Size distribution and clinicoradiological signs of aggressiveness in odontogenic myxoma-three-dimensional analysis and systematic review.

OBJECTIVES: Therapeutic assessment of odontogenic myxoma (OM) is poorly standardized. Unidimensional size criteria have shown to be unreliable in therapeutic decision-making. We evaluate the size distribution of OM and scan for associated clinicoradiological signs of aggressiveness. Additionally, we evaluate three-dimensional size delineation of OM aiming to improve future therapeutic assessment of this destructive neoplasm.

METHODS: Primarily, we reviewed the database "PubMed" for data concerning the size of OMs as radiologically determined. Afterwards, the impact of age, sex, locularity and location on the size was investigated by χ² test, Student's t-test and regression analysis. Furthermore, we statistically evaluated the impact of size on the occurrence of clinicoradiological signs of aggressiveness. Secondly, we approximated the volume of five unpublished cases of OM by semi-automatic image segmentation of cone-beam CT images.

RESULTS: Multilocular OMs were significantly larger than unilocular ones (p < 0.002). Age (0.042) and multilocularity (<0.002) significantly impacted size. Size was significantly associated with cortical perforation (0.032) and multilocularity (<0.002), further regression analysis revealed tooth resorption (0.019), cortical perforation (0.005) and multilocularity (<0.002) as significant predictors of size. Employing the volume as a mean of comparison, we found that the biggest OM (38.42 ml; multilocular) was 124 times larger than the smallest (0.31 ml; unilocular). However, using the maximum diameter (cm) as a surrogate for size, the biggest lesion (6.3) was only 5.25 times larger than the smallest (1.2).

CONCLUSIONS: Locularity and volumetric size characterization might help in therapeutic decision-making and could help to improve our understanding of OM.