision received January 14, 2016.
- Accepted January 19, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved