*3.1.1 Risk factors which have been utilised traditionally*

Oral cavity cancers are commonly managed with single modality treatment in early stages and a combined modality approach in advanced stages. Radical, oncologically-sound surgical approaches form the backbone of early cancers with a meticulous assessment of the histopathological specimen. Even with appropriate excisions, local and regional recurrences are an extremely concerning aspect which can alter the eventual outcomes in these cancers.

Many retrospective studies and single institution prospective trials demonstrated the benefit of adding radiation adjuvantly for oral cancer patients who have undergone radical surgery. Cooper et al. collated data from RTOG #85-03 & #88-24 to retrospectively sort into 3 risk groups of presumed progressive risk to enable adjuvant treatment decisions [4]. Group I included fewer than 2 involved nodes, no ECE & negative surgical margins. Group II included at least 2 involved nodes or presence of ECE of tumour with uninvolved margins. Group III included microscopically involved surgical margins. Accordingly, comparing outcomes in groups I to III, the loco-regional recurrence rates at 5 years was 17%, 27% & 61% and median survival was 5.6 yrs., 2.6 yrs. and 1.5 yrs. respectively.

Langendijk et al. used the Classification and Regression Tree(CART) or the Recursive Partitioning Analysis(RPA) method to construct homogenous subgroups of well-known prognostication parameters to be able to identify locoregional recurrence risk [5]. Accordingly, 801 patients were divided into RPA 1 group (Intermediate risk) which included no ECE and free surgical margins; RPA 2 (High Risk) which included T1,T2 and T4 tumours with close or positive surgical margins or one positive node with ECE and RPA 3 (Very High Risk) which included T3 tumours with close or positive margin, N3 neck or multiple nodes with ECE. The 5-year LRC was 92%, 78% and 58% and OS rates were 67%, 50% and 36% for the three RPA classes respectively.

Salama JK et al. used pooled multivariate analysis to give a consensus statement on use of adjuvant therapy in head and neck cancers [6]. They divided patients into two groups to identify their risk for developing loco-regional recurrences based on pathological features. High risk group included those with involved surgical margins and extranodal spread. Low risk group included other adverse features such as T3/4 tumours, perineural invasion, lymphovascular space invasion, multiple nodal involvement and lower neck adenopathy. They advised for postop radiotherapy alone for patients with low-risk features and addition of concurrent chemotherapy for those with high-risk features. They also suggested postop radiotherapy dose of upto 63Gy for high-risk features and 57Gy for those with low-risk disease.

Perineural Invasion (PNI): There is no standardised definition of this entity, but the most accepted one is when tumour cells are present in any one of the three layers of the nerve sheath & when tumour cells are in close proximity to the nerve & involves more than one-third of circumference. Bur et al. identified an incidence of PNI in literature of between 3 and 52% [7]. Though many authors have not been able to assign prognostic significance to presence of PNI, some features such as multiple foci of PNI, involvement of large nerves (>1 mm) and higher maximum extent of PNI were associated with increased local failure and reduced disease specific survival.

In the systematic review collated data from 13 retrospective studies, they identified local recurrence rates of 4.4% - 22.9%, regional recurrence rates of 12.3% - 30.8% and 5-yr overall survival estimates of 48–89.6% in those patients with presence of PNI [7]. In two studies from this review, where neck dissection was conducted, regional failure rates were 12.3% and 17.6% which undermines the fact of ineffective salvage options after this. However, the authors suggest that in the absence of any prospective trial to assess the impact of PNI on loco-regional recurrence or survival, it would be imperative to discuss the options of treatment in detail in those patients where positive PNI is the only risk factor post radical surgery.

Depth of Invasion (DOI): Tumour thickness or depth of invasion has been consistently identified as a predictor for cervical lymph node metastasis in oral cavity cancers. Oral cavity cancers have occult nodal metastases of up to 40% in clinically negative neck, which is usually managed with elective nodal dissection as opposed to just observation. Huang et al. in their meta-analysis, have tried to address this aspect and have concluded that the optimal cut-off point for DOI is 4 mm, to consider for neck management [8].

In another study, Liao et al. followed-up patients who underwent surgery of early-stage oral cavity cancers(pT1-2 N0) to identify poor prognostic features and suggest for adjuvant therapy [9]. They found that poorly differentiated tumours and DOI of 4 mm and above were both independent poor prognostic factors, and when present together accounts for 2-yr regional failure of 42%. Hence, they suggest for the use of PORT in this subset of early OC cancers.

Bulbul et al. conducted a meta-analysis of 8 studies (1427 patients), which used frozen section (FS) evaluation to define margin status in early(T1/2N0) oral cavity cancers [10]. They compared positive/close FS margins which was cleared by further resection (R1 – R0), positive margin not cleared (R1) to those with negative margins upfront (R0). They found that patient with R1-R0 had poorer local recurrence free survival (LRFS) when compared to R0, regardless of clearance which was statistically significant. Furthermore, R1-R0 patients showed almost equal LRFS to

that of R1, though the trend to worse result was with R1 patients. They concluded that upfront positive/close margin was a marker of a locally aggressive disease regardless of re-resection & correction. They also indicated that there should be standardisation of the FS sampling method too.
