Abstract
BACKGROUND: Perfusion computed tomography (PCT) can reflect the blood flow and the capillary condition, which is valuable for the preoperative differentiation of brain tumors. We investigated PCT parameters at tumor and peritumoral lesions, and assessed the efficacy of the parameters for differentiating glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and brain metastasis (METs). MATERIALS AND METHODS: We performed PCT in patients with 39 supratentorial malignant brain tumors (22 GBM, 6 PCNSL, 11 METs). Regions of interests were placed manually at the enhanced tumor area confirmed on contrast-enhanced CT image, peritumoral lesion and contralateral white matter. CBV (cerebral blood volume), CBF (cerebral blood flow), MTT (mean transit time) and PS (permeability surface) were measured. Normalized ratio of the PCT parameters obtained by using normal-appearing white matter as the denominator were used for analyses in all cases, respectively. Mann-Whitney U Test was used to compare all of the parameters. Logistic regression and receiver operating characteristic analyses were constructed to evaluate the preoperative diagnostic performance. RESULT: nCBV and nCBF of GBM were statistically higher than those of PCNSL (p= 0.0002, 0.0005) and METs (p= 0.0028, 0.0044) at tumor area. At peritumoral lesion, nCBV and nCBF of GBM were statistically higher than those of PCNSL (p= 0.0002, 0.0005) and MET (p= 0.0028, 0.0044). nMTT of METs was statistically higher than that of GBM and PCNSL at peritumoral lesion (p= 0.0001, p= 0.0007). The logistic regression and receiver operating characteristic analyses indicated that GBM could differentiate from PCNSL and METs with sensitivity and specificity with 86.3% and 94.1%. METs could differentiate from PCNSL and MET with sensitivity and specificity with 90.9% and 96.4%. CONCLUSION: PCT parameters of GBM, PCNSL, METs and their peritumoral lesion provide the valuable information for the preoperative differentiation among these three types of tumors.