The evolution of nasopharyngeal carcinoma staging

Abstract

The prevalence of nasopharyngeal carcinoma is characterized by an unbalanced distribution: the disease is particularly prevalent in East and Southeast Asia. In this article, we review the evolution of the International Union Against Cancer/American Joint Committee on Cancer staging system for nasopharyngeal carcinoma. With the increasing using of newer imaging methods, more advanced radiotherapy techniques and systemic chemotherapy, we also discuss newer clinical features that might affect staging. Finally, we propose the future direction of staging and potential prognostic factors that have a major influence on the treatment outcomes of this disease.

INTRODUCTION

Nasopharyngeal carcinoma (NPC) is a squamous cell carcinoma arising from the epithelial lining of the nasopharynx. According to the International Agency for Research on Cancer, 129,079 new cases of NPC were reported in 2018.¹ Compared to other cancer types, NPC is uncommon, with a very unique geographical distribution, whereby more than 70% of new cases are reported in East and Southeast Asia.¹

The current International Union Against Cancer (UICC) and the American Joint Committee on Cancer (AJCC) TNM Classification for NPC remains the most accurate prognostic tool to stratify patients for treatment and evaluation of treatment outcomes.^2–4 The system is based purely on the following three anatomical criteria: the local extent of the cancer within the site of origin (T), the degree of metastatic involvement of the regional lymph nodes (N), and the presence or absence of distant metastatic disease (M). The introduction of the TNM Staging System fifth edition jointly adopted by AJCC and the UICC in 1997 was a milestone.⁵ This consensus on a unified NPC staging and classification system was reached based on the results of retrospective studies worldwide.^6–9 The seventh edition of this staging system was released in 2009, and it was recently updated (eighth edition) in 2016,¹⁰ further refining the staging of primary and nodal disease and the stage groups.^11,12

In the past two decades, there have been significant advancements in imaging technology, radiotherapy (RT) delivery, and systemic chemotherapy for NPC patients.^13–17 With these changes in diagnosis and treatment, the TNM classification systems have also been significantly refined. This article presents in detail the evolution of the TNM classification from the fifth to the eighth edition.^5,10,18,19

THE EIGHth EDITION OF THE TNM CLASSIFICATION: – A SUMMARY OF THE CHANGES (Table 1)

Table 1. .

Comparison between the seventh and eighth editions of the UICC/AJCC staging system for nasopharyngeal carcinoma

Seventh edition		Eighth edition
Primary tumor (T)
TX	Primary tumor cannot be assessed	TX	Primary tumor cannot be assessed
		T0	No tumor identified, but EBV-positive cervical node(s) involvement
T1	Nasopharynx, oropharynx or nasal cavity without parapharyngeal extension	T1	Nasopharynx, oropharynx or nasal cavity without parapharyngeal extension
T2	Parapharyngeal extension	T2	Parapharyngeal extension, adjacent soft tissue involvement (medial pterygoid, lateral pterygoid, prevertebral muscles)
T3	Bony structures of skull base and/or paranasal sinuses	T3	Bony structures (skull base, cervical vertebra) and/or paranasal sinuses
T4	Intracranial, cranial nerves, hypopharynx, orbit, infratemporal fossa/masticator space	T4	Intracranial extension, cranial nerve, hypopharynx, orbit, extensive soft tissue involvement (beyond the lateral surface of the lateral pterygoid muscle), parotid gland
Regional lymph nodes (N)
NX	Regional lymph nodes cannot be assessed	NX	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis	N0	No regional lymph node metastasis
N1	Unilateral cervical, unilateral or bilateral retropharyngeal lymph nodes, above the supraclavicular fossa;≤6 cm	N1	Unilateral cervical, unilateral or bilateral retropharyngeal lymph nodes, above the caudal border of cricoid cartilage;≤6 cm
N2	Bilateral metastasis in lymph node(s), 6 cm or less in greatest dimension, above the supraclavicular fossa	N2	Bilateral metastasis in lymph node(s), 6 cm or less in greatest dimension, above the caudal border of cricoid cartilage
N3a	Greater than 6 cm in dimension	N3	>6 cm and/or below caudal border of cricoid cartilage (regardless of laterality)
N3b	Supraclavicular fossa
Distant metastasis (M)
M0	No distant metastasis	M0	No distant metastasis
M1	Distant metastasis	M1	Distant metastasis
Stage group
I	T1 N0 M0	I	T1 N0 M0
II	T2 N0-1 M0, T1 N1 M0,	II	T2 N0-1 M0, T0-1 N1 M0
III	T1-3 N2 M0, T3 N0–1 M0	III	T3N0-2 M0, T0-2 N2 M0
IVA	T4 N0–2 M0	IVA	T4 or N3 M0
IVB	Any T N3 M0	IVB	Any T, any N M1
IVC	Any T, any N M1
Summary of changes
T category
T0 is added for EBV-positive cervical node(s) involvement despite unidentified primary tumor
Involvement of medial pterygoid, lateral pterygoid, and prevertebral muscles is now staged as T2
In T4, the original “ infratemporal fossa/masticator space” is replaced by specific description of soft tissue involvement to avoid potential ambiguity
N category
The previous N3b criterion of “supraclavicular fossa” is now changed to “ below caudal border of cricoid cartilage”
N3a and N3b are merged into a single N3 Stage group
The previous Stages IVA and IVB are merged into IVA •The previous IVC is now upstaged to IVB

EBV, Epstein-Barr virus; UICC/AJCC, International Union Against Cancer/American Joint Committee on Cancer.

One of the revisions in the current edition is the adjustment in the T0 stages defined in the previous edition.¹⁰ The T0 stage was added for Epstein-Barr virus (EBV)-positive unknown primary tumors with cervical lymph node (CLN) involvement. The second revision is the specific description of T2, from parapharyngeal extension to tumor involvement of the medial pterygoid, lateral pterygoid, and prevertebral muscles.²⁰ The third revision is that the previous T4 criteria “masticator space” and “infratemporal fossa” are now replaced by a specific description of soft-tissue involvement to avoid ambiguity. The fourth revision is that the previous N3b criterion for the supraclavicular fossa (SCF) has been changed to the lower neck (as defined by nodal extension below the caudal border of the cricoid cartilage)^21,22 : N3a and N3b are merged into a single N3 category, which is now defined as unilateral or bilateral metastasis in any direction, with 6 cm being the largest dimension in CLN(s) in the cricoid cartilage.

IMAGING STUDIES

Significant stage migration has been observed with the increasing use of advanced imaging technology. With the advantages of excellent tissue contrast, multiplanar capacity, and a lack of radiation and bone beam-hardening artefacts, MRI allows for a more accurate evaluation of local disease extension, especially parapharyngeal tumor extension, oropharyngeal extension, and ethmoid sinus.^23–25 These resulted in changes in 49.8% of T stage cases, 10.7% of N stage cases, and 38.6% of clinical stage cases according to the sixth AJCC/UICC staging system,²⁶ and enabled a more accurate delineation of tumor margins in the nasopharynx.^26–29 On the other hand, ¹⁸F-FDG-positron emission tomography (PET)/CT affords more sensitivity than does the conventional work-up (i.e. chest radiography, abdominal ultrasound, and skeletal scintigraphy) in detecting distant metastasis,³⁰ small CLN metastases, and local residual/recurrent disease.^27,31–33 Peng et al showed PET/CT examination resulted in modification of N categories and overall stage for 135 (28.7%) and 46 (9.8%) patients, respectively.³² In a study by Law et al, PET-CT was valuable in terms of impacting management in 33% of patients (major impact in detecting M1 disease in 8% and medium impact by upstaging the N category or showing the exact lymph node in 25%).³⁴

UNSETTLED ANATOMIC ISSUES

Classification of T0

In the eighth edition of the AJCC staging manual for NPC, the definition of T0 is added for EBV-positive unknown primary tumors with CLN involvement.¹⁰ This assignment is based on evidence level III, which means the supporting evidence is somewhat problematic. In fact, EBV is not only related to NPC but also lymphoepithelioma-like carcinoma, which arises from mucosal sites outside the nasopharynx, including the salivary gland, oropharynx, sinonasal tract, and non-head and neck regions.^35–38 Lymphoepithelioma-like carcinoma is morphologically similar to undifferentiated NPC with a high frequency of concurrent CLN metastases but could occur in multiple sites. Therefore, it is an arbitrary assumption that all occult primary tumors with EBV-related CLN metastases originate from the nasopharynx. In patients with EBV-positive CLNs, Mao et al showed the most common primary sites after the nasopharynx (52.3%) were the salivary gland (24.8%), lung (7.9%), oropharynx (3.4%), nasal cavity/maxillary (3.4%), oral cavity (1.1%), orbit (1.1%), and liver (0.4%). The origins of EBV-positive CLNs may not be restricted to the nasopharynx alone, and are likely to involve the head and neck or non-head and neck regions. The NPC T0 classification should be cautiously assigned to such tumors.³⁹

PROGNOSTIC SIGNIFICANCE OF PREVERTEBRAL SPACE INVOLVEMENT

In the current eighth edition of the staging system, prevertebral muscle involvement was added as a T2 criterion. Several retrospective studies have shown that prevertebral space involvement (PSI) is an independent prognostic factor.^40–42 For patients with treatment using two-dimensional RT, PSI was associated with a poor prognosis on local and distant control.⁴¹ In Lee”s study, distant metastasis-free survival (DMFS) increased from 72 to 100% after adjuvant chemotherapy in the patients with PSI. However, this effect was not observed in the group without PSI.⁴² Zhou et al retrospectively reviewed 506 patients in the intensity modulated radiotherapy (IMRT) era and found that PSI was an independent prognostic factor for both overall survival (OS) and DMFS, and suggested a classification of T4.⁴⁰ In a study by Pan et al, in which prevertebral muscle involvement was added as a T2 criterion, the OS of patients with PSI was similar to that of those with parapharyngealspace involvement.²⁰

CERVICAL VERTEBRAE AND PAROTID GLAND INVOLVEMENT

The eighth TNM staging system added the cervical vertebrae and parotid gland, which were not classified in the previous versions of the staging system. The incidence rate of concurrent tumor invasion into the cervical vertebrae and parotid gland observed by MRI was <5%.⁴³ In the study by Pan et al, the subgroup of patients with parotid gland infiltration showed poor OS, similar to that of the subgroup with intracranial extension and/or cranial nerve palsy. Moreover, there was no statistically significant difference in OS between those showing involvement of pterygoid structures alone and those presenting with erosion of the skull base and/or cervical vertebra (86% vs 79%).²⁰ However, there is no separate list covering the prognosis in cases with involvement of the cervical vertebra and parotid gland. At present, there is no research on the prognosis of cervical vertebra and parotid gland invasion, and further investigation is needed on the staging of cervical vertebra and parotid gland invasion.

STAGING: NODAL METASTASIS

Supraclavicular fossa

The SCF, originally described by Ho,^44,45 is the triangular region defined by the superior margin of the sternal end of the clavicle, the superior margin of the lateral end of the clavicle, and the point where the neck meets the shoulder. This definition of SCF involvement is based primarily on clinical examination, and there was no reliable method to define the SCF radiologically by using the above clinical landmarks before CT cross-sectional imaging was widely employed for N classification in NPC.^46,47 Ng et al suggested replacing the SCF with the radiological nodal levels IV and Vb, and found that this substitution was predictive for both distant control and overall survival.⁴⁸ The study by Mao et al on 924 patients staged by MRI also supported the use of levels as defining criteria.⁴⁹ For NPCs treated with IMRT, Li et al²¹ also reported that the involvement of levels IV and Vb and the SCF as defined by the International Consensus Guidelines^47,50 was an independent prognostic factor for distant failure and disease failure. Compared to the seventh UICC/AJCC, Yue et al showed that N-categories based on the lower cervical levels (extension below the caudal edge of the cricoid cartilage) provided more satisfactory distinction between hazard ratios for distant and disease failure for each N category. N3a and N3b as defined by the seventh UICC/AJCC showed similar DMFS (p = 0.31) and Disease-free survival (DFS) (p = 0.21) rates.²² Hence, one modification brought along with the eighth edition will be changing the definition of lower neck extension to lymphatic spread below the caudal edge of the cricoid cartilage as N3, and merging N3a and N3b into a single N3 category.

MEASUREMENTS OF THE MAXIMAL NODAL DIAMETER ON MRI SLICES

A study from Hong Kong that included almost 5000 NPC cases without distant metastasis found that, when adjusted for other meaningful parameters, palpation-determined LN size (greatest diameter ≤6 cm vs. >6 cm) was a significant prognostic factor.⁵¹ On the basis of that study, cervical LN size (greatest diameter >6 cm) has been included in the UICC/AJCC NPC staging system N3 subset since 1992 (i.e. the fourth edition). Subsequent staging systems have also adopted this criterion, including the eighth edition of the UICC/AJCC staging system. However, the criterion is based on the evaluation of palpable LNs, which would differ between clinicians. Furthermore, the widespread use of CT, MRI, PET-CT, and other advanced techniques in NPC can allow detection of LN metastases at a higher rate of 77–88.1%,^52,53 compared to only about 60% by palpation.⁵⁴ Both the LN levels and laterality are assessed based on MRI, but in the eighth staging system, the nodal size is defined according to the largest dimension without specifying the plane to be used. Mao et al found that the MRI-determined longest CLN diameter is a significant prognostic factor for both disease failure and distant failure in NPC in univariate analyses.⁵⁵ However, only CLN level and laterality and extranodal neoplastic spread (ENS) had such significant prognostic value for distant failure and disease failure in multivariate analyses. Li et al showed that maximal axial diameter (MAD) values of >4 cm were associated with a higher risk of nodal failure (p = 0.05). Multivariate analysis showed level was the only independent prognostic factor for distant failure (p < 0.01) and disease failure (p < 0.01).²¹ There is a need to investigate the significant cut-off values of MRI-determined LN size, the measurement (axial plane MAD or the longest diameter of all planes) for the N classification in the future.

Parotid area lymph node metastases

The pattern of nodal metastasis in NPC follows an orderly spread down the neck with infrequent node skipping; the most commonly involved regions include the retropharyngeal space, levels II, III, IV, and V, followed by level Ib and the supraclavicular nodes.^52,56–60 The parotid lymph nodes (PLNs), classified as level VIII in the latest International Consensus Guidelines for nodal levels,⁶¹ are also at risk of harboring metastasis from the nasopharynx. However, the classification of PLN metastasis was not identified until the eighth edition of the AJCC/UICC staging system. Zhang et al reported that the incidence of PLN metastasis was 1.2% (21/1811), and PLN involvement was associated with significantly poorer progression-free survival (PFS), OS, and DMFS, with a significantly higher hazard ratio for distant failure than N2 disease and similar to N3 disease.⁶² In the study by Xu et al, the results also showed that the 5 year DMFS and regional recurrence-free survival (RRFS) curves for PLN metastasis were significantly separated from that for N2 disease but crossed that for N3 disease.⁶³ Although the rate of parotid lymph node metastasis is actually low, the inconsistent classification may affect treatment strategies for patients with PLN metastasis. Therefore, the staging of parotid lymph node metastasis needs to be further clarified.

Future direction

The following improvements should be considered: regrouping of T and N categories for better distinction of hazard between stages and incorporation of other prognostic factors beyond anatomic staging.

Simplification of T staging

The eighth edition of the AJCC NPC staging system has been proven to provide more accurate prediction of treatment outcomes than the seventh edition.⁶⁴ The local control rate for NPC is currently 90–95% for patients treated using modern techniques.^65–67 However, these advances have altered the prognostic value of staging parameters for local failure,⁶⁸ and the prognostic value of the T category may have become weaker. The survival curves for T2 and T3 almost overlap, with no significant differences in locoregional relapse-free survival and DFS.⁶⁴ Considering the improved local control provided by modern techniques, Tang et al re-evaluated the prognostic value of the eighth edition T categories with an independent, external validation cohort. The results showed that in the IMRT era, merging T2 and T3 to a proposed T2 provided significant differences in local recurrence-free survival, DFS, and OS between the proposed T categories.This article also suggested a new regrouping of T and N classification, and defining proT1N0 as Stage I; proT1N1/proT2N0-1 as Stage II; proT3N0-2/proT1-2N2 as Stage III; and proT1-3N3 as stage IVa generated orderly, significant differences in DFS and OS between stages in the training set and external validation cohort⁶⁹ (Figure 1).

Figure 1.

The proposed T classification for the study of the tang et al. The proposed T2 was combination of T2 and T3 for the 7th AJCC/UICC staging system

The role of other prognostic factors beyond anatomic staging

Many studies have been conducted to evaluate whether the incorporation of other clinical factors and molecular biomarkers into the staging system would better predict survival. Larger tumors are related to an increased number of clonogenic tumor cells, as well as tumor hypoxia, which is associated with increased radioresistance and possibly altered levels of intercellular communication factors.^70,71 Several studies have shown a positive correlation between the primary tumor volume and various survival outcomes in NPC.^72–82 However, NPC is a highly infiltrative tumor with a propensity to spread along the adjacent soft tissue as well as to the skull base and foramina. Therefore, the traditional measurement of tumor size in NPC is time-consuming, labor-intensive, or limited in availability. In recent years, IMRT has gradually replaced two-dimensional-CRT as the primary RT technique. With the increased use of computerized treatment planning systems for the delineation of three-dimensional tumor volume in radiation therapy, the calculation of NPC tumor volume has become easier. Among these studies, the series by Guo et al is the largest (n = 694), and their results suggest an improved predictive ability of T-classifications when the primary tumor volume is incorporated.⁷⁶ Furthermore, cell-free EBV DNA was correlated with the tumor burden, TNM stage, and survival of NPC patients.^83–87 Analysis of plasma/serum EBV DNA using real-time polymerase chain reaction is now widely used under clinical conditions for primary diagnosis, prognostication, and monitoring relapse and response to treatment, as well as population screening.⁸⁸ The addition of pretreatment plasma EBV DNA into the eighth edition staging system has been shown to significantly improve its prognostic performance.^89,90 Most importantly, the cutoff levels vary widely between different studies and international efforts are still ongoing to standardize the quantitative assay. Other biomarkers such as LDH, DNA methylation, miRNAs, and mRNAs also have demonstrated their prognostic value and potential clinical applications in nasopharyngeal carcinoma. A six-hypermethylated gene panel⁹¹ and a five-miRNA signature⁹² were found to be associated with survival in nasopharyngeal carcinoma patients, respectively.

Important changes have been introduced in the 8th AJCC/UICC staging system. However, new parameters which are prognostic and practical for future staging updates would be needed. More individualized treatment strategy would continue to be altered with the modified staging system.

Many studies have been conducted to evaluate whether the incorporation of other clinical factors and molecular biomarkers into the staging system would better predict survival. Larger tumors are related to an increased number of clonogenic tumor cells, as well as tumor hypoxia, which is associated with increased radioresistance and possibly altered levels of intercellular communication factors.^70,71 Several studies have shown a positive correlation between the primary tumor volume and various survival outcomes in NPC.^72–82 However, NPC is a highly infiltrative tumor with a propensity to spread along the adjacent soft tissue as well as to the skull base and foramina. Therefore, the traditional measurement of tumor size in NPC is time-consuming, labor-intensive, or limited in availability. In recent years, IMRT has gradually replaced two-dimensional-CRT as the primary RT technique. With the increased use of computerized treatment planning systems for the delineation of three-dimensional tumor volume in radiation therapy, the calculation of NPC tumor volume has become easier. Among these studies, the series by Guo et al is the largest (n = 694), and their results suggest an improved predictive ability of T-classifications when the primary tumor volume is incorporated.⁷⁶ Furthermore, cell-free EBV DNA was correlated with the tumor burden, TNM stage, and survival of NPC patients.^83–87 Analysis of plasma/serum EBV DNA using real-time polymerase chain reaction is now widely used under clinical conditions for primary diagnosis, prognostication, and monitoring relapse and response to treatment, as well as population screening.⁸⁸ The addition of pretreatment plasma EBV DNA into the eighth edition staging system has been shown to significantly improve its prognostic performance.^89,90 Most importantly, the cutoff levels vary widely between different studies and international efforts are still ongoing to standardize the quantitative assay. Other biomarkers such as LDH, DNA methylation, miRNAs, and mRNAs also have demonstrated their prognostic value and potential clinical applications in nasopharyngeal carcinoma. A six-hypermethylated gene panel⁹¹ and a five-miRNA signature⁹² were found to be associated with survival in nasopharyngeal carcinoma patients, respectively.