gms | German Medical Science

Objective: Tumor perfusion is a crucial key feature in malignant gliomas. So far, depicting cerebral perfusion is cumbersome by expensive and rarely available Xenon-CT, 15-O H₂O PET, or by less accurate and time consuming MRI. Measuring the tumor perfusion by 128-row CT scanners has being shown useful in determinating histology of cerebral neoplasms. This study was set up to evaluate the regional histological features responsible for these volume perfusion CT (VPCT) measurements.

Methods: Eight patients with malignant gliomas and candidates for tumor removal received a contrast enhanced MRI and VPCT. The latter was performed using a 128-detector row CT with constant periodic bidirectional table movement. Thus, entire organs larger than the detector width can be imaged. Total CT data acquisition time was 45 s. The intravenous contrast injection protocol consisted of 36 ml of highly iodinated contrast agent at a flow rate of 6ml/sec, followed by a saline chasing bolus of 20 ml. During fluorescence guided navigated tumor resection a total of 29 samples were obtained. The biopsy site was captured by the neuronavigational device. According to these sites, absolute values for intratumoral cerebral blood flow (CBF), cerebral blood volume (CBV) and permeability (K^trans) had been retrospectively determined from the VPCT data using commercial software. For histological evaluation cell and vascular morphology and density were determined on HE and anti-CD34 slides. The latter was quantified by pixel-based segmentation. Pearson correlation and Spearman test was applied for statistical analysis.

Results: Histology revealed glioblastoma in four, malignant gliomas in two and gliosarcoma and anaplastic astrocytoma in one patient each. All but two patients with newly diagnosed tumors had recurrent disease. A high correlation could be found between vascular density and CBF, CBV and K^trans (Pearson correlation p=0.0039, p=0.0141 and p=0.0282, respectively). No significant correlation was found according semiquantitative cell and vascular morphology. The correlation of VPCT data and cell density was also not significant.

Conclusions: Regional intratumor perfusion determined by VPCT correlated with histological vascular density. Thus VPCT might advance as a quick diagnostic tool in glioma patients specific for the tumors vascular pathology. These finding are currently validated in a larger study.