HPV-associated Oropharyngeal Squamous Cell Carcinoma: An Imaging Framework for AJCC Version 9 TNM Staging

Abstract

Human papillomavirus (HPV)–associated oropharyngeal squamous cell carcinoma (OPSCC) is biologically and prognostically distinct from HPV-negative disease, and staging directly impacts treatment. The American Joint Committee on Cancer (AJCC) and Union for International Cancer Control version 9 TNM staging system incorporates unequivocal imaging-detected and/or clinically detected extranodal extension (iENE) into clinical nodal staging and definitive pathologic extranodal extension into pathologic nodal staging. Additionally, it provides more explicit anatomic definitions for advanced local disease (clinical T4) while maintaining T and M category definitions from the eighth edition. Radiologists play a central role in the multidisciplinary team by evaluating endophytic tumor extent, regional nodal disease and iENE, and distant metastasis using CT, MRI, and PET. This article reviews clinically relevant oropharyngeal anatomy and proposes a practical radiologic approach to assigning T, N, and M categories in HPV-associated OPSCC. It emphasizes cautious interpretation of iENE because of interreader variability and imperfect diagnostic performance, with further high-grade evidence needed to establish its consistently reproducible, incremental prognostic value as a clinical N category modifier. It also highlights multidisciplinary factors that guide selection of surgical resection versus nonsurgical therapy. Posttreatment evaluation is discussed in the context of blood-based biomarkers such as tumor tissue–modified viral HPV DNA. The goal of this article is to provide a practical, staging-focused imaging framework for evaluation of HPV-associated OPSCC.

Keywords: CT, MR Imaging, Head/Neck, Neck, Pharynx, Surgery, Human Papillomavirus, Squamous Cell Carcinoma, Oropharynx, Transoral Robotic Surgery, Oropharyngeal Carcinoma, TTMV HPV DNA

© RSNA, 2026

Summary

The American Joint Committee on Cancer/ Union for International Cancer Control version 9 TNM staging system for human papillomavirus (HPV)–associated oropharyngeal squamous cell carcinoma provides an updated framework for radiology reporting, incorporating cautious assessment of unequivocal imaging-detected extranodal extension and acknowledging the emerging, complementary role of circulating tumor tissue–modified viral HPV DNA in surveillance.

Essentials

• ■ The American Joint Committee on Cancer/ Union for International Cancer Control version 9 TNM staging system incorporates unequivocal imaging-detected or clinically detected extranodal extension (ENE) into the clinical N category, while biopsy-proven ENE is incorporated into the pathologic N category assignment for human papillomavirus (HPV)–associated oropharyngeal squamous cell carcinoma; however, imaging reporting is variable, thus requiring cautious, standardized evaluation to avoid misclassification.

• ■ CT and MRI help evaluate primary tumor extent and nodal disease, which guides the multidisciplinary team in determining treatment.

• ■ Baseline posttreatment fluorine 18 fluorodeoxyglucose PET/CT is often performed at approximately 12 weeks and, if available, should be interpreted in the context of circulating tumor tissue–modified viral HPV DNA results.

Introduction

Human papillomavirus (HPV)–associated oropharyngeal squamous cell carcinoma (OPSCC) has become the predominant OPSCC subtype in many developed countries, with contemporary estimates reaching up to nearly 90% in selected U.S. tertiary cancer center cohorts, and is biologically and clinically distinct from HPV-negative disease (1–3). Compared with traditional tobacco- and alcohol-associated OPSCC, HPV-associated tumors predominantly affect younger male individuals, carry a more favorable prognosis, and commonly manifest with small or clinically occult primaries and nodal-predominant disease (3). Although HPV-associated OPSCC has historically been associated with younger patients (40s to early 50s) at diagnosis, recent data point to a "birth cohort effect," with disease burden shifting toward older adults and an increasing median age at diagnosis (4–8). These distinctions are reflected in prior staging systems, including the American Joint Committee on Cancer (AJCC) staging system eighth edition, which separated HPV-associated from non-HPV OPSCC. In 2026, however, the practical challenge for radiologists lies in applying the updated AJCC/Union for International Cancer Control TNM staging system version 9 (AJCC v9/TNM 9) criteria consistently in daily practice, particularly for assessment of local tumor extent, nodal disease, and imaging-detected extranodal extension (iENE) within multidisciplinary care.

The purpose of this article is to address a practical gap between staging system updates and day-to-day radiology workflow by reviewing clinically relevant oropharyngeal anatomy and proposing a systematic radiologic approach to assigning T, N, and M categories according to AJCC v9/TNM 9 in HPV-associated OPSCC. Particular emphasis is placed on iENE, including areas of ongoing controversy and limitations in the current version, as well as multidisciplinary factors that guide surgical resection versus nonsurgical therapy. Posttreatment evaluation is also discussed, including the role of blood-based biomarkers such as tumor tissue–modified viral (TTMV) HPV DNA.

Pathophysiology and Histopathologic Features

HPV-associated OPSCC is primarily driven by two high-risk viral oncoproteins (E6 and E7), resulting in marked p16 overexpression (critical for cell survival), which is widely used as a surrogate marker for HPV-driven disease (9). Because p16 is not entirely specific (eg, expression may also be observed in cutaneous and pulmonary squamous cell carcinomas), HPV attribution may require confirmatory testing with HPV DNA or RNA in situ hybridization when clinically relevant. High-risk HPV DNA polymerase chain reaction alone is not sufficient for staging assignment according to AJCC v9/TNM 9 (10,11). p16 positivity is defined as strong, diffuse nuclear and cytoplasmic staining in more than 70% of tumor cells (12). HPV-associated OPSCCs, compared with HPV-negative tumors, typically retain wild-type p53, have a lower overall mutational burden with specific genomic alterations (eg, PIK3CA mutations, loss of TRAF3), and exhibit an "inflammatory" immune microenvironment with abundant tumor-infiltrating CD8+ T cells (13,14). These biologic differences underlie increased intrinsic radiosensitivity, a more favorable prognosis, and the distinct staging system for HPV-associated OPSCC (11,15). HPV-driven tumor cells undergoing apoptosis shed viral DNA fragments into the circulation as circulating tumor HPV DNA, and newer approaches, such as TTMV HPV DNA assays, leverage droplet digital polymerase chain reaction with fragmentation pattern analysis to improve specificity for TTMV HPV DNA. These assays are now used in the pre- and posttreatment setting (16,17).

The histopathology of HPV-associated OPSCC differs fundamentally from that of HPV-negative disease in terms of cellular morphology and nodal presentation. HPV-associated OPSCC typically demonstrates nonkeratinizing, basaloid morphology with high nuclear-to-cytoplasmic ratios, syncytial cytoplasm, and absence of keratin pearls. In contrast, HPV-negative OPSCC is usually keratinizing squamous cell carcinoma with keratin pearl formation and intercellular bridges (6,12). Nodal metastases may show cystic or necrotic change, although this imaging appearance is variable (18,19). However, despite these histologic differences, immunohistochemistry, most commonly p16 staining, is often required to support the diagnosis of HPV-associated OPSCC.

Oropharynx Anatomy

Knowledge of the oropharyngeal boundaries and adjacent critical structures is essential for precise localization of the primary tumor and identification of specific routes of deep extension on HPV-associated OPSCC imaging studies. The oropharynx is located between the nasopharynx and hypopharynx and lies posterior to the oral cavity, extending from the soft palate to the inferior boundary of the hyoid bone or vallecular floor. Anteriorly, it is bordered by the circumvallate papillae of the tongue and the junction of the hard and soft palate. Laterally, it is bounded by the middle layer of the deep cervical fascia, which lies deep to the superior and middle pharyngeal constrictor muscles (11). Posteriorly, it is bounded by the posterior pharyngeal wall. HPV-associated OPSCC can arise from various subsites within the oropharynx, most commonly the base of the tongue (defined as the posterior third of the tongue and the lingual tonsil) and the palatine tonsils (including the tonsillar fossa and anterior and posterior tonsillar pillars), followed by the inferior mucosal surface of the soft palate and uvula and the lateral and posterior walls of the oropharynx (20).

Other adjacent structures important for accurate local-regional evaluation and treatment planning include the extrinsic tongue musculature (genioglossus, hyoglossus, styloglossus, and palatoglossus muscles), the lingual surface of the epiglottis, the pterygoid muscles and plates, the retromolar trigone, and adjacent osseous structures such as the hard palate, mandible, and skull base. Laterally, the carotid artery is a key landmark for assessing tumor encasement (20). Pertinent radiologic anatomy of the oropharynx and surrounding structures is shown in Figure 1.

Figure 1:

Images in a 35-year-old male patient with normal CT scan of the oropharynx. (A) Sagittal contrast-enhanced CT image demonstrates boundaries of the oropharynx (outlined area). (B) Axial contrast-enhanced image demonstrates the bilateral palatine tonsils (stars), posterior oropharyngeal wall (dashed arrow), and bilateral pterygoid muscles (solid arrows). (C) Axial contrast-enhanced image just inferiorly demonstrates the tongue base (outlined area, posterior third of the tongue); extrinsic tongue muscles include hyoglossus muscle (solid arrows) and genioglossus muscle (dashed arrow).

Multidisciplinary Evaluation Using AJCC v9/TNM 9 Staging Protocol

Accurate AJCC staging of OPSCC requires biopsy-proven HPV-positive status, established with p16 immunohistochemistry and/or positive high-risk HPV RNA or DNA in situ hybridization, depending on the diagnostic context. Once HPV status is confirmed, clinical examination with flexible fiberoptic endoscopy and, when needed, panendoscopy is used to define the primary site and exophytic tumor extent (11,21). Otolaryngologists, radiation oncologists, and medical oncologists assess mucosal extent of disease, functional status, regional lymphadenopathy, and clinical evidence of extranodal extension (ENE), including nodal fixation. Radiologists within the multidisciplinary team use CT or MRI to evaluate extent of endophytic disease and integrity of deep structures, as well as iENE, a key adverse feature incorporated into the N category of the AJCC v9/TNM 9 staging system (11). Multidisciplinary assessment is therefore fundamental to staging and treatment planning. Emerging blood-based assays, such as TTMV HPV DNA, are increasingly incorporated into selected clinical workflows; however, their optimal use remains under investigation, and they are not part of AJCC v9/TNM 9 staging (17,22).

Clinical (c) staging is assigned based on history, physical examination, endoscopy, imaging, and biopsy findings. It provides a pretreatment framework, but selection of transoral surgery versus nonsurgical therapy depends on resectability, surgical exposure, expected functional morbidity, and nodal disease pattern, as summarized in Table 1. The cT0 category includes patients presenting with a p16-positive cervical node and no detectable primary tumor at examination or imaging (11). Pathologic (p) staging, in contrast, is reserved for patients undergoing primary surgery and is derived from histologic assessment of the resected primary tumor and lymph nodes. It refines risk by quantifying nodal number, level, and pathologic ENE (pENE) to guide adjuvant therapy.

Table 1:

Factors for Treatment Decision-Making of TORS versus NSOPT for Human Papillomavirus-associated Oropharyngeal Squamous Cell Carcinoma

Factor	Favors TORS	Favors NSOPT
Primary tumor size and T category	T1–T2 and carefully selected small T3 lesions	Bulky T3–T4 disease
	En bloc oncologic resection with adequate margins feasible	Unlikely to achieve clear margins
	Limited volume tongue base or tonsillar primaries
Tongue base involvement	Involvement of <50% of deep tongue base musculature	Tumor requiring resection of ≥50% of deep tongue base musculature (high risk of severe dysphagia and aspiration)
Soft palate/tonsillar pillar/lateral wall	Lateral tonsillar or soft palate tumors with good exposure	Extensive posterior pharyngeal wall involvement (>50%)
	Limited posterior pharyngeal wall involvement (<50%)	Circumferential pharyngeal involvement
	No need for major reconstructive flap
Parapharyngeal space/carotid proximity	Minimal parapharyngeal fat involvement	Deep extension into parapharyngeal fat
	Clean margin from carotid artery and major neurovascular structures	Carotid artery encasement
		RP/medialized carotid artery
Mandible/maxilla/hard palate	No osseous invasion	Radiologic invasion of mandible, maxilla, or hard palate (often requires open composite resection)
Larynx/hypopharynx	Mucosal abutment onto lingual epiglottis only	Laryngeal or hypopharyngeal invasion
Skull base/prevertebral fascia	No skull base or prevertebral involvement	Involvement of skull base, prevertebral fascia, or paraspinal musculature
Nodal disease	Unilateral, amenable to selective neck dissection	Grossly unresectable nodal disease
	No RP nodes	RP nodes involved
	Five nodes or fewer involved	Nodal carotid artery encasement
		Nodal disease likely to mandate adjuvant chemoradiation
Exposure/access	Good mouth opening	Severe trismus due to tumor
	Favorable mandibular width and dental anatomy	Bulky tongue
	Adequate neck extension allowing full TORS exposure	High-arched or narrow mandible, torus mandibularis, or cervical spine disease preventing optimal positioning

Source.–References 47, 55–59.

Note.–NSOPT = nonsurgical oropharyngeal therapy, RP = retropharyngeal, TORS = transoral robotic surgery.

T Category

The eighth edition of the AJCC staging manual introduced a dedicated staging system for HPV-associated OPSCC (p16 positive), acknowledging its distinct biology and prognosis relative to HPV-negative disease. The Tis (carcinoma in situ) category was eliminated, and T4a/T4b were consolidated into a single T4 category. AJCC v9/TNM 9 maintains the same T category definitions for HPV-associated OPSCC as the eighth edition, including the same T0–T4 structure for both clinical and pathologic staging. However, for cT4 disease, version 9 provides a more explicit anatomic description of advanced local invasion, including structures not clearly listed in the abbreviated AJCC eighth edition wording, and specifies that cT4 assignment based solely on extrinsic oral tongue muscle involvement requires unequivocal imaging evidence supported by clinical findings (11).

The current T category definitions are as follows: T0, no evidence of primary tumor (HPV-associated OPSCC metastatic cervical node, typically at level II, with unknown primary); T1, tumor ≤ 2 cm in greatest dimension; T2, tumor > 2 cm but ≤ 4 cm in greatest dimension; T3, tumor > 4 cm in greatest dimension or involvement of the lingual surface of the epiglottis; and T4, advanced local disease with invasion of the larynx, deep/extrinsic tongue muscles (genioglossus, hyoglossus, palatoglossus, and styloglossus), pterygoid muscles (medial and/or lateral) or plates (medial and/or lateral), hard palate, mandible, nasopharynx, skull base, or carotid artery encasement (11). This differs from the AJCC staging manual eighth edition, in which cT4 disease was described more succinctly as moderately advanced local invasion of the larynx, extrinsic tongue muscles, medial pterygoid, hard palate, mandible, or beyond (23).

Several imaging pitfalls exist and can be mitigated through multidisciplinary tumor board discussion. Small T1 tumors can be difficult to visualize with imaging (CT or MRI) alone, and correlation with endoscopy as well as biopsy findings is essential; fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) PET/CT may be useful when there is clear focal hypermetabolic activity (Fig 2). T2 tumors are generally well-defined at imaging, and the role of imaging is to identify unexpected features that may increase the T category or alter management (Fig 3). Base of tongue and vallecular primaries may extend onto the lingual surface of the epiglottis. Mucosal spread confined to the lingual epiglottis is classified as T3 disease and should not be mistaken for laryngeal invasion. In addition, tumors abutting the lingual epiglottis may be difficult to distinguish from infiltration at imaging; in these cases, direct visualization is essential for determining T2 versus T3 disease when tumor size is between 2 and 4 cm (Fig 4). Distinguishing T1 from T2 or T3 tumors may also be difficult with imaging alone because superficial mucosal spread can be underestimated. This makes direct visualization or endoscopic assessment necessary for accurate measurement of maximum tumor dimension and highlights the importance of a multidisciplinary approach (11).

Figure 2:

Images in a 55-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma. (A) Axial contrast-enhanced CT image demonstrates a subtle enhancing mass within the right palatine tonsil, which measured 1.9 cm (arrow), consistent with T1 disease. (B) Axial fused PET/CT image demonstrates corresponding hypermetabolic activity (arrow), supporting localization of the primary tumor.

Figure 3:

Image in a 68-year-old male patient with biopsy-proven human papillomavirus–associated oropharyngeal carcinoma. Axial contrast-enhanced CT image demonstrates a mass within the right palatine tonsil measuring 2.5 cm in maximum dimension (arrow), consistent with T2 disease. There was no extension along the lingual surface of the epiglottis, involvement of the extrinsic tongue musculature, or extension beyond the oropharynx. Enlarged, heterogeneous lymph node with loss of normal morphology is seen at right level II, consistent with metastatic nodal disease (arrowhead).

Figure 4:

Images in a 43-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma. (A) Axial postcontrast image demonstrates predominantly exophytic tumor centered within the deep tongue base (solid arrows), filling the vallecula (dashed arrow). There is a metastatic left level II lymph node (arrowhead). (B) Sagittal postcontrast image shows tumor at the tongue base (solid arrow) with extension along the lingual surface of the epiglottis (dashed arrow), confirmed with direct visual inspection, consistent with T3 disease.

Invasion of extrinsic tongue muscles is a defining criterion for T4 disease. As per AJCC v9/TNM 9, a cT4 designation on this basis should be reserved for unequivocal imaging findings, ideally corroborated by clinical signs such as tongue tethering, fixation, or deviation. In cases of equivocal findings at CT, MRI can be useful for confirming invasion of extrinsic tongue muscles (Fig 5). MRI is best viewed as complementary for local T category assignment, particularly when superior soft tissue contrast is required to assess deep tumor extent or to adjudicate equivocal invasion of adjacent muscles, laryngeal structures, or skull base–related soft tissues.

Figure 5:

Images in a 74-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma. (A) Axial contrast-enhanced CT image demonstrates an infiltrative mass centered within the right oropharynx (arrowhead) with extension into the right genioglossus muscle (solid arrow) and right hyoglossus muscle (dashed arrow). (B) Axial contrast-enhanced fat-saturated T1-weighted image confirms invasion of the right genioglossus muscle (solid arrow) and right hyoglossus muscle (dashed arrow), compatible with T4 disease. Deep extension into the extrinsic tongue muscle is a contraindication for transoral robotic surgery.

N Category

In the AJCC staging manual eighth edition, HPV-associated OPSCC was separated from non-HPV OPSCC and assigned a distinct nodal staging framework based primarily on nodal size and laterality. Early data demonstrated that HPV-associated disease had a more favorable clinical course than HPV-negative disease, and there was limited evidence supporting ENE as a prognostic factor in HPV-associated OPSCC, as it had not been studied as extensively as in non-HPV OPSCC. However, accumulating data have since shown that ENE remains an important independent prognostic factor in HPV-associated OPSCC, particularly for distant metastasis and overall survival, prompting AJCC v9/TNM 9 to incorporate ENE into clinical/imaging and pathologic N category assignment (24). Both pN2 disease and ENE have been shown to be independent prognostic factors (25,26).

In AJCC v9/TNM 9, N category assignment for HPV-associated OPSCC continues to depend on nodal size and laterality but now also incorporates unequivocal clinically evident ENE or iENE (11). Unequivocal clinically evident ENE or iENE increases the clinical N category by one level. The updated categories are as follows: cN1, ipsilateral nodes ≤ 6 cm (upstaged to cN2 if iENE is present); cN2, bilateral or contralateral nodes ≤ 6 cm (upstaged to cN3 if iENE is present); and cN3, nodes > 6 cm with or without iENE (11). This update emphasizes the role of radiology in identifying iENE because it can potentially impact staging and treatment. Importantly, although iENE and pathologic ENE are distinct entities, both are independently associated with higher rates of distant metastasis and worse overall survival in HPV-associated OPSCC, and their adverse impact persists even when discordant (27,28).

Pathologic N category assignment in AJCC v9/TNM 9 has also been revised. In the AJCC staging manual eighth edition, pN assignment for HPV-associated OPSCC was based solely on the number of metastatic lymph nodes, with laterality, nodal size, and ENE deliberately excluded; most surgically treated patients were therefore grouped as pN1/stage I (23). In contrast, AJCC v9/TNM 9 incorporates pENE into the nodal classification, recognizing it as an adverse prognostic factor and using it, together with nodal burden, to redefine pN categories: pN1 is limited to one to four nodes without pENE (subdivided into pN1a [one node] and pN1b [two to four nodes]), pN2 includes one to four nodes with pENE or more than four nodes without pENE, and pN3 includes more than four nodes with pENE (11).

In AJCC v9/TNM 9, unequivocal iENE, first introduced for nasopharyngeal carcinoma staging and now applied to HPV-positive OPSCC (29), is explicitly accepted as a basis for clinical nodal staging and does not require histopathologic confirmation. Larger nodes (>3 cm) have a higher likelihood of ENE, but size alone is not diagnostic, and necrosis, cystic change, and capsular thickening should not be used as surrogates for iENE (24,30). Contrast-enhanced CT or MRI may serve as the primary cross-sectional anatomic modality for evaluating cervical nodal stations and possible iENE, while ¹⁸F-FDG PET/CT plays a complementary role by identifying metabolically active but morphologically subtle nodal disease (31). In HPV-associated OPSCC, however, as mentioned previously, nodal metastases may show cystic/necrotic change and variable FDG uptake, making contrast-enhanced CT essential for depicting their characteristic morphology (11).

Two complementary consensus-based radiologic frameworks have been proposed for iENE assessment. First is the international Delphi consensus, which has provided standardized broad criteria and a four-tier grading system for iENE, emphasizing that ENE lies on a continuum from subtle fat infiltration to overt invasion of adjacent structures. The consensus recommends that radiology reports comment on indistinct or irregular nodal margins; extension of tumor into perinodal fat; conglomerate, matted, or coalescent nodes with loss of intervening fat planes; and extension into adjacent muscles (typically the sternocleidomastoid), salivary glands, skin, or neurovascular bundle (30). The grades are defined as follows: grade 0, no iENE; grade 1, irregular or ill-defined margins or limited extension into perinodal fat; grade 2, tumor extending beyond the margins of two or more adjoining nodes forming a conglomerate; and grade 3, unequivocal invasion into surrounding structures.

Second is the Asian-Oceanian Society of Neuroradiology and Head & Neck Radiology, American Society of Head and Neck Radiology, and European Society of Head and Neck Radiology task force radiologic framework, referenced in AJCC v9/TNM 9 for HPV-associated OPSCC for identifying iENE (32). This framework similarly defines four categories (no iENE, perinodal fat invasion, coalescent nodes, and adjacent structure invasion) and limits grading to definitive radiologic evidence. It also provides structure-specific guidance to reduce overcalling (eg, fat-plane loss alone is insufficient for muscle invasion; arterial involvement requires near-circumferential encasement; venous involvement requires tumor thrombus) and includes a checklist for coalescent nodes and common pitfalls (treated or inflamed nodes, postbiopsy changes, and single necrotic nodes mimicking conglomerates) (32). Although AJCC v9/TNM 9 does not incorporate consensus iENE grades into N category definitions, it explicitly requires unequivocal clinical ENE and/or unequivocal iENE for cN category assignment and definitive pENE for pN category assignment.

Both these frameworks are useful for standardizing terminology, but neither should be interpreted as proof that iENE is reproducibly identified at routine imaging. Use of consensus-based criteria is encouraged to standardize reporting of imaging features. Published data support this cautious approach. Tran et al (33), in a multicenter study of 75 consecutive HPV-associated OPSCC cases, reported poor reproducibility of previously described CT features among five expert head and neck neuroradiologists despite standardized training (Fleiss κ < 0.6). Similarly, Sahin et al (34), in a prospective blinded study of 34 expert observers across radiology, surgery, and radiation oncology, found weak interobserver agreement (κ < 0.6), with no specialty performing better than chance. Earlier aggregate evidence, including the meta-analysis by Park et al (35), also demonstrated only moderate diagnostic performance for CT and MRI, with substantial heterogeneity; notably, CT specificity was lower in HPV-associated OPSCC than in HPV-negative oral cavity cancer or mixed head and neck squamous cell carcinoma cohorts. More recently, Ebrahimi et al (36) demonstrated substantial prognostic heterogeneity among patients with ENE with nodal metastatic head and neck cutaneous squamous cell carcinoma, suggesting that staging systems may overweight ENE when applied without sufficient context from other clinicopathologic factors. Accordingly, although AJCC v9/TNM 9 and recent international consensus statements provide a structured lexicon for iENE assessment, their adoption should not be interpreted as evidence of robust radiologic reproducibility.

Because iENE now modifies cN category, the threshold for diagnosis should be high (11,30), and multidisciplinary discussion is essential to establish a consistent institutional framework for reporting and to ensure uniform patient care. For cN category assignment purposes, only unequivocal iENE should be considered positive; at our institution, this corresponds to the highest-grade category (grade 3), defined as clear extension into adjacent musculature, salivary glands, cutaneous soft tissue, or the neurovascular bundle (Fig 6). By contrast, equivocal or suggestive findings (eg, capsular irregularity or fat stranding) should be considered ENE-negative to avoid stage migration and preserve prognostic discrimination (11,37). Lymph nodes may indent adjacent muscle with apparent loss of the intervening fat plane, which should not be interpreted as iENE (Fig 7). Recent core biopsy is another important pitfall, as procedure-related perinodal fat stranding and nodal irregularity can mimic iENE; careful correlation with procedural history is essential.

Figure 6:

Images in a 65-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma and metastatic nodal disease. (A) Axial contrast-enhanced CT image demonstrates a left level II conglomerate nodal mass with clear infiltration into the adjacent sternocleidomastoid muscle (arrowheads) and partial encasement of the left internal carotid artery (arrow). (B) Coronal contrast-enhanced CT image demonstrates extension of nodal disease inferiorly into the sternocleidomastoid muscle (arrowheads). There is also tumor thrombus within the left internal jugular vein (solid arrow). Note atrophy of the left sternocleidomastoid muscle (dashed arrow), suggestive of denervation change secondary to perineural invasion of the left spinal accessory nerve. Definitive extension into adjacent musculature or neurovascular bundle (tumor thrombus, encasement of the carotid artery, or perineural invasion) is considered unequivocal imaging-detected extranodal extension.

Figure 7:

Image in a 55-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma and metastatic nodal disease. Axial postcontrast image demonstrates a necrotic metastatic right level II lymph node (solid arrow). There is loss of the normal fat plane with the adjacent sternocleidomastoid muscle (dashed arrows) with no definite infiltration into the muscle. There is also mass effect on the adjacent right internal jugular vein (arrowhead). None of these features qualify for imaging-detected extranodal extension (ENE). Patient underwent surgical resection; pathology findings confirmed no ENE of the metastatic lymph node.

Clinically, patients with limited nodal disease and no iENE constitute the lowest-risk group and are most suitable for de-escalation strategies in primary radiation therapy or surgical settings. Phase 3 data have shown that transoral robotic surgery (TORS) and de-escalated adjuvant radiation therapy in this cohort achieve equivalent event-free survival with a statistically significant reduction in severe (grade ≥ 3) treatment-related toxicity (25,38). In contrast, the presence of iENE identifies patients who should generally be excluded from de-escalation trials and managed with more intensive systemic and radiation therapy (24,27). By shifting the N category from a node count– and size-based paradigm to one incorporating iENE, AJCC v9/TNM 9 underscores the central role of radiologic evaluation in multidisciplinary decision-making. Future refinements will likely require high-quality evidence demonstrating that iENE can be identified reproducibly and that it provides incremental prognostic value beyond established clinical and pathologic factors.

M Category

AJCC v9/TNM 9 preserves the definitions of distant metastases established in the eighth edition: cM0, no distant metastasis identified at history, clinical examination, or imaging; and cM1, distant metastasis detected clinically or radiologically (11). pM1 denotes the presence of distant metastasis confirmed microscopically, including tumor identified at biopsy, resection specimen, or cytology obtained through fine-needle aspiration. Importantly, pM1 assignment does not require microscopic confirmation of every distant metastatic site.

Although the formal M definitions are unchanged, detection of recurrent or metastatic disease remains highly important because prognosis in HPV-associated OPSCC varies substantially according to metastatic burden and distribution. Patients with solitary or oligometastatic recurrence have better outcomes (5-year overall survival: ~46%) than those with multisite metastatic relapse (5-year overall survival: ~9%), making accurate detection and characterization of recurrence clinically relevant (39). In practice, imaging remains central to assessing metastatic disease because recurrent HPV-associated OPSCC has a substantial distant-failure component (often comparable to local-regional failure), and distant metastases may occur later and at atypical sites. Chest CT and ¹⁸F-FDG PET are the main tools for detecting pulmonary and osseous metastases (Fig 8), which are the most common sites in HPV-associated OPSCC. Abdominal, pelvic, or brain involvement is uncommon in the absence of lung metastases (11). Compared with HPV-negative disease, distant metastases in HPV-associated OPSCC may manifest beyond 2 years and can involve less typical sites (40). Blood-based biomarkers such as TTMV HPV DNA correlate with metastatic burden; higher levels have been identified in patients with two or more metastatic sites, and persistent detectability at last follow-up portends worse survival. At present, these biomarkers function as adjuncts rather than replacements for imaging in assigning M category, and their optimal use remains under investigation (17).

Figure 8:

Image in a 65-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma. Coronal maximum intensity projection image from fluorodeoxyglucose PET shows the primary oropharyngeal lesion (arrowhead) and left neck metastatic nodal disease (solid arrow). Extensive metastatic disease within bone (dashed arrows), liver (solid star), and lungs (hollow star).

From a therapeutic standpoint, the presence of M1 disease generally shifts management from curative local-regional strategies to systemic or palliative approaches. Most patients with widespread metastatic disease are treated with systemic regimens such as anti–PD-1 immunotherapy and/or platinum-based chemotherapy, with cross-sectional imaging and ¹⁸F-FDG PET used to assess treatment response and detect progression (41). In carefully selected patients with oligometastatic disease (most commonly, limited lung metastases), metastasis-directed therapies such as surgical resection or stereotactic body radiation therapy may be considered, often in conjunction with systemic therapy (39). Accurate radiologic characterization of the number, size, and distribution of metastases is therefore essential not only for assigning M category and disease staging but also for identifying candidates for metastasis-directed approaches versus systemic therapy alone.

Prognostic Stage Groups

Clinical stage grouping in AJCC v9/TNM 9 remains unchanged from the eighth edition. Stage I includes T0/N1/M0 and T1–T2/N0–N1/M0. Stage II includes T0–T2/N2/M0 and T3/N0–N2/M0. Stage III includes T0–T3/N3/M0 and T4/any N/M0, and stage IV includes any T/any N/M1 (11).

A key structural principle carried forward into AJCC v9/TNM 9 from the eighth edition is that stage IV is reserved exclusively for patients with M1 disease. Locally or regionally advanced HPV-associated OPSCC (including T4 or N3) is classified as stage III; only the presence of distant metastasis qualifies as stage IV, underscoring the binary impact of M status on overall stage grouping (11).

The revised pN categories also drive corresponding changes in prognostic stage grouping: Stage I remains T0–T2, pN0–N1, M0; stage II includes T0–T2 tumors with pN2–pN3 or T3 tumors with pN0-pN2; stage III comprises T3 tumors with pN3 or any T4 tumor; and stage IV includes any T/any N/M1 (11).

Optimizing Radiology Reports for Staging and Treatment

Imaging evaluation of HPV-associated OPSCC can be challenging; radiologists can improve proficiency by aligning their search patterns with the AJCC staging system and systematically evaluating anatomy according to the T, N, and M categories. Begin by identifying the tumor subsite within the oropharynx and measuring its greatest dimension. The tumor should then be assessed in all directions to define its peripheral margins. Posteriorly, evaluate for extension along the lingual surface of the epiglottis. Anteriorly, assess involvement of the extrinsic tongue musculature and hard palate. Laterally, evaluate for extension into the pterygoid muscles, pterygoid plates, and mandible and for carotid artery encasement. Superiorly, assess involvement of the nasopharynx and skull base. Inferiorly, evaluate for extension into the larynx.

Lymph node assessment should follow, beginning with level II and III nodal stations and extending to other nodal groups bilaterally, including the retropharyngeal region. Retropharyngeal nodes are particularly important because they are not routinely dissected at the time of TORS; their involvement is associated with a higher risk of regional and distant failure and may favor primary chemoradiation over upfront surgery (15). If ¹⁸F-FDG PET has been performed, hypermetabolic activity can identify metastatic lymph nodes that appear normal at CT or MRI based on size and morphology. Asymmetrically prominent ipsilateral lymph nodes are often metastatic, and ¹⁸F-FDG PET can help confirm nodal disease. If ¹⁸F-FDG PET has not been performed or is not scheduled, its use should be recommended in the report impression to ensure complete staging. At a minimum, the report impression should describe the primary tumor, including involved structures that determine the highest T category, and the presence or absence of lymph node involvement. If nodal disease is present, the report should document location (ipsilateral vs contralateral to the primary tumor or retropharyngeal), number of abnormal lymph nodes, and presence or absence of ENE. When ¹⁸F-FDG PET is available, the impression may also include assessment for distant metastatic disease, providing a complete radiologic evaluation of T, N, and M categories. Table 2 provides a reporting checklist for radiologic evaluation of HPV-associated OPSCC.

Table 2:

Radiologic Checklist for Reporting Human Papillomavirus-associated Oropharyngeal Squamous Cell Carcinoma

Staging Component	Reporting Elements
Tumor extent
General (all T categories)	Primary site: palatine tonsil, ATP, PTP, base of tongue, vallecula, GTS, soft palate/uvula, lateral/posterior pharyngeal wall
	Side: right/left
	>1 cm from midline: yes/no
	Biopsy changes: yes/no
T1-T2	Greatest dimension: (≤2 cm; >2–4 cm)
T3	Greatest dimension: (>4 cm)
	Extension into lingual surface of epiglottis
T4	Deep/extrinsic tongue muscles (genioglossus, hyoglossus, palatoglossus, styloglossus)
	Pterygoid muscles or pterygoid plates (medial and/or lateral)
	Hard palate or mandible
	Larynx (true invasion)
	Nasopharynx
	Skull base or central skull base foramina
	Internal carotid artery encasement
Nodal disease	Laterality: ipsilateral, contralateral, bilateral
	RP node presence: yes/no
	Quantity
	Largest diameter (≤6 cm vs >6 cm)
	iENE presence: unequivocal present/absent
Metastatic disease	Absent or present
	Site(s) (ie, lung, bone, liver, brain, etc)

Note.–ATP = anterior tonsillar pillar, iENE = imaging-detected extranodal extension, GTS = glosso-tonsillar sulcus, PTP = posterior tonsillar pillar, RP = retropharyngeal.

Familiarity with treatment options and contraindications allows radiologists to add value to the report. TORS enables resection of oropharyngeal carcinoma while preserving speech and swallowing function and is generally offered for T1 and T2 tumors with a high likelihood of negative surgical margins. Importantly, TORS addresses the primary oropharyngeal tumor only; cervical nodal disease requires separate management with neck dissection, radiation therapy, and/or chemoradiation, depending on nodal burden and pathologic risk features. TORS may be considered for selected T3 tumors when resection is unlikely to result in substantial functional deficit. In contrast, T4 disease, poor exposure, anticipated positive margins, grossly unresectable nodal disease, or unequivocal clinical ENE/iENE are most often managed nonsurgically (42). The presence of ENE reduces the appeal of upfront surgery because many such patients will require adjuvant chemoradiation after resection, resulting in trimodality therapy (43). Additional anatomic features to document in the radiologic report, beyond those used for T category assignment, include involvement of the soft palate, tumor distance from midline, tumor abutment of the carotid artery, and presence of a retropharyngeal carotid artery (Fig 9). Torus mandibularis and cervical spine disease may limit optimal patient positioning (44).

Figure 9:

Images in a 46-year-old male patient with human papillomavirus–associated oropharyngeal carcinoma and trismus. (A) Axial contrast-enhanced CT image demonstrates tumor centered within the right oropharynx (star), extending laterally, infiltrating the pterygoid musculature (solid arrow) and encasing the right internal carotid artery (dashed arrow). There is also extension to the left of midline within the soft palate (arrowhead). (B) Sagittal contrast-enhanced CT image demonstrates craniocaudal extent of tumor with involvement of the soft palate (star) and nasopharynx (arrow). Involvement of tumor within the pterygoid muscles and nasopharynx and encasement of the carotid artery is consistent with category T4 disease. Tumor surrounding the internal carotid artery, extension to the contralateral side, and tumor-related trismus are contraindications for transoral robotic surgery.

In carefully selected lateralized tonsillar cancers (disease confined to the tonsil and lateral soft palate without tongue base involvement and located more than 1 cm from midline) and with limited ipsilateral nodal disease, ipsilateral therapy provides excellent control with a low risk of contralateral neck failure (45,46). Clinical and radiologic factors considered during multidisciplinary tumor board discussions when determining treatment approach are summarized in Table 1.

Posttreatment Evaluation and Emerging Role of TTMV HPV DNA

Baseline posttreatment imaging should be performed approximately 12 weeks after completion of therapy, using ¹⁸F-FDG PET/CT (47). This interval balances the need for early response assessment against the high likelihood of treatment-related changes, such as edema and inflammation, which can generate false-positive findings and lead to unnecessary concern for patients and the multidisciplinary team (4,48).

Circulating TTMV HPV DNA is an emerging adjunct in the posttreatment setting. Serial measurements demonstrate excellent negative and positive predictive values (more than 98% and 95%, respectively) for recurrence, and rising or persistently detectable levels often precede radiologic or clinical detection of disease by several weeks to months (17,49). Clearance of TTMV HPV DNA after treatment is associated with favorable outcomes, whereas persistent or recurrent detectability is associated with worse survival. However, its current role is largely complementary. Emerging expert panel commentary recommends that circulating tumor DNA assays be interpreted alongside imaging, endoscopic evaluation, and clinical findings, rather than as a stand-alone test for adjudicating indeterminate posttreatment abnormalities (17,21).

For example, indeterminate hypermetabolic activity at ¹⁸F-FDG PET in the treatment bed without a corresponding abnormality at CT or physical examination, in the setting of negative TTMV HPV DNA, is unlikely to represent recurrence; follow-up imaging in 3 months is useful to confirm resolution of hypermetabolic activity and absence of disease (Fig 10). Nodal involution is often delayed in HPV-associated OPSCC, with only a minority of residual nodes containing viable tumor (47). Accordingly, patients with undetectable TTMV HPV DNA and residual nodal tissue at imaging are appropriate for close imaging surveillance; follow-up imaging typically demonstrates stable or decreased nodal tissue, thereby avoiding unnecessary biopsy or patient anxiety (Fig 11). In cases of rising TTMV HPV DNA without evidence of recurrence in the neck at CT or physical examination, evaluation for metastatic disease elsewhere with CT or ¹⁸F-FDG PET should be strongly considered (Fig 12). The optimal use of TTMV HPV DNA remains under investigation and it is not incorporated into AJCC v9/TNM 9 staging.

Figure 10:

Images in a 69-year-old male patient with human papillomavirus (HPV)–associated oropharyngeal carcinoma, after transoral robotic surgery and chemoradiation therapy. (A) Axial PET/CT image obtained 3 months after completion of therapy demonstrates hypermetabolic activity at the junction of the right posterior oral tongue and base of tongue (arrow). (B) Axial contrast-enhanced CT image acquired the same day as PET/CT shows no corresponding mass-like enhancement (arrow). Tumor tissue–modified viral (TTMV) HPV DNA was undetectable, and direct visual inspection was negative for disease; therefore, multidisciplinary tumor board recommendation was a follow-up CT examination. (C) Follow-up axial contrast-enhanced CT image obtained 3 months later demonstrates an unchanged appearance of the primary site (arrow) with no findings for recurrent disease. TTMV HPV DNA remained undetectable, and direct visual inspection was normal. This case highlights integration of TTMV HPV DNA testing in interpreting posttreatment imaging of HPV-associated oropharyngeal carcinoma; discordant PET and CT findings with negative circulating tumor DNA make residual/recurrent disease less likely; follow-up is useful to document stability and absence of residual/recurrent disease.

Figure 11:

Images in a 74-year-old female patient with human papillomavirus–associated oropharyngeal carcinoma, after definitive chemoradiation therapy. (A) Axial contrast-enhanced CT image before treatment demonstrates a right level II conglomerate metastatic lymph node (arrow) with infiltration into the sternocleidomastoid muscle (arrowhead), compatible with imaging-detected extranodal extension. (B) Axial contrast-enhanced CT image 3 months after completion of therapy demonstrates residual nodal tissue at the anterior margin of the right sternocleidomastoid muscle (arrow). Tumor tissue–modified viral (TTMV) HPV DNA test findings were negative, and therefore, residual nodal tissue was considered treated disease. (C) Axial contrast-enhanced CT image 1 year after completion of therapy demonstrates decreased volume of nodal tissue (arrow). TTMV HPV DNA remained negative, and therefore, no findings to suggest residual/recurrent disease.

Figure 12:

Images in a 64-year-old male patient with history of human papillomavirus–associated oropharyngeal carcinoma. (A) Axial contrast-enhanced pretreatment CT image demonstrates a large mass within the tongue base (arrows) and metastatic right level IIa lymph node (arrowhead). (B) Axial contrast-enhanced CT image 1 year following completion of treatment demonstrates a normal appearance of the oropharynx and no metastatic nodal disease. However, tumor tissue–modified viral (TTMV) HPV DNA level was trending upward, worrisome for presence of disease. Given negative CT scan of the neck, CT scans of the chest, abdomen, and pelvis were performed. (C) Axial contrast-enhanced CT image of the chest shows a spiculated lesion within the left lower lobe (arrow), which was pathology-proven metastatic disease. Rising TTMV HPV DNA level and a normal posttreatment neck should prompt search for distant metastatic disease.

Future Directions

Beyond AJCC v9/TNM 9, future efforts in HPV-associated OPSCC will likely focus on moving from purely anatomic staging toward more nuanced risk stratification. The eighth edition successfully downstaged most HPV-positive tumors to reflect their favorable prognosis but resulted in a heavily stage I–weighted distribution that obscured outcome heterogeneity. Although AJCC v9/TNM 9 incorporates iENE into clinical nodal staging for HPV-associated OPSCC, overall stage grouping remains largely weighted toward stage I and II disease, and stage IV continues to be reserved for patients with metastatic disease. Future AJCC versions may further standardize quantitative ENE metrics, for example, distinguishing minor from major ENE based on depth of soft tissue spread or incorporating iENE grades into the N category (11,28).

Concurrently, incorporation of nonanatomic biomarkers such as TTMV HPV DNA into risk stratification represents a logical next step. Detectable posttreatment TTMV HPV DNA is a strong predictor of recurrence, and patterns of clearance during chemoradiation correlate with disease control, raising the possibility of response-adapted de-escalation in future trials (50). Pretreatment TTMV HPV DNA levels may also reflect total tumor burden and nodal stage. At present, expert panels consider TTMV HPV DNA testing an adjunct to conventional imaging and clinical staging rather than a replacement; however, future staging systems or risk models may formally integrate TTMV HPV DNA as a modifier of anatomic stage, particularly in posttreatment and surveillance settings (47,49,51).

Therapeutically, the dominant research focus is de-escalation without loss of cure. TORS-based strategies enable pathologic risk stratification to guide adjuvant radiation dose and the need for chemotherapy, as demonstrated in ECOG 3311 and ongoing trials such as PATHOS. Parallel efforts in definitive chemoradiation are evaluating reduced radiation doses or altered fractionation for favorable-risk patients; however, in the absence of mature phase III data, these approaches remain investigational (52,53). Future de-escalation strategies will likely rely on more individualized patient selection using factors such as ENE status, circulating tumor DNA kinetics, hypoxia imaging, and early treatment response to identify patients who may safely undergo less intensive therapy while reserving full-dose chemoradiation or treatment intensification for higher-risk groups.

Systemic therapy for recurrent and metastatic HPV-associated OPSCC is also evolving. Genomic profiling of recurrent HPV-positive tumors has demonstrated acquisition of mutations more typical of HPV-negative disease, suggesting that resistance biology may diverge from that of the primary tumor and may require different targeted strategies (51,54). For radiologists, these developments underscore an expanding role. Beyond assigning T, N, and M categories, imaging will be integral to biomarker-guided trials, response-adapted de-escalation, early detection of oligometastatic disease for metastasis-directed therapy, and longitudinal correlation with liquid biopsy and molecular data as staging systems evolve toward a combined anatomic-biologic paradigm.

Conclusion

Knowledge of oropharyngeal anatomy and the AJCC v9/TNM 9 staging protocol for HPV-associated OPSCC, particularly the nuances and cautious interpretation of iENE, is essential, as imaging directly contributes to clinical stage assignment and influences multidisciplinary treatment planning. As blood-based biomarkers increasingly enter clinical practice, high-value care will depend on their integration with standardized imaging interpretation. Future refinements to staging criteria should be guided by stronger evidence demonstrating whether iENE can be identified reproducibly and whether it adds incremental prognostic value beyond established clinical, imaging, and pathologic factors.

Disclosures of conflicts of interest

P.R. is an associate editor for the Radiology Early Career Editorial Board. Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplemental materials.