Discussion:
Our study suggests that OOSCC heterogeneity as evaluated by CT-texture analysis is an independent predictor of overall survival in patient treated by surgical resection. Indeed, pretreatment mean gray-level values were significantly associated with OS at medium texture scale (SFF=5) and significantly associated with EFS at coarse texture scale (SFF=6).
Tumors are heterogeneous both on genetic and histopathological levels with intratumoral spatial variation in the cellularity, angiogenesis, haemorrhage, and areas of necrosis. Tumors with high intratumoral heterogeneity have been shown to have a poor prognosis, which could be secondary to intrinsic aggressive biology or treatment resistance(5).
Over the past decade, texture analysis has increasingly been investigated as a method to predict survival in patients with a wide spectrum of cancers(4,8–10). In head and neck cancers, Zhang et al. (6) found entropy and skewness on CT-TA to be associated with the OS of patients with locally advanced OOSCC treated by induction chemotherapy with cisplatin, 5-fluorouracil, and docetaxel.
In the filtration-histogram approach for CT-TA, “mean” reflects the measure of uniformity of the histogram corresponding to the gray-level values within a ROI. We hypothesize that gray level uniformity of primary oral and oropharyngeal tumors may be linked with hypervascularization and thus may reflect tumor angiogenesis.
Limitations of our study include the use of a monocentric retrospective design with possible known bias. However, to reduce this bias, the evaluation of imaging and histopathology was conducted blinded and independently to each other. Moreover, patient sample is small. Many patients had to be eliminated from the study before performing CT-TA because of marked dental artifacts. The effects of metallic artifacts on CT-TA are not fully known or described; therefore, for this pilot study, we tried to collect subjects without any dental artifacts to limit this factor as a potential confounder.