Head and neck cancer staging: Balancing innovation with consistency

Abstract

In this issue of Cancer, Ho et al recommend a convergence of human papillomavirus (HPV)-negative (HPV−) and HPV-positive (HPV+) oropharyngeal cancer (OPC) staging. The specific challenges they issue are to merge the nodal pathologic characterization of OPC, incorporate extranodal extension (ENE) in HPV+ OPC, and stage HPV− OPC using nodal number. In addition, they advocate a change of the breakpoint for the number of nodes to 6, as opposed to the current 5. Their article presents an opportunity to discuss some of the nuances and future potential for staging head and neck cancer. However, before any change can be undertaken, several fundamentals of staging should be reviewed. Head and neck cancer staging undertook substantial change with the publication of the 8 Editions of the Union for International Cancer Control (UICC) and American Joint Committee on Cancer (AJCC) manuals. The major changes include the addition of ENE in HPV− nodal characterization, the separation of OPC into an HPV-mediated and -nonmediated classification, and the inclusion of depth of invasion (DOI) in oral cavity cancer. It is important to look at these changes in totality and balance them against the long-standing precepts of staging. The first is continuity across time. No change should be enacted lightly or without strong evidence. In the cases of ENE and DOI, decades of evidence consistently emerged to support them as strong independent predictors of prognosis. ENE and DOI fulfilled the criteria for robust strength of data, thus providing strong support for a change. A second major concern was a balance of consistency across head and neck sites while keeping each change relevant within that particular site. This is important to keep the staging system manageable for clinicians. Too many differences between sites makes it difficult for clinicians to remember and implement a system. However, for those sites that had substantive opportunity for improvement, this principle should give way. The oral cavity and skin are the only sites with clear evidence of DOI as a predictor. ENE, on the other hand, has been demonstrated in multiple mucosal sites. Therefore, the former is introduced only in the oral cavity and skin, while the latter is universal to all mucosal sites. The case of HPV-mediated OPC is a special one. Given the dramatic emergence of this disease and the unique characteristics of the patients and the outcome contrasted with HPV-nonmediated OPC, a special staging system is warranted. Staging is an inherently conservative endeavor. Too-frequent change leads to inconsistency across time, whereas a too-conservative approach stifles innovation. The UICC and AJCC felt it essential that the 8th Editions of their staging manuals reflect the emergence of HPV+ OPC. To acknowledge and account for this new disease, available data were used to create a staging system that best reflected clinical outcomes. Thus, a clinical staging system modeled after the pioneering work of O’Sullivan et al was created using an internationally validated assessment of clinical staging. All patients should be staged clinically, but only those that undergo surgery can be staged pathologically. For sites that are treated primarily with an initial operation (eg, the oral cavity), pathologic staging has a critical role in determining adjuvant therapy; however, for sites that are treated primarily with radiation therapy (eg, nasopharyngeal cancer), treatment decisions are based almost exclusively on clinical stage. A somewhat unique aspect of OPC is the fact that its treatment is currently evolving, with some centers favoring a primary surgical approach, while others favor radiation therapy. Only OPC that was surgically treated could be used to develop a pathologic staging system. Some of the key thought leaders in this arena, Haughey and colleagues, performed an assessment of their institutional data as well as data from multiple other institutions. These datasets were used to derive pathologic staging. Heretofore, clinical and pathologic staging of the head and neck were consonant. However, in HPV+ OPC, available data support 2 distinct mechanisms of staging: 1) clinical node characterization based on laterality, size, and number and 2) pathologic node characterization based solely upon the number of positive nodes. A difficult