gms | German Medical Science

Objective: Tumor measurement is important in children with unresectable low-grade gliomas (pLGG) to determine the need for treatment or assessment of response. Standard measurements are usually ascertained by the biggest dimension in the transverse, anterior-posterior and cranio-caudal and the product of largest two dimension (SM, cm²). This study evaluates 3-D tumor volume measurements (VM, cm³) in pLGG.

Method: Of 32 pLLG patients treated between 2005 - 2015, 29 patients had at least three evaluable studies of post contrast Spoiled Gradient Recalled (SPGR) acquisition, a 3D volumetric sequence to allow calculation of tumor volume. SM and VM were performed independently and blinded to patient’s data by two investigators. The visualization and rendering software program (3D-Analyze, ToMMTi-Systems, Erfurt, Germany) was used for VM; the exterior boundaries of each tumor across slices were traced and a volume calculated. Inter-rater reliabilities for both SM and VM were determined. In addition percent change from the first MRI to the MRI, which revealed the maximum difference in tumor size, was compared between both approaches.

Results: Mean tumor size at the first MRI scan was 20 cm² based on SM approach and 398 cm³ using the VM approach. Inter-rater reliability was high for both approaches (.98 versus .96). The change scores, based on absolute values, revealed an overall 49% change in size from the first scan when using SM compared to 65% with the VM approach. The difference in percent change between methods approached significance for this small sample (t= 2.25, p< 0.06).

Conclusions: Inter-rater reliability was found to be very high for both SM and VM. Because of the relatively higher sensitivity to change in size, volumetric assessments in pLGG may provide greater consistency between imaging studies to assess changes in these mostly irregularly shaped tumors and allow a more accurate evaluation of treatment response.