Available online at www.sciencedirect.com

www.elsevier.com/locate/semdp

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3

http://dx.doi.org/10.0740-2570/& 2015 El

The author has nnCorrespondence adE-mail address:

The pathology of HPV-related head and neckcancer: Implications for the diagnostic pathologist

William H. Westra, MDa,b,c,n

aDepartment of Pathology, The Johns Hopkins Medical Institutions, Baltimore, MarylandbDepartment of Oncology, The Johns Hopkins Medical Institutions, Baltimore, MarylandcDepartment of Otolaryngology/Head and Neck Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland

a r t i c l e i n f o

1053/j.semdp.2015.02.023sevier Inc. All rights rese

o relevant financial or codress: Room 2242, 401 Nwwestra@jhmi.edu

a b s t r a c t

A subset of head and neck squamous cell carcinomas are caused by the human papillomavirus

(HPV). This HPV-related form of head and neck squamous cell carcinoma (HPV-HNSCC) has

captured the attention of the oncology community for its rising incidence, its link to non-

traditional risk factors, and its divergent clinical behavior. To diagnose this special form of head

and neck squamous cell carcinoma is to provide important prognostic information and, in

some instances, redirect clinical therapy. The diagnosis of HPV-HNSCC is aided by a strong

appreciation for its characteristic microscopic findings and by an awareness of aberrant

features that set apart a growing list of HPV-HNSCC morphologic variants. This review will

delineate the microscopic appearance of HPV-HNSCC, spotlight ways in which the misinter-

pretation of these microscopic features can lead to diagnostic confusion, offer recommenda-

tions for appropriate terminology when diagnosing HPV-HNSCC, and provide examples of

specific diagnostic scenarios where HPV testing can inform the diagnostic process.

& 2015 Elsevier Inc. All rights reserved.

HPV as a causative agent in head and necksquamous cell carcinoma

The detection of human papillomavirus (HPV) in head andneck squamous cell carcinoma (HNSCC) dates back to the mid1980s, but its role in the initiation and maintenance ofmalignancy was not decisively established until 2000 whenevidence began to mount confirming its nature as a causalagent (Table 1).1 The biologic plausibility of HPV as a carci-nogen of human epithelia is deeply rooted in an extensiveexperience with cervical cancers.2 Like cervical cancer, epi-demiologic studies have observed a consistent associationbetween HPV exposure risk and HPV-related HNSCC (HPV-HNSCC).3,4 Studies documenting HPV integration into the

rved.

mmercial relationships torth Broadway, Baltimore

host genome, expression of viral mRNA transcripts, trans-lation of viral oncoproteins, and disruption of key tumorsuppressor pathways support the biologic activity of HPV anddismiss the notion of “passenger HPV” that is coincidental totumor development.5 Indeed, the consistent presence of HPVacross all stages of clinical progression underscores itsobligatory role in both the initiation and maintenance ofthe malignant phenotype.6–8 Disruption of the p53 and RBpathways by the viral oncoproteins E6 and E7 precludes theneed for various genetic alterations induced by chroniccigarette exposure, thus giving HPV-HNSCCs a molecular-genetic profile that is quite distinct from smoking-relatedcancers.9–11 The presence of HPV in squamous cell carcino-mas of the oropharynx (HPV-OPSCC) also marks a clinically

o disclose., MD 21231-2410.

Table 1 – Evidence supporting the causal role of high-risk human papillomavirus (HPV) in oropharyngeal carcinoma.

Biologic plausibility of HPV as a carcinogen of human epitheliaSpecificity of exposure risk to oropharyngeal site in a dose–response relationshipBiologically active virus as indicated by 1 or more HPV copy per tumor cell, transcriptional activity, and expression of viral oncoproteins

resulting in disruption of the key pathways regulating cell growthPresence and persistence of HPV across all stages of clinical progressionDistinctive genetic profile compared to HPV-negative head and neck squamous cell carcinoma (HNSCC)Distinctive clinical behavior relative to HPV-negative HNSCCDistinctive morphologic features relative to HPV-negative HNSCC

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 43

distinct form of HNSCC characterized by improved clinicaloutcomes relative to its HPV-negative counterpart.12,13 Ineffect, HPV is an important causative agent in a subset ofHNSCCs, and its recognition amounts to nothing less thanthe identification of a pathologically, genetically, and clin-ically distinct tumor entity.

Fig. 1 – PDL1 expression within the tonsillar cryptepithelium.

The oropharynx as a preferential site of HPV-related tumorigenesis

Numerous studies have documented a strong predilection ofHPV-HNSCC for the oropharynx, specifically the lingual andpalatine tonsils.14 At this site, high-risk HPV, particularlyHPV16, is detected in over 80% of squamous cell carcino-mas.15 This preferential targeting likely reflects complex bio-logical interactions between HPV and the specializedlymphoepithelium (i.e., reticulated epithelium) lining thetonsillar crypts including the exploitation of critical immuno-logic checkpoints. As one important example, PD-1:PD-L1signaling is selectively activated in the reticulated epithelium(Figs. 1 and 2). Strong expression of PDL1 in the tonsils cryptsresults in a diminished cytotoxic T-cell response to viralantigens thus creating a potential “immune-privileged” sitefor viral infection, viral persistence, and viral inducedtumorigenesis.16

The microscopic and ultrastructural features of the reticu-lated epithelium lining the tonsillar crypts are distinct fromthe squamous epithelium that lines the surface of the tonsiland other head and neck sites. As the stratified squamousepithelium covering the surface of the tonsils extends intothe recesses of the tonsillar crypts, a dense lymphoid infil-trate obscures the junction between the lymphoid and epi-thelial components and splinters the epithelial sheath intoirregular nests and cords.17 These nests and cords no longerexhibit the polarization and surface maturation that charac-terize the stratified squamous epithelium lining the surfaceof the tonsils. At the cytologic level, the epithelial cells takeon a basaloid appearance (i.e., high nuclear to cytoplasmicratio, absence of cytoplasmic keratinization) with vesicularnuclei and loss of distinct cytoplasmic borders and intercel-lular bridges. These basaloid cells reside on a non-contiguousbasement membrane that is innately disjointed in a way thatfacilitates the unrestrained migration of immune modulatorycells between the lymphoid stroma and lining epithelium in amanner that blurs the interface of the epithelium andlymphoid stroma (Fig. 2).18 The unique microscopic, ultra-structal, and immunologic makeup of the tonsillar cryptsimpacts on fundamental histopathology interpretive tasks

such as tumor grading, tumor classification and the recog-nition of invasive growth (see below).

HPV-related HNSCC in non-oropharyngeal sites

HPV-related carcinomas of the oral cavity, larynx andhypopharynx

Although HPV-related HNSCCs are encountered in non-oropharyngeal sites, the incidence of true HPV-related non-oropharyngeal carcinomas is vastly over stated. There arevarious factors that contribute to inflated incidence rates(Table 2). Among these, differences in HPV detection methodshave a dramatic impact on incidents rates. In particular,highly sensitive assays that are unable to distinguish bio-logically active from inactive infections tend to overestimatethe incidence of HPV-related HNSCCs at non-oropharyngealsites. In a systematic review of PCR-based HPV incidencestudies, Isayeva et al.19 noted highly variable detection rateswith upper limits of HPV positivity reaching 100% in laryngealcarcinomas, 74% in oral cavity carcinomas, and 70% in oralsamples from individuals without cancer. The seeminglyubiquitous presence of HPV in these head and neck samplesstrongly argues for the implementation of detection strat-egies that couple the presence of virus with evidence of itsbiologic activity, and that correlate HPV status with mean-ingful clinical parameters such as disease specific survival.Other factors may contribute to the erroneous diagnosis of

an HPV-related cancer outside of the oropharynx. Occasion-ally, ectopically displaced tonsillar tissue is encountered in

Fig. 2 – Schematic rendition of the tonsillar crypt epithelium highlighting a disrupted basement membrane and a porousepithelial–subepithelial interface.

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 344

non-oropharyngeal sites such as the floor of mouth andhypopharynx.20,21 Potentially, HPV could target these displacedtonsillar tissues and induce tumors in these non-oropharyngeallocations. In some instances, identification of tumor origin maybe elusive when large and bulky tumors involving multiplecontiguous anatomic sites. For example, some locally advancedHPV-positive squamous cell carcinomas of presumed supra-glottic origin actually represent tumors originating from thebase of the tongue. The ability to masquerade as a primarycancer in a non-oropharyngeal site is further enhanced by thepropensity of HPV-related oropharyngeal carcinomas to arisefrom deep within the tonsillar crypts in the absence of surfacedysplasia, propagate beneath the surface epithelium in theabsence of stromal desmoplasia, and then emerge at someadjacent site along Waldheyerʼs ring. In effect, inspection of thepharyngeal membranes often fails to recognize the full extentof tumor spread. Once these factors are properly accounted for,the true incidence of HPV-related HNSCC of the oral cavity,larynx and hypopharynx is only about 5%.22–25

HPV-related carcinomas of the nasopharynx

Several studies have detected the presence of biologically activeHPV in nasopharyngeal carcinomas.26–29 The true incidence of

Table 2 – Factors that contribute to the detection of HPV in he

Presence of biologically inactive HPV (passenger virus, viral contaminanLarge bulky tumor involving multiple contiguous anatomic subsites withTumor migration along Waldheyerʼs ringEctopically displaced tonsillar tissuePresence of biologically active HPV (i.e., true HPV-positive non-orophary

HPV-related nasopharyngeal carcinoma is highly variable andis strongly influenced by the ethnicity of the study populationand by the tumor stage. Although Epstein–Barr virus (EBV) isrecognized as a major etiologic agent for non-keratinizingcarcinomas of the nasopharynx, EBV is not detected in allnasopharyngeal carcinomas. In particular, the constant associ-ation between EBV and non-keratinizing nasopharyngeal carci-nomas noted in Asian patients is not maintained in Caucasianpatients. Those studies of non-keratinizing nasopharyngealcarcinomas that have included a sizeable number of NorthAmerican Caucasian patients have confirmed a substantialproportion of tumors that are HPV, not EBV, positive.26–29

Tumor stage may also influence the likelihood of detectingHPV in a presumed nasopharyngeal carcinoma. Singhi et al.28

observed that HPV was detected only in high-stage nasophar-yngeal carcinomas that showed some degree of involvementof the oropharynx. Accordingly, they proposed that mostHPV-positive nasopharyngeal carcinomas actually arise inthe oropharynx, and then involve the nasopharynx seconda-rily via tumor migration along Waldheyerʼs ring.28 Indeed,clonal migration of neoplastic cells along this track oflymphoepithelial tissue has been proposed as a mechanismto account for tumor multifocality in patients with synchro-nous HPV-related squamous cell carcinomas of Waldheyerʼs

ad and neck carcinomas from non-oropharyngeal sites.

t) detected by highly sensitive assays (e.g., PCR-based assays)no clear epicenter

ngeal carcinomas)

Fig. 3 – HPV-related adenoid cystic-like carcinoma of the sinonasal tract. The surface epithelium exhibits severe squamousdysplasia (A), while the invasive carcinoma resembles a minor salivary carcinoma including the formation of cylindricalstructures and true ducts (B).

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 45

ring including dual involvement of the oropharynx andnasopharynx.30,31

HPV-related carcinomas of the sinonasal tract

The sinonasal tract is not a common site of HNSCC, but up to20% of carcinomas that arise from this anatomic subsiteharbor transcriptionally active HPV.32 The likelihood ofdetecting HPV in sinonasal carcinomas is influenced bycertain phenotypic features. Like the oropharynx, HPV pos-itivity closely tracks with the non-keratinizing morphology.In non-keratinizing squamous cell carcinomas of the sino-nasal tract, HPV detection rates have ranged from 34% to50%.33,34 The likelihood of detecting HPV is also increased invariant forms of squamous cell carcinoma such as those thatexhibit papillary (i.e., papillary squamous cell carcinoma) orglandular (i.e., adenosquamous cell carcinoma) features. Theprognostic importance of detecting HPV is still unclear due tosmall study populations. One of the larger studies to datedemonstrated a strong trend toward improved overall sur-vival in the HPV-positive patients, but a trend that was notstatistically significant.33,34

Bishop et al.35,36 have described an unusual form of HPV-related carcinoma that appears to be restricted to the sinonasaltract. These tumors are set apart by their distinctive adenoidcystic-like appearance and thus have been termed “HPV-related adenoid cystic-like carcinomas of the sinonasal tract”(Fig. 3). Unlike adenoid cystic carcinomas of other head andneck sites, these tumors harbor transcriptionally active HPV,are often associated with surface squamous dysplasia, and arenot associated with the MYB gene rearrangement that charac-terizes most adenoid cystic carcinomas of the head and neck.35

In the sentinel report of 8 patients, there was no documentedregional or distant metastasis; however, the number of studycases was small and clinical follow-up was limited.35

Although some carcinomas of the sinonasal tract harborbiologically active HPV, clear cut differences between HPV-positive tumors of the sinonasal tract and oropharynx should

discourage any conjecture that they represent the samedisease entity. It is not known how HPV is transmitted tothe sinonasal tract. Unlike HPV-OPSCC, sexual transmissionhas not been established for those HPV-positive tumorsarising in the sinonasal tract. The rate of OPSCC has beenincreasing, whereas the rate of sinonasal SCC has been slowlydecreasing. Finally, the clinical implication of detecting HPVin sinonasal carcinomas, in contrast to OPSCCs, is unclear.Given the relatively low incidence of sinonasal carcinomas,multi-institutional studies will be required to address manyunanswered questions relating to the epidemiology, pathol-ogy, and clinical significance of HPV-positive sinonasalcarcinomas.

The morphology of HPV-related oropharyngealcarcinoma

The histologic features of the reticulated epithelium areretained, to varying degrees, in HPV-OPSCC. HPV-OPSSCsconsistently arise from tonsillar crypts. Involvement of thetonsillar surface, when it occurs, is generally a secondaryphenomenon reflecting colonization of the surface epithe-lium as the carcinomas spill over from the tonsillar crypts(Fig. 4). This transition between HPV-HNSCCs and the adja-cent surface epithelium tends to be abrupt without transi-tional zones of epithelial precursor lesions. Indeed, thehistologic progression through the sequential stages of dys-plasia culminating in carcinoma in situ and invasive growththat characterize non-HPV-HNSCCs is not generally evidentfor HPV-OPSCCs. The inability to histologically characterizethe early stages of HPV-induced tumorigenesis continues todeter efforts to assess cancer risk and diagnose early cancers.As these carcinomas infiltrate, they tend to invade as

sheets, lobules, or ribbons of cells (Fig. 5). Central necrosiswithin expanding tumor lobules, sometimes giving rise tocystic degeneration, is a frequent finding. Invasive growthoften does not elicit a strong desmoplastic stromal reaction.Instead, the tumor nests are often surrounded by a zone of

Fig. 4 – HPV-related squamous cell carcinoma of the oropharynx (OPSCC) consistently arises from the tonsillar crypts and notthe surface epithelium (A). A p16 immunohistochemical stain highlights the distribution of the carcinoma in the tonsil (B).

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 346

lymphoid cells that penetrate the tumor nests as tumor-infiltrating lymphocytes. The tumor cells often display a highnuclear to cytoplasmic ratio, syncytial cytoplasm withoutintercellular bridges, and lack significant cytoplasmic kerati-nization. These cellular features can impart a distinct basa-loid appearance. When these features are present and welldeveloped in an oropharyngeal carcinoma, the likelihood thatthe tumor is HPV related is very high. Although the finding ofkeratinization certainly does not exclude the possibility of anHPV-OPSCC,36 its presence in association with surface dys-plasia and a prominent desmoplastic stroma strongly pointsaway from an HPV etiology.In lymph node metastases, the presence of cystic degener-

ation is a common finding. Its presence should warrantstrong consideration of an HPV-related metastasis from theoropharynx. In cytologic preparations from fine-needle aspi-rates of lymph node metastases, the tumor cells formcohesive sheets and clusters and exhibit hyperchromatic,

Fig. 5 – HPV-related OPSCC typically invade the lymphoidstroma as expanding lobules that are permeated by tumor-infiltrating lymphocytes. The absence of a desmoplasticstromal reaction is not unusual.

pleomorphic, and overlapping nuclei (Fig. 6).37 Their highnuclear to cytoplasmic ratio is in contrast to conventionalkeratinizing HNSCC where the keratinized cells exhibit abun-dant orangeophilic cytoplasm.

Histologic grading of HPV-OPSCC

Tumor grade is a semi-quantitative measurement of differ-entiation expressed as the degree to which a tumor resem-bles the normal tissue from which it arises. As a rule ofthumb, the more poorly differentiated a tumor, the moreaggressive its behavior. Numerous studies have uniformlyreported a strong and direct correlation between HPV pos-itivity and a high histologic grade.38 Accordingly, HPV-OPSCCs are widely perceived as poorly or undifferentiatedcarcinomas. This perception is largely based on the immatureappearance of a tumor cell that widely diverges from the

Fig. 6 – Fine-needle aspiration of HPV-related squamous cellcarcinoma yields cellular material in cohesive sheets. Thetumor cells lack overt keratinization, and exhibit a highnuclear to cytoplasmic ratio with nuclear hyperchromasiaand nuclear overlapping.

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 47

stratified squamous epithelium that lines the surface of thetonsil. Using the surface epithelium as a point of reference forphenotypic divergence, however, may not be appropriate,given the common origin of HPV-OPSCCs from the tonsillarcrypts. HPV-OPSCCs often retain the appearance of thereticulated epithelium from which they arise (i.e., permeatinglymphocytes and basaloid cells), and thus might best beregarded as highly differentiated tumors based on this moresuitable comparison (Fig. 7). Recognition of HPV-OPSCC aswell differentiated, not poorly or undifferentiated, is appro-priate as it more accurately reflects histogenic derivation,more fittingly associates tumor grade with expected clinicalbehavior, and seems to explain the perplexing epidemiologictrends toward improving patient survival in the face ofworsening tumor grade for patients with OPSCC.39

Few studies have tried to identify histopathologic parame-ters that might help identify the subgroup of HPV-OPSCCassociated with unfavorable clinical outcomes. Lewis et al.40

made a preliminary observation that the presence of anapla-sia and tumor cell multinucleation are predictive of poorerclinical outcomes in patients with non-keratinizing p16-positive oropharyngeal carcinomas, but this initial observa-tion has yet to be confirmed by others.

Diagnosing invasion in HPV-OPSCC

For many of the HPV-OPSCCs, the difficulty in distinguishingcarcinoma in situ from invasive carcinoma is heightened by(1) origin from the tonsillar crypts beneath the surfaceepithelium; (2) the oft absent tell-tale sign of invasion—stromal desmoplasia; (3) the well recognized propensity ofsmall, sometimes occult, tonsillar carcinomas to metastasizeto regional nodes in the absence of clear cut stromal invasion;(4) the blurred junction of the reticulated epithelium and theunderlying lymphoid stroma; and (5) the porous nature of

Fig. 7 – Although HPV-related OPSCCs are often regarded aspoorly differentiated based on the immature and basaloidappearance of the tumor cells (red asterisk), they often retaina close resemble to the non-neoplastic reticular epitheliumlining the tonsillar crypts (black asterisk). (For interpretationof the references to colour in this figure legend, the reader isreferred to the web version of this article.)

this epithelial/lymphoid junction where the basal cell layer isincomplete and its supporting basement membrane is dis-rupted and non-contiguous (Fig. 2).18,41 In effect, the timehonored microscopic approach to the recognition of tumorinvasion may not be valid for those HPV-OPSCCs arising fromthe tonsillar crypts. Until the histologic progression of HPV-related neoplasia of the tonsils is better characterized and thecritical transition marking infiltrative growth is more clearlydemarcated, an aggressive approach that regards all HPV-related neoplasia of the tonsils as potentially malignant, evenin the absence of those histologic features that have beentraditionally used to diagnosis invasion, may be warranted.

Diagnostic terminology in the reporting of HPV-OPSCC

The same morphologic features that characterize HPV-OPSCCs also can cause considerable confusion when it comesto reporting these tumors. Given the distinctiveness of HPV-OPSCC as a biological and clinical subtype of HNSCC, thereporting of these cancers must make careful use of diag-nostic terminology that asserts their relationship with HPV,and yet does not confound the treating clinician. As noted,the current and widespread practice of assigning thesetumors with an advanced histologic grade (e.g., poorly differ-entiated or undifferentiated) is misleading for the way itinappropriately infers an aggressive clinical course. Similarly,the common use of the term “basaloid” as a diagnosticdescriptor, though morphologically correct, invites an erro-neous connection with the basaloid variant of squamous cellcarcinoma—the subtype of HNSCC notorious for aggressiveclinical behavior.41,42 When reporting HPV-OPSCCs, it is ourcurrent practice to suspend the use of inapt descriptors suchas “basaloid,” do away with the conventional grading schemefor HNSCC, and routinely report on the HPV status of allHNSCCs arising in the oropharynx.For OPSCCs that are found to be HPV positive, we prefer the

term “HPV-related squamous cell carcinoma” for its simplic-ity, clarity, and directness. At the same time, blunt use of theterm “HPV-related” can have a powerful psychological impacton patients, patient partners, and family members, givenconsiderable uncertainty regarding transmission, latency,and communicability of oral HPV. Use of the term shouldonly be used in those HNSCCs where the presence of bio-logically active HPV has been confirmed by reliable HPVdetection assays, and only for those HPV-related carcinomasarising in the oropharynx. Clearly, there is a growing need fora consensus panel to put forward a classification schemethat, like the nasopharynx, is customized for carcinomas ofthe oropharynx.

Morphologic variants of HPV-related OPSCC

On occasion, HPV-OPSCC can deviate from its usual morpho-logic profile. Recognized morphologic variants include papil-lary squamous cell carcinoma,43,44 adenosquamouscarcinoma,45 basaloid squamous cell carcinoma,42

lymphoepithelial-like carcinoma,46 sarcomotoid carcinoma,47

Fig. 9 – This HPV-OPSCC exhibits well-developedlymphoepithelial features including syncytial cytoplasm,vesicular nuclei, prominent nucleoli, and a denselymphoplasmacytic infiltrate.

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 348

and small cell carcinoma.48,49 With the notable exception ofHPV-related small cell carcinoma, morphologic variance doesnot seem to influence clinical behavior. An awareness ofthese variant forms is important so that HPV-OPSCC is notoverlooked and mistaken as some more aggressive type ofcarcinoma. Accordingly, a few of these variants warrantspecial consideration.

HPV-HNSCC with papillary features

Papillary squamous cell carcinoma (PSCC) is an uncommonvariant of HNSCC that is characterized by exophytic papillarygrowth. The papillary fronds may be lined by keratinizedsquamous cells or non-keratinized basaloid cells. The pres-ence of biologically active HPV is most commonly noted inthose PSCCs of the non-keratinized variety.43,44 Most HPV-positive PSCCs arise in the oropharynx, but they also occur inthe sinonasal tract and, less commonly, the larynx.43,44

Patients with PSCCs are generally believed to have a betterprognosis than patients with conventional SCCs.50 In largepart this improved prognosis reflects its limited invasiveness.It is not clear whether the presence of HPV adds any survivalbenefit.

HPC-HNSCC with basaloid features

The lobular arrangement of compact tumor cells with a highnuclear to cytoplasmic ratio in the absence of keratinizationconveys a distinctly “basaloid” appearance to HPV-HNSCCs(Fig. 8). When the basaloid morphology is highly developed,HPV-related SCC may be histologically indistinguishable fromthe basaloid squamous variant of SCC—a variant of HNSCCthat is set apart as a distinct subtype based on its strikingbasaloid morphology and its aggressive clinical behavior.51 Atthe microscopic level, the distinction between HPV-HNSCCand the basaloid variant of HNSCC can be difficult as bothtumors are characterized by the lobular growth of basaloidcells. A subset of basaloid squamous cell carcinomas harborsbiologically active HPV. These HPV-positive tumors usuallyarise in the oropharynx, while the HPV-negative tumors are

Fig. 8 – In this HPV-related OPSCC, the basaloid features areso highly developed that it closely resembles the clinicallyaggressive basaloid squamous cell carcinoma.

more likely to be encountered in the supraglottic larynx andhypopharynx. Morphologic similarities aside, HPV-OPSCCs donot share the same aggressive clinical behavior that charac-terizes the basaloid variant of squamous cell carci-noma.42,52,53 In their evaluation of basaloid HNSCCs, Begumet al.42 found that the presence of HPV was significantlyassociated with improved overall survival even thoughpatients with HPV-HNSCCs were more likely to present withlymph nodes metastases. In effect, the detection of HPVessentially downgrades what would otherwise be regardedas a high-grade HNSCC.

HPC-OPSCC with lymphoepithelial features

Some HPV-related OPSCCs demonstrate lymphoepithelialfeatures including tumor cells with syncytial cytoplasm,vesicular nuclei, and large central nucleoli (Fig. 9). The cellsare dispersed in a lymphoplasmacytic background as cords,clusters, or single cells.46,54 When these lymphoepithelialfeatures are highly developed, an HPV-HNSCC may be mis-taken for an Epstein–Barr virus (EBV)-induced undifferenti-ated carcinoma of the nasopharynx. Based on thismorphologic overlap, one cannot assume an EBV-drivenprocess by phenotype alone. Despite its “undifferentiated”histologic appearance, the clinical outcome associated withthe lymphoepithelial variant seems to be favorable andcomparable to that associated with conventional HPV-related OPSCC.46,54

HPV-HNSCC with small cell features

Recent reports have drawn attention to a variant form ofHPV-HNSCC characterized by a small cell phenotype that isindistinguishable from small cell carcinoma of the lung andother sites.48,49 We have also detected HPV in high-gradeneuroendocrine carcinomas of the oropharynx that are bestclassified as large cell neuroendocrine carcinomas (unpub-lished observations). The small cell variant is comprised ofsmall anaplastic cells with hyperchromatic nuclei, scant

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 49

cytoplasm, a high mitotic rate and tumor necrosis. In 4 of the9 tumors reported by Bishop and Westra,48 the small cellcomponent arose in close association with a more conven-tional HPV-HNSCC component. In these cases, HPV wasdetected in both the small cell and conventional components.For patients with HPV-related cancer of the oropharynx,recognition of the small cell variant and its distinction fromHPV-related squamous cell carcinoma is important but notnecessarily straightforward. Both tumor types share certainmorphologic features including small hyperchromatic cellswith scant cytoplasm and tumor necrosis. In these instances,immunohistochemistry may play a role in confirming thepresence of a small cell component. By immunohistochem-istry, the small cell component consistently demonstratesneuroendocrine differentiation (e.g., positive for synaptophy-sin and chromogranin) and loss of squamous differentiation(e.g., negative for p63 and CK5/6).Importantly, HPV does not convey a favorable prognosis

when its presence is detected in this small cell variant. HPV-related small cell carcinoma of the oropharynx appears toshare the same aggressive clinical features of its counterpartin the uterine cervix and lung where the small cell phenotype

Fig. 10 – HPV-related sarcomatoid carcinoma. This example of sonly in the epithelial component (B), but both the conventionalimmunohistochemistry (C) and HPV16 positive by in situ hybrid

is associated with early distant spread and poor overallsurvival.48,49 Consequently, the small cell phenotype shouldbe regarded as an undifferentiated form of HPV-relatedoropharyngeal carcinoma where tumor morphology super-sedes HPV positivity as a prognostic indicator. In turn, thepresence of a small cell component should disqualify anypatient with an HPV-OPSCC from consideration as a candi-date for less intensive multimodality therapy (i.e., de-escalation therapy) solely on the grounds of HPV positivity.

HPV-HNSCC with spindle cell features (HPV-relatedsarcomatoid carcinoma)

Sarcomatoid carcinoma (SA) of the upper respiratory tract is adistinct variant of squamous cell carcinoma trademarked bya prominent or even exclusive spindle cell component.55

Although the spindle cells may take on a highly mesenchy-mal appearance, they are in fact derived from the surfacesquamous epithelium. Recently, HPV has been detected in asubset of sarcomatoid carcinomas arising in the orophar-ynx.47,56 Importantly, HPV is present in both the conventionaland spindle cell components (Fig. 10). The strict epithelial

arcomatoid carcinoma (A) exhibited pancytokeratin stainingand spindle cell components were p16 positive byization (D).

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 350

tropism of HPV infection confirms the epithelial nature of thespindle cell component and serves as an important diagnos-tic tool in those difficult cases where, at the one extreme, thespindle cell proliferation may closely resemble a reactivefibroblastic proliferation, and at the other extreme, where itmay resemble a true sarcoma.

Methods of HPV detection

Given the clinical implications of HPV status, even to thepoint of dictating the type and intensity of cancer therapy,the need for routine and accurate HPV testing of oropharyn-geal carcinomas is compelling and urgent. Indeed, the Collegeof American Pathologists has recommended routine HPVtesting as part of the standard pathologic evaluation ofresected oropharyngeal squamous cell carcinomas (http://www.cap.org/apps/docs/committees/cancer/cancer_protocols/2013/Pharynx_13protocol_3300.pdf), and Cancer Care Ontariohas published evidence based guidelines for routine testing ofHNSCCs (https://www.cancercare.on.ca/common/pages/UserFile.aspx?fileId=279836). Despite the expectation for routineHPV testing, there is currently no standard approach. Instead,methods of HPV testing across laboratories vary considerably,reflecting the biases and tendencies of individual investigatorsand the cost to benefit ratio of each technique.57,58

Immunostaining for p16 protein has recently been regardedas a practical alternative or complimentary procedure for HPVtesting of oropharyngeal cancers based on a high correlationbetween the HPV detection and p16 overexpression in recentstudies.15,59–61 The simplicity, low cost, and high sensitivity ofp16 immunohistochemistry have prompted consideration ofreplacing more intensive PCR-based methods as a standaloneHPV test.60 To be truly useful as a surrogate marker of HPVinfection, the interpretation of p16 immunohistochemistrymust be informed by various histological, anatomical, andclinical considerations.62 First, p16 IHC may substitute forHPV testing when strong staining is present in the nucleusand cytoplasm of the tumor cells throughout all or most(470%) of the tumor. Focal or weak staining should besupported by other forms of HPV testing. Second, while thesensitivity and specificity of p16 staining as a marker of HPVinfection is sufficiently high to serve as a reliable test forsquamous cell carcinomas of oropharyngeal origin, thesevalues are either unknown or unacceptably low for HNSCCsarising in non-oropharyngeal sites.22,23,25 Third, interpreta-tion of p16 staining must be informed by the morphologicfeatures of the tumor as outlined above. P16 IHC staining maysubstitute for HPV detection in those oropharyngeal carcino-mas that demonstrate the typical morphology of HPV-relatedHNSCC. Additional HPV testing should be performed in p16

Table 3 – Diagnostic applications for HPV testing.

1) Determine site of tumor origin in patients presenting with a lymph n2) Discern tumor relationships in patients with oropharyngeal carcinom3) Distinction between HPV- and EBV-related cancer in those tumors ex4) Separation of HPV-related carcinoma from basaloid variant of squam5) Recognition of HPV-related cancer and its distinction from non-neopl6) Distinction of an HPV-related cystic metastasis from a branchial cleft

negative oropharyngeal carcinomas that exhibit classic HPV-related histomorphology, and in p16-positive oropharyngealcarcinomas that do not exhibit classic HPV morphology.Fourth, p16 IHC is currently used primarily as a prognosticindicator for patients with oropharyngeal carcinoma, and anyexpanded clinical role for HPV detection (e.g., selection forHPV therapeutic vaccine or therapeutic de-escalation) maynecessitate more stringent detection methods.

Diagnostic applications of HPV testing

Although HPV testing is generally performed for prognosti-cation purposes, there are specific diagnostic scenarios whereknowledge of HPV status can directly inform the diagnosticprocess. As a result, HPV testing is finding increasing use bythe pathologist as a very helpful diagnostic adjunct (Table 3).

Determination of site of tumor origin

In HPV-related OPSCCs, the presence of HPV persists acrossall stages of clinical progression such that it is just as readilydetected in metastatic implants as in the correspondingprimary cancers.7 Consequently, a lymph node metastasis isquite suitable as a substrate for HPV testing, obviating theneed for additional tissue acquisition in those patients withsmall or even occult primary cancers. For those patients whopresent with neck metastases in the absence of an obviousprimary tumor, HPV testing of a lymph node metastasis is aneffective strategy for localizing the site of origin. In thesepatients, the detection of HPV in a lymph node metastasis is areliable predictor of oropharyngeal origin.63–65 Similarly, forthe squamous cell carcinoma in the lung of a patient with aprior HPV-related oropharyngeal carcinoma, HPV detectionprovides a direct link between the two tumors and providescompelling evidence that the tumor in the lung represents ametastasis rather than a new primary lung primary.66,67 Thislink to an oropharyngeal primary is particularly useful giventhe observation that HPV-related OPSCCs can sometimesmetastasize to distant sites long after treatment of theirprimary cancers, even well beyond the 2- to5-year intervalthat is usually set as the threshold for distinguishing aprimary lung cancer from a metastatic HNSCC.66

Distinction from non-HPV types of HNSCC

As noted above, lymphoepithelial features may be presentand fully developed in a subset of HPV-related OPSCCs.Failure to recognize that the lymphoepithelial phenotype isnot entirely restricted to EBV-driven undifferentiated naso-pharyngeal carcinomas may be problematic when it is

ode metastasisa who develop a squamous carcinoma in the lunghibiting a lymphoepithelial morphologyous cell carcinomaastic tissuescyst

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 51

encountered in cervical lymph node metastases. Assump-tions about nasopharyngeal origin based solely on the mor-phologic findings run the risk of inappropriately divertingtreatment away from the oropharynx and toward the naso-pharynx. Accordingly, testing for both HPV and EBV isadvisable when a lymphoepithelial phenotype is encounteredin lymph node metastases in the absence of a known primarytumor site.Alternatively, HPV-related OPSCCs may be easily confused

with the basaloid variant of squamous cell carcinoma basedon similar morphologic features, yet the distinction is impor-tant as these tumors set apart by very different clinicalbehaviors. When dealing with basaloid tumors of the oro-pharynx, HPV testing is advisable to separate the clinicallyfavorable HPV-positive carcinoma with basaloid morphologyfrom the highly aggressive HPV-negative basaloid variant ofsquamous cell carcinoma.42

Squamous-lined cysts of the lateral neck

Patients with squamous cell carcinomas of the head and neckoften present with a neck mass as the initial and onlymanifestation of their disease. These cervical metastasesfrequently undergo cystic degeneration causing clinical con-fusion with benign cystic squamous lesions. Even fine-needleaspiration cytopathology is not always helpful in making aclear cut distinction between benign and malignant cysticsquamous lesions of the neck, especially when it comes toseparating benign lymphoepithelial cysts with reactive squ-amous atypia from well-differentiated squamous cell carci-nomas.68 This diagnostic difficulty has prompted a pursuit ofadditional diagnostic techniques including the immunohis-tochemical detection of differentially expressed markers.When struggling with this diagnostic dilemma, detection ofHPV in the cyst wall provides very compelling evidence for acystic metastasis from the oropharynx. Of note, p16 is over-expressed in almost one-half of benign branchial cleft cysts,rendering p16 immunostaining of little if any value in thisdiagnostic scenario.69 Accordingly, some other method ofHPV detection such as in situ hybridization is necessary toconfirm the malignant nature of a squamous-lined cyst inthe neck.

Recognition of HPV-related cancer and its distinction fromnon-neoplastic tissues

The histologic recognition of an HPV-related OPSCC is notalways straightforward and may benefit from HPV detectionassays such as p16 immunohistochemical staining that high-light the presence of HPV infected tumor cells. Several factorscontribute to diagnostic difficulty. First, HPV-related OPSCCsare often very well differentiated, retaining a close resem-blance to the reticulated epithelium lining the tonsillarcrypts. Second, the tumors are often masked by an obscuringlymphoid background. Third, the tumors may be inconspic-uously small, even in patients who present with advancemetastatic spread to regional lymph nodes. Indeed, inpatients with HPV-positive lymph node metastases fromtumors of an unknown primary site, p16 staining of thetonsillectomy specimens sometimes discloses small clusters

of HPV-carcinoma that were not apparent by routine histol-ogy alone.64

In summary, the emergence of a distinct subset of headand neck cancer that is caused by high-risk HPV is re-shaping the oncology landscape and is impacting on currentdiagnostic practices. The morphology of these HPV-relatedare characteristic, but these same histologic features arechallenging traditional diagnostic paradigms relating to fun-damental concerns such as tumor grading, tumor classifica-tion and the histologic recognition of tumor invasion.Confirmation of HPV status is not only important for esti-mating clinical outcomes and directing therapy, but candirectly inform the diagnostic process in specific diagnosticscenarios.

r e f e r e n c e s

1. Gillison ML, Shah KV. Human papillomavirus-associated headand neck squamous cell carcinoma: mounting evidence foran etiologic role for human papillomavirus in a subset of headand neck cancers. Curr Opin Oncol. 2001;13(3):183–188.

2. Walboomers JM, Jacobs MV, Manos MM, et al. Human papil-lomavirus is a necessary cause of invasive cervical cancerworldwide. J Pathol. 1999;189(1):12–19.

3. DʼSouza G, Kreimer AR, Viscidi R, et al. Case-control study ofhuman papillomavirus and oropharyngeal cancer. N Engl JMed. 2007;356(19):1944–1956.

4. Gillison ML, DʼSouza G, Westra W, et al. Distinct risk factorprofiles for human papillomavirus type 16-positive andhuman papillomavirus type 16-negative head and neck can-cers. J Natl Cancer Inst. 2008;100(6):407–420.

5. Gillison ML, Koch WM, Capone RB, et al. Evidence for a causalassociation between human papillomavirus and a subset ofhead and neck cancers. J Natl Cancer Inst. 2000;92(9):709–720.

6. Laborde RR, Janus JR, Olsen SM, et al. Human papillomavirusin oropharyngeal squamous cell carcinoma: assessing viruspresence in normal tissue and activity in cervical metastasis.Laryngoscope. 2012;122(12):2707–2711.

7. Mehrad M, Zhao H, Gao G, Wang X, Lewis JS Jr.Transcriptionally-active human papillomavirus is consis-tently retained in the distant metastases of primary orophar-yngeal carcinomas. Head Neck Pathol. 2014;8(2):157–163.

8. Ruzevick J, Olivi A, Westra WH. Metastatic squamous cellcarcinoma to the brain: an unrecognized pattern of distantspread in patients with HPV-related head and neck cancer.J Neuro-Oncol. 2013;112(3):449–454.

9. Agrawal N, Frederick MJ, Pickering CR, et al. Exome sequencingof head and neck squamous cell carcinoma reveals inactivatingmutations in NOTCH1. Science. 2011;333(6046):1154–1157.

10. Fallai C, Perrone F, Licitra L, et al. Oropharyngeal squamouscell carcinoma treated with radiotherapy or radiochemother-apy: prognostic role of TP53 and HPV status. Int J Radiat OncolBiol Phys. 2009;75(4):1053–1059.

11. Marur S, DʼSouza G, Westra WH, Forastiere AA. HPV-associated head and neck cancer: a virus-related cancerepidemic. Lancet Oncol. 2010;11(8):781–789.

12. Ang KK, Harris J, Wheeler R, et al. Human papillomavirus andsurvival of patients with oropharyngeal cancer. N Engl J Med.2010;363(1):24–35.

13. Fakhry C, Westra WH, Li S, et al. Improved survival of patientswith human papillomavirus-positive head and neck squa-mous cell carcinoma in a prospective clinical trial. J NatlCancer Inst. 2008;100(4):261–269.

14. Begum S, Cao D, Gillison M, Zahurak M, Westra WH. Tissuedistribution of human papillomavirus 16 DNA integration in

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 352

patients with tonsillar carcinoma. Clin Cancer Res. 2005;11(16):5694–5699.

15. Singhi AD, Westra WH. Comparison of human papillomavirusin situ hybridization and p16 immunohistochemistry in thedetection of human papillomavirus-associated head and neckcancer based on a prospective clinical experience. Cancer.2010;116(9):2166–2173.

16. Lyford-Pike S, Peng S, Young GD, et al. Evidence for a role ofthe PD-1:PD-L1 pathway in immune resistance of HPV-associated head and neck squamous cell carcinoma. CancerRes. 2013;73(6):1733–1741.

17. Mills SE. Histology for Pathologists3rd ed. Philadelphia: Lippin-cott Williams & Wilkins; 2007.

18. Perry ME, Jones MM, Mustafa Y. Structure of the cryptepithelium in human palatine tonsils. Acta Oto-LaryngologicaSuppl. 1988;454:53–59.

19. Isayeva T, Li Y, Maswahu D, Brandwein-Gensler M. Humanpapillomavirus in non-oropharyngeal head and neck cancers:a systematic literature review. Head Neck Pathol. 2012;6(suppl1):S104–S120.

20. Baba Y, Kato Y, Ogawa K. Hyperplasia of lymphoid structuresin the hypopharynx: a case report. J Med Case Rep. 2010;4:388.

21. Patel K, Ariyaratnam S, Sloan P, Pemberton MN. Oral tonsils(ectopic oral tonsillar tissue). Dent Update. 2004;31(5):291–292.

22. Poling JS, Ma XJ, Bui S, et al. Human papillomavirus (HPV)status of non-tobacco related squamous cell carcinomas ofthe lateral tongue. Oral Oncol. 2014;50(4):306–310.

23. Lingen MW, Xiao W, Schmitt A, et al. Low etiologic fractionfor high-risk human papillomavirus in oral cavity squamouscell carcinomas. Oral Oncol. 2013;49(1):1–8.

24. Bishop JA, Ma XJ, Wang H, et al. Detection of transcriptionallyactive high-risk HPV in patients with head and neck squa-mous cell carcinoma as visualized by a novel E6/E7 mRNAin situ hybridization method. Am J Surg Pathol. 2012;36(12):1874–1882.

25. Chernock RD, Wang X, Gao G, et al. Detection and significanceof human papillomavirus, CDKN2A(p16) and CDKN1A(p21)expression in squamous cell carcinoma of the larynx. ModPathol. 2013;26(2):223–231.

26. Maxwell JH, Kumar B, Feng FY, et al. HPV-positive/p16-positive/EBV-negative nasopharyngeal carcinoma in whiteNorth Americans. Head Neck. 2010;32(5):562–567.

27. Stenmark MH, McHugh JB, Schipper M, et al. NonendemicHPV-positive nasopharyngeal carcinoma: association withpoor prognosis. Int J Radiat Oncol Biol Phys. 2014;88(3):580–588.

28. Singhi AD, Califano J, Westra WH. High-risk human papilloma-virus in nasopharyngeal carcinoma. Head Neck. 2012;34(2):213–218.

29. Dogan S, Hedberg ML, Ferris RL, Rath TJ, Assaad AM, ChioseaSI. Human papillomavirus and Epstein–Barr virus in naso-pharyngeal carcinoma in a low-incidence population. HeadNeck. 2014;36(4):511–516.

30. McGovern SL, Williams MD, Weber RS, et al. Three synchro-nous HPV-associated squamous cell carcinomas of Wal-deyerʼs ring: case report and comparison with Slaughterʼsmodel of field cancerization. Head Neck. 2010;32(8):1118–1124.

31. Joseph AW, Ogawa T, Bishop JA, et al. Molecular etiology ofsecond primary tumors in contralateral tonsils of humanpapillomavirus-associated index tonsillar carcinomas. OralOncol. 2013;49(3):244–248.

32. Lewis JS Jr, Westra WH, Thompson LD, et al. The sinonasaltract: another potential "hot spot" for carcinomas withtranscriptionally-active human papillomavirus. Head NeckPathol. 2014;8(3):241–249.

33. Bishop JA, Guo TW, Smith DF, et al. Human papillomavirus-related carcinomas of the sinonasal tract. Am J Surg Pathol.2013;37(2):185–192.

34. El-Mofty SK, Lu DW. Prevalence of high-risk human papillo-mavirus DNA in nonkeratinizing (cylindrical cell) carcinomaof the sinonasal tract: a distinct clinicopathologic and molec-ular disease entity. Am J Surg Pathol. 2005;29(10):1367–1372.

35. Bishop JA, Ogawa T, Stelow EB, et al. Human papillomavirus-related carcinoma with adenoid cystic-like features: a pecu-liar variant of head and neck cancer restricted to the sino-nasal tract. Am J Surg Pathol. 2013;37(6):836–844.

36. Cai C, Chernock RD, Pittman ME, El-Mofty SK, Thorstad WL,Lewis JS Jr. Keratinizing-type squamous cell carcinoma of theoropharynx: p16 overexpression is associated with positivehigh-risk HPV status and improved survival. Am J Surg Pathol.2014;38(8):809–815.

37. Jarboe EA, Hunt JP, Layfield LJ. Cytomorphologic diagnosis andHPV testing of metastatic and primary oropharyngeal squ-amous cell carcinomas: a review and summary of the liter-ature. Diagn Cytopathol. 2012;40(6):491–497.

38. Mendelsohn AH, Lai CK, Shintaku IP, et al. Histopathologicfindings of HPV and p16 positive HNSCC. Laryngoscope.2010;120(9):1788–1794.

39. Mehta V, Yu GP, Schantz SP. Population-based analysis of oraland oropharyngeal carcinoma: changing trends of histopa-thologic differentiation, survival and patient demographics.Laryngoscope. 2010;120(9):2203–2212.

40. Lewis JS Jr, Scantlebury JB, Luo J, Thorstad WL. Tumor cellanaplasia and multinucleation are predictors of diseaserecurrence in oropharyngeal squamous cell carcinoma,including among just the human papillomavirus-relatedcancers. Am J Surg Pathol. 2012;36(7):1036–1046.

41. Thariat J, Badoual C, Faure C, Butori C, Marcy PY, Righini CA.Basaloid squamous cell carcinoma of the head and neck: roleof HPV and implication in treatment and prognosis. J ClinPathol. 2010;63(10):857–866.

42. Begum S, Westra WH. Basaloid squamous cell carcinoma ofthe head and neck is a mixed variant that can be furtherresolved by HPV status. Am J Surg Pathol. 2008;32(7):1044–1050.

43. Jo VY, Mills SE, Stoler MH, Stelow EB. Papillary squamous cellcarcinoma of the head and neck: frequent association withhuman papillomavirus infection and invasive carcinoma. AmJ Surg Pathol. 2009;33(11):1720–1724.

44. Mehrad M, Carpenter DH, Chernock RD, et al. Papillarysquamous cell carcinoma of the head and neck: clinicopatho-logic and molecular features with special reference to humanpapillomavirus. Am J Surg Pathol. 2013;37(9):1349–1356.

45. Masand RP, El-Mofty SK, Ma XJ, Luo Y, Flanagan JJ, Lewis JS Jr.Adenosquamous carcinoma of the head and neck: relation-ship to human papillomavirus and review of the literature.Head Neck Pathol. 2011;5(2):108–116.

46. Singhi AD, Stelow EB, Mills SE, Westra WH. Lymphoepithelial-like carcinoma of the oropharynx: a morphologic variant ofHPV-related head and neck carcinoma. Am J Surg Pathol.2010;34(6):800–805.

47. Bishop JA, Montgomery EA, Westra WH. Use of p40 and p63immunohistochemistry and human papillomavirus testing asancillary tools for the recognition of head and neck sarcoma-toid carcinoma and its distinction from benign and malignantmesenchymal processes. Am J Surg Pathol. 2014;38(2):257–264.

48. Bishop JA, Westra WH. Human papillomavirus-related smallcell carcinoma of the oropharynx. Am J Surg Pathol. 2011;35(11):1679–1684.

49. Kraft S, Faquin WC, Krane JF. HPV-associated neuroendocrinecarcinoma of the oropharynx: a rare new entity with potentiallyaggressive clinical behavior. Am J Surg Pathol. 2012;36(3):321–330.

50. Thompson LD, Wenig BM, Heffner DK, Gnepp DR. Exophyticand papillary squamous cell carcinomas of the larynx: aclinicopathologic series of 104 cases. Otolaryngol-Head NeckSurg. 1999;120(5):718–724.

S E M I N A R S I N D I A G N O S T I C P A T H O L O G Y 3 2 ( 2 0 1 5 ) 4 2 – 5 3 53

51. Wain SL, Kier R, Vollmer RT, Bossen EH. Basaloid-squamouscarcinoma of the tongue, hypopharynx, and larynx: report of10 cases. Hum Pathol. 1986;17(11):1158.

52. Thariat J, Badoual C, Faure C, et al. Basaloid squamous cellcarcinomas of the head and neck. Bull Cancer. 2009;96(10):989–1004.

53. Chernock RD, Lewis JS Jr, Zhang Q, El-Mofty SK. Humanpapillomavirus-positive basaloid squamous cell carcinomasof the upper aerodigestive tract: a distinct clinicopathologicand molecular subtype of basaloid squamous cell carcinoma.Hum Pathol. 2010;41(7):1016–1023.

54. Carpenter DH, El-Mofty SK, Lewis JS Jr. Undifferentiatedcarcinoma of the oropharynx: a human papillomavirus-associated tumor with a favorable prognosis. Mod Pathol.2011;24(10):1306–1312.

55. Batsakis JG, Suarez P. Sarcomatoid carcinomas of the upperaerodigestive tracts. Adv Anat Pathol. 2000;7(5):282–293.

56. Watson RF, Chernock RD, Wang X, et al. Spindle cell carcinomasof the head and neck rarely harbor transcriptionally-activehuman papillomavirus. Head Neck Pathol. 2013;7(3):250–257.

57. Westra WH. Detection of human papillomavirus (HPV) inclinical samples: evolving methods and strategies for theaccurate determination of HPV status of head and neckcarcinomas. Oral Oncol. 2014;50(9):771–779.

58. Mirghani H, Amen F, Moreau F, et al. Human papilloma virustesting in oropharyngeal squamous cell carcinoma: what theclinician should know. Oral Oncol. 2014;50(1):1–9.

59. Hoffmann M, Ihloff AS, Gorogh T, et al. p16(INK4a) overexpres-sion predicts translational active human papillomavirus infec-tion in tonsillar cancer. Int J Cancer. 2010;127(7):1595–1602.

60. Lewis JS Jr. p16 Immunohistochemistry as a standalone testfor risk stratification in oropharyngeal squamous cell carci-noma. Head Neck Pathol. 2012;6(suppl 1):S75–S82.

61. Jordan RC, Lingen MW, Perez-Ordonez B, et al. Validation ofmethods for oropharyngeal cancer HPV status determination

in US cooperative group trials. Am J Surg Pathol. 2012;36(7):945–954.

62. El-Naggar AK, Westra WH. p16 expression as a surrogatemarker for HPV-related oropharyngeal carcinoma: a guide forinterpretative relevance and consistency. Head Neck. 2012;34(4):459–461.

63. Begum S, Gillison ML, Ansari-Lari MA, Shah K, Westra WH.Detection of human papillomavirus in cervical lymph nodes:a highly effective strategy for localizing site of tumor origin.Clin Cancer Res. 2003;9(17):6469–6475.

64. Begum S, Gillison ML, Nicol TL, Westra WH. Detection ofhuman papillomavirus-16 in fine-needle aspirates to deter-mine tumor origin in patients with metastatic squamous cellcarcinoma of the head and neck. Clin Cancer Res. 2007;13(4):1186–1191.

65. Zhang MQ, El-Mofty SK, Davila RM. Detection of humanpapillomavirus-related squamous cell carcinoma cytologi-cally and by in situ hybridization in fine-needle aspirationbiopsies of cervical metastasis: a tool for identifying the siteof an occult head and neck primary. Cancer. 2008;114(2):118–123.

66. Bishop JA, Ogawa T, Chang X, et al. HPV analysis in distin-guishing second primary tumors from lung metastases inpatients with head and neck squamous cell carcinoma. Am JSurg Pathol. 2012;36(1):142–148.

67. Weichert W, Schewe C, Denkert C, Morawietz L, Dietel M,Petersen I. Molecular HPV typing as a diagnostic tool todiscriminate primary from metastatic squamous cell carci-noma of the lung. Am J Surg Pathol. 2009;33(4):513–520.

68. Briggs RD, Pou AM, Schnadig VJ. Cystic metastasis versusbranchial cleft carcinoma: a diagnostic challenge. Laryngo-scope. 2002;112(6):1010–1014.

69. Cao D, Begum S, Ali SZ, Westra WH. Expression of p16 inbenign and malignant cystic squamous lesions of the neck.Hum Pathol. 2010;41(4):535–539.