Relationship between tumor cell infiltration and 5-aminolevulinic acid fluorescence signals after resection of MR-enhancing lesions and its prognostic significance in glioblastoma.

PURPOSE: This study investigated the degree of tumor cell infiltration in the tumor cavity and ventricle wall based on fluorescent signals of 5-aminolevulinic acid (5-ALA) after removal of the magnetic resonance (MR)-enhancing area and analyzed its prognostic significance in glioblastoma. METHODS: Twenty-five newly developed isocitrate dehydrogenase (IDH)-wildtype glioblastomas with complete resection both of MR-enhancing lesions and strong purple fluorescence on resection cavity were retrospectively analyzed. The fluorescent signals of 5-ALA were divided into strong purple, vague pink, and blue colors. The pathologic findings were classified into massively infiltrating tumor cells, infiltrating tumor cells, suspicious single-cell infiltration, and normal-appearing cells. The pathological findings were analyzed according to the fluorescent signals in the resection cavity and ventricle wall. RESULTS: There was no correlation between fluorescent signals and infiltrating tumor cells in the resection cavity (p = 0.199) and ventricle wall (p = 0.704) after resection of the MR-enhancing lesion. The median progression-free survival (PFS) and median overall survival (OS) were 12.5 (± 2.1) and 21.1 (± 3.5) months, respectively. In univariate analysis, the presence of definitive infiltrating tumor cells in the resection cavity and ventricle wall was significantly related to the PFS (p = 0.002) and OS (p = 0.027). In multivariate analysis, the absence of definitive infiltrating tumor cells improved PFS (hazard ratio: 0.184; 95% CI: 0.049-0.690, p = 0.012) and OS (hazard ratio: 0.124; 95% CI: 0.015-0.998, p = 0.050). CONCLUSIONS: After resection both of the MR-enhancing lesions and strong purple fluorescence on resection cavity, there was no correlation between remnant fluorescent signals and infiltrating tumor cells. The remnant definitive infiltrating tumor cells in the resection cavity and ventricle wall significantly influenced the prognosis of patients with glioblastoma. Aggressive surgical removal of infiltrating tumor cells may improve their prognosis.

MR imaging of hepatic iron overload in rat.

To investigate the relationship of hepatic signal intensity and T2 with histologic grading in an animal model of oral iron overload and to determine the duration of feeding necessary to produce abnormalities detectable on magnetic resonance (MR) images, hepatic iron overload was induced in 12 rats by feeding them a diet supplemented with 4% carbonyl iron for 2-11 weeks. Iron overload seen on MR images was graded independently and blindly by two radiologists as normal, mild, moderate, or severe. The rats were killed, and histologic findings were graded blindly by four pathologists using a similar subjective scale. Hepatic T2 values were estimated from spin-echo images. In the rats with iron overload, intracellular iron deposition was noted on histologic studies. On MR images, hepatic signal intensity and T2 decreased after only 2 weeks of dietary iron overload, and both continued to decrease with longer duration of feeding. There was significant correlation between iron overload duration and changes on MR images and between MR images and histologic grading (r = .92, P = .0001 for both). The mean T2 of hepatic iron overload decreased with longer duration of feeding.