Quantitative analysis of CT-perfusion parameters in the evaluation of brain gliomas and metastases.

ABSTRACT

BACKGROUND: The paper reports a quantitative analysis of the perfusion maps of 22 patients, affected by gliomas or by metastasis, with the aim of characterizing the malignant tissue with respect to the normal tissue. The gold standard was obtained by histological exam or nuclear medicine techniques. The perfusion scan provided 11 parametric maps, including Cerebral Blood Volume (CBV), Cerebral Blood Flow (CBF), Average Perfusion (Pmean) and Permeability-surface area product (PS). METHODS: The perfusion scans were performed after the injection of 40 ml of non-ionic contrast agent, at an injection rate of 8 ml/s, and a 40 s cine scan with 1 s interval was acquired. An expert radiologist outlined the region of interest (ROI) on the unenhanced CT scan, by using a home-made routine. The mean values with their standard deviations inside the outlined ROIs and the contralateral ROIs were calculated on each map. Statistical analyses were used to investigate significant differences between diseased and normal regions. Receiving Operating Characteristic (ROC) curves were also generated. RESULTS: Tumors are characterized by higher values of all the perfusion parameters, but after the statistical analysis, only the PS, PatRsq (Patlak Rsquare) and Tpeak (Time to Peak) resulted significant. ROC curves, confirmed both PatRsq and PS as equally reliable metrics for discriminating between malignant and normal tissues, with areas under curves (AUCs) of 0.82 and 0.81, respectively. CONCLUSION: CT perfusion is a useful and non invasive technique for evaluating brain neoplasms. Malignant and normal tissues can be accurately differentiated using perfusion map, with the aim of performing tumor diagnosis and grading, and follow-up analysis.