Oral Cancer: New System May Improve Prognostic Accuracy
FROM CANCER
The TNM staging system is used by most facilities for cancer reporting, as defined by the National Cancer Institute. This system combines the size and extent of the primary tumor (T), the number of neighboring lymph nodes with cancer and subcategories (N), and whether or not metastasis has occurred (M).
In a new study published in the journal Cancer, the researchers created a novel classification system to better account for extranodal extension (ENE). The study population included 1460 adults with OSCC (696 with no lymph node involvement and 764 with positive lymph nodes), who underwent surgical resections at four centers.
“Our findings build on the growing evidence base that historical factors do not improve staging performance and that their omission results in improved N‐classification [i.e., the nodal status or lymph node involvement in cancer staging] performance,” John R. de Almeida, MD, of the University of Toronto, and colleagues, wrote in their new paper.
For patients with OSCC, this system, known as the 8th edition of American Joint Committee on Cancer/International Union Against Cancer TNM N‐classification (TNM‐8‐N), has several limitations, the researchers explained. These limitations include redundancy in the rare N3a category (i.e., having single or multiple lymph nodes greater than 6 cm or 3-7 lymph nodes without ENE) and the impact of ENE as a new prognostic feature, they said.
“Recent studies have shown that major ENE is associated with a significantly worse outcome than minor ENE, suggesting that these two subgroups should be considered as separate entities,” the authors wrote.
Study Methods and Results
The researchers created N-classifications based on adjusted hazard ratios and statistical analysis (recursive partitioning) with a focus on lymph node (LN) size and number and the extent of ENE. They compared their classifications of OSCC cases to those of the TNM-8-N’s classifications of the same cases.
Using the new classification system, lymph node number and size and the extent of ENE were associated with overall survival. The adjusted hazard ratios for LN counts of 1 vs. zero and greater than 1 vs. 0 were 1.92 and 3.21, respectively. The adjusted hazard ratios (aHRs) for LN size of greater than 3 cm vs. 3 cm or less, and for major vs. minor ENE were 1.88 and 1.40, respectively.
The use of an aHR improved cancer staging compared to the TNM-8-N by eliminating the N2c and 6-cm threshold, stratifying the extent of ENE, and stratifying N2b by 3-cm threshold, the researchers wrote.
The researchers compared their new system to the TNM-8 and also two other classification systems and their own recursive partitioning analysis (another statistical model).
The aHR-based system ranked first out of five in terms of correctly staging cancer, while the TNM-8 was fifth in the discovery cohort and fifth in the validation cohorts.
Outcome predictions (percentage variance explained) were 19.81 with the aHR vs. 18.95 in theTNM-8 in the discovery cohort, and similarly were 11.72 vs. 10.13, respectively, in the validation cohort.
“Overall, 25 patients staged as IVa in TNM‐8 were upstaged to IVb in the aHR proposal, and one patient staged as IVb was downstaged to IVa. Otherwise, overall stage between TNM‐8 and aHR remained the same,” the authors wrote.
“Our proposed N-classification based on aHR challenges previous tenets such as the importance of the 6-cm threshold and the importance of contralateral nodes,” the researchers wrote in their discussion.
The results from the new classification system were limited by the relatively small sample sizes and may not generalize to nonsquamous oral cancers, the researchers noted.
Further validation is needed before this system may be routinely applied in practice, but the results support evidence in favor of eliminating historical factors from staging, they said.
