A Case of HPV-Associated Oropharyngeal Squamous Cell Carcinoma with Block-Like, Partial Loss of p16 Expression

Abstract

Background

Oropharyngeal squamous cell carcinoma is frequently associated with high-risk HPV infection, which confers a good prognosis. Immunohistochemistry for p16 is used as a surrogate for HPV status, but discrepant results are occasionally seen. Here, we report a case with a unique pattern of partial loss of p16.

Methods

A 63 year old male presented with a base of tongue nonkeratinizing squamous cell carcinoma and a large metastatic neck mass. The primary lesion and multiple regions of the metastatic mass were assessed with p16 immunohistochemistry, RNA in situ hybridization for high-risk HPV, and HPV16 genome sequencing.

Results

The primary lesion was p16 negative, and the metastatic neck mass had large, confluent regions that were either strongly p16 positive or entirely p16 negative. All of these regions were positive for high-risk HPV with identical HPV16 genomes.

Conclusion

This unusual case illustrates a potential diagnostic pitfall, and it raises important questions regarding molecular mechanisms and prognostic implications of p16 staining in oropharyngeal squamous cell carcinoma.

Introduction

Oropharyngeal squamous cell carcinoma (OPSCC) has increased in incidence over the last several decades, driven by a rising number of tumors positive for high-risk human papillomavirus (HPV) [1, 2]. HPV-positivity in OPSCC confers a markedly better prognosis [3], so determining HPV status is an essential component of pathologic assessment and requisite for correct tumor staging. Because transcriptionally-active high-risk HPV is strongly associated with expression of p16 by immunohistochemistry (IHC), p16 IHC is recommended as a prognostic marker and surrogate marker for HPV status [4].

There is a growing body of literature describing discordance between p16 IHC and HPV status assessed by other methods [5–7]. Most commonly, these tumors are positive for p16, but have undetectable HPV DNA or RNA. The incidence of these tumors is not entirely clear, and may vary by geographic region, but has been reported in 2 to 6% of OPSCC patients [7–9]. Partial p16 expression is not uncommon in OPSCC, and usually consists of patchy nuclear and cytoplasmic staining with < 70% of tumor cells being positive. When the staining is equivocal (i.e. 50 to 70%), as many as 2/3rds of these patients have transcriptionally-active high-risk HPV in their tumors [10, 11]. Here, we present a unique case of HPV16 mRNA positive OPSCC in which large, confluent areas of the tumor were entirely negative for p16 by IHC despite high-risk HPV throughout.

Methods

p16 Immunohistochemistry

IHC for p16 was performed on 4 µm sections from formalin-fixed, paraffin-embedded tissue using the Ventana Benchmark Ultra autostainer with predilute monoclonal antibody E6H4 (Ventana/Roche, Tucson, AZ). Detection utilized heat-induced epitope retrieval (32 min with CC1 buffer) and the OptiView detection kit (Ventana Medical Systems, Inc., Tucson, AZ).

High-Risk HPV RNA In Situ Hybridization

In situ hybridization for high-risk HPV E6/E7 mRNA was performed by Propath Laboratories (Dallas, TX) using the RNAscope 2.5 HD-BROWN Manual Assay (Advanced Cell Diagnostics, Inc., Hayward, CA) targeting HPV-associated RNA in the nucleus and cytoplasm of the target cells. The high-risk HPV RNAscope probe used at ProPath covers HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68. Tissue samples that previously stained positive for high-risk HPV, as well as some that were negative, were used as batch control tissues and reacted appropriately. A positive control was provided on the patient’s tissue for ubiquitin C which served as an internal control for RNA integrity.

HPV Genome Sequencing

DNA was extracted from 2 mm sterile punches taken from tumor tissue blocks and sent to the DCEG-NCI Cancer Genomics Research Laboratory (Rockville, MD) for HPV16 genome sequencing. A custom Thermo Fisher Ion Torrent AmpliSeq HPV16 panel approach was used to amplify the HPV16 genome as previously described [12–14]. This next-generation sequencing assay uses Thermo Fisher Life Sciences’ Ion Torrent S5 in combination with a custom HPV16 Ion Ampliseq panel of 47 multiplexed primers designed to cover the entire viral genome for all HPV16 variant lineages. An in-house developed pipeline was used for all SNP variant calls and variant annotations using the Torrent Variant Caller v.5.0.3 and snpEff v.3.6c [15]. Pipeline settings and parameters can be found at https://github.com/NCI-CGR/cgrHPV16. HPV16 lineage (A, B, C, D) and sublineage assignment (A1-4, B1-4, C1-4, D1-4) was based on the maximum likelihood tree topology using RAxML MPI v7.2.8.27 and 16 HPV16 sublineage reference sequences. In the event that multiple HPV16 variants were detected, the predominant variant was assigned based on presence in at least 60% of the sequence reads. The HPV16 genome sequence from each of the three tumor sections were aligned together and compared at every nucleotide position across the genome. We estimated the number of nucleotide positions across the genome that differed among the samples to determine if these HPV16 samples were the exact same HPV16 isolate (no nucleotide differences). The HPV16 genomes were examined manually to confirm nucleotide differences using the Integrative Genomics Viewer (IGV) [16].

Case Report

Clinical Features

A 63 year old male presented to the emergency department with a several month history of right shoulder pain with an enlarging right sided neck mass. He had a past medical history significant for smoking, type 2 diabetes mellitus, and hypertension. A CT demonstrated a 4.2 × 2.6 cm heterogeneously enhancing mass deep to the sternocleidomastoid involving level II and level III. The right tonsil was noted to be slightly enlarged relative to the left, but no definite primary lesion was identified. Flexible pharyngoscopy showed a normal nasopharynx, oropharynx, and hypopharynx. A fine needle aspiration cytology specimen from the neck mass was negative for malignancy, but the level of clinical suspicion was high, so a right neck dissection (levels II-V) was completed. This revealed metastatic nonkeratinizing squamous cell carcinoma (SCC) in multiple lymph nodes and extranodal tissue including skeletal muscle (see below). A subsequent PET scan demonstrated no clear primary lesion and no evidence of distant metastasis. In particular, thoracic imaging demonstrated no pulmonary lesion at that time. The patient underwent a right radical tonsillectomy and base of tongue resection. Both specimens were entirely submitted for microscopic examination, revealing a 5 mm nonkeratinizing SCC of the right tongue base. He went on to receive postoperative radiation therapy, consisting of 70 Gy in 33 fractions. One year after surgery, he developed a 10 mm right upper lobe lung mass and multiple enlarged thoracic lymph nodes. Endobronchial ultrasound guided needle core biopsies of subcarinal and right lower paratracheal lymph nodes both revealed metastatic squamous cell carcinoma.

Microscopy and p16 IHC

The neck dissection specimen demonstrated metastatic nonkeratinizing SCC (Fig. 1), composed of infiltrative nests of squamous cells associated with a desmoplastic stromal reaction. The tumor cells had scant to moderate eosinophilic cytoplasm, and there were scattered foci of squamous maturation. The tumor had brisk mitotic activity, and there was abundant necrosis. The largest metastatic deposit was 6.0 cm, with extensive involvement of adjacent skeletal muscle. Perineural and lymphovascular invasion were identified. Apart from the largest mass, 32 separate lymph nodes were submitted, 8 of which also showed metastatic SCC. IHC for p16 was completed on all tumor blocks. The neck mass showed strong, confluent nuclear and cytoplasmic positivity throughout approximately 90% of the sampled tumor. However, these areas transitioned abruptly to discrete regions that were entirely negative for p16 (Fig. 1). Scattered regions throughout the tumor also demonstrated weaker partial staining. Similar discrete areas of p16 positivity and negativity were observed on multiple tissue blocks from the neck mass. The components of the tumor that were positive and negative for p16 were similar morphologically with no clearly discernible morphologic correlate to the p16 staining patterns.

Fig. 1.

Sections from the neck mass revealed nonkeratinizing squamous cell carcinoma (A). The majority of the tumor was strongly, diffusely positive for p16 by IHC, but discrete regions were completely negative (B). High power H&E (C) and p16 IHC (D) images taken from the same area demonstrate the interface between the distinct tumor regions

The right base of tongue lesion was a nonkeratinizing SCC with typical nonkeratinizing morphology and no evidence of maturing squamous differentiation. This lesion was completely negative for p16 (Fig. 2). Appropriate staining of internal and external control tissue was confirmed.

Fig. 2.

Nonkeratinizing squamous cell carcinoma of the right tongue base (A), which was negative for p16 by IHC (B). The background mucosa showed a normal patchy pattern of p16 positivity (C)

Core needle biopsies of thoracic lymph nodes demonstrated nonkeratinizing squamous cell carcinoma, consistent with metastases from the oropharyngeal lesion. IHC for p16 was diffusely positive in the sampled tissue.

Additional Studies

RNA ISH for high-risk HPV was conducted on sections from both the base of tongue lesion and the metastatic neck mass. Both lesions were positive for HPV by ISH (Fig. 3). This raised the question of whether the different components of the two lesions may have been involved by different HPV types, leading to their distinct p16 IHC staining patterns. To resolve this, 2 mm punches of tissue blocks were taken from three areas: primary base of tongue tumor, p16 positive metastatic tumor, and p16 negative metastatic tumor. HPV16 genome sequencing of these samples revealed identical HPV16 genomes in all three regions. Given the large amount of HPV16 isolate diversity previously identified among individuals in any given population [13], finding identical HPV16 genomes in all three samples indicates that they were all related to the same causative virus. RNA ISH and HPV16 genome sequencing were not performed on the thoracic metastases.

Fig. 3.

RNA ISH for high-risk HPV was positive throughout the metastatic neck mass (A) and the base of tongue lesion (B)

Discussion

In this unique case of HPV-associated nonkeratinizing SCC, the primary base of tongue tumor was negative for p16, and the metastatic tumor in cervical lymph nodes demonstrated discrete, confluent regions that were either strongly positive for p16 or completely negative. All these regions were positive for the same HPV16, making it unlikely that the metastasis represented a collision tumor from two distinct primary lesions. This is a strange phenomenon that we have not seen in practice, and that has not been reported in the literature.

This case raises several issues worth considering. First, depending on the region sampled, a core biopsy or fine needle aspiration might have demonstrated either p16 positive for negative tumor. This represents a potential diagnostic pitfall, as a specimen with < 70% p16 nuclear and cytoplasmic staining would be interpreted as HPV-negative. The CAP guidelines [4] suggest considering high-risk HPV-specific testing when p16 IHC is equivocal (50–70% nuclear and cytoplasmic staining), which may be helpful in a case like this one, depending on the proportion of the different tumor areas that might have been sampled on biopsy. However, given that the primary tumor was completely negative, if it had been biopsied or resected initially, it could have labeled this patient’s tumor as HPV-independent (by showing negative p16) even though it was driven by transcriptionally-active high-risk HPV.

The second interesting issue is the potential molecular mechanism(s) underlying the variable p16 IHC in this case. Typically, the E7 viral oncoprotein destabilizes the retinoblastoma protein (coded by RB1), which leads to loss of p16 suppression and subsequent p16 overexpression by IHC [4, 17]. While p16 expression by IHC is highly correlated with HPV status in OPSCC, it is understood that this correlation is not perfect. Several mechanisms may account for discordance of p16 IHC and HPV status, including genetic and epigenetic alterations. For example, CDKN2A mutations are relatively frequent in head and neck SCC, and may be associated with p16 overexpression by IHC [17, 18]. Deletion of RB1 is occasionally seen in OPSCC, where it may co-exist with HPV infection and contribute to p16 expression by IHC [19]. Conversely, p16 deletion or promoter hypermethylation may lead to loss of p16 expression in some cases of head and neck SCC [20, 21]. Guidelines recommend p16 IHC as a surrogate for HPV-specific testing because p16 is inexpensive, readily available, and reasonably accurate in OPSCC, but it is important to be aware of the various settings in which p16 and HPV-specific testing may be discrepant.

A final interesting issue is the prognostic implication of variable p16 staining as we saw in this case. Positivity for p16 generally confers a favorable prognosis regardless of HPV status [5, 7]. However, it is not clear how this would apply to the current case. Would prognosis be driven by the majority of the tumor cells (p16 +) or by the minority population (p16-)? Is this impacted by the status of the primary tumor (p16-)? In tumors with patchy p16 staining, might prognosis be assessed by completely different molecular features? The answers to these questions are unknown, and—given the rarity of cases like this one—would be exceptionally difficult to study. What does seem clear is that patients with p16 and HPV specific testing discrepant tumors do not have the same favorable prognosis across the board as do patients with p16 positive/high-risk HPV positive tumors [7–9]. In our patient, where p16 is still expressed in the nuclei and cytoplasm of > 70% of the overall tumor area, and where HPV16 mRNA is detectable by in situ hybridization and HPV16 DNA by next generation sequencing, the favorable outcome may be maintained. However, his extensive nodal involvement and metastatic disease suggests that his prognosis may be more guarded.

The principles of p16 and HPV testing in OPSCC are complex and they continue to evolve as our understanding improves. The current case highlights a rare pattern of confluent, block-like loss of p16 expression—including complete absence in the primary base of tongue tumor—in an HPV-associated OPSCC, raising additional questions about the pathogenesis and pathologic assessment of these tumors.

Author Contributions

Surgical care was provided by MJC. Pathology investigation was performed by MJB, JSL Jr., and SAR. Genome sequencing and interpretation were performed by LM, SB, and MY. The first draft of the manuscript was written by SAR. All authors contributed to subsequent revision of the manuscript.

Funding

This study was funded by the intramural research program of the Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH.

Data Availability

Not applicable.

Code Availability

https://github.com/NCI-CGR/cgrHPV16.

Declarations

Conflict of interest

The authors have no conflicts of interest to declare.

Ethical approval

Data was collected in accordance with guidelines from the institutional Research Ethics Board.

Informed Consent

Written informed consent was obtained from the patient for data collection and publication of this case report and accompanying images.