Evaluating The Usefulness Of Pseudo-continuous Arterial Spin Labeling Imaging In Distinguishing High-grade Gliomas From Primary Cerebral

Primary malignant brain tumors commonly encountered in clinicalpractice include high-grade gliomas and lymphomas. Lymphomas arerelatively less seen,but recently it,s incidence is rising. Differentiation ofhigh-grade gliomas from lymphomas is crucial prior to designing the optimaltreatment plan for a patient, as curative surgery is unnecessary in the latter, inwhich combined high-dose chemotherapy and radiation therapy usually playthe pivotal therapeutic role.The study of tissue perfusion provides functional information about thelesion characterization and the prognosis in a great many disorders. Differenttechniques are used in everyday clinical practice: the perfusion CT-scan,T2*-weighted perfusion MRI (or dynamic susceptibility contrast MRI:DSC-MRI) and the techniques of nuclear medicine, in particular single-photonemission computed tomography (SPECT) whose clinical use is much easierthan positron emission tomography (PET). However, all of these techniquesare invasive since they require the exogenous injection of a contrast media oruse ionizing radiation(CT-scan and nuclear medicine). Since the early1990’s,a quantitative MRI technique has been developed to study perfusion withoutcontrast media, the arterial spin labeling (ASL)perfusion technique.Pseudo-continuous arterial spin labeling is a modern ASL technique,it ismore accurate and reliable. As an endogenous tracer, ASL uses arterial waterprotons labeled by radiofrequency pulses,so it is more easily accepted bypatients. Now, the usefulness of this new technique in the differentialdiagnosis of brain tumors has not been sufficiently evaluated in the previousliterature. Purpose:1Know more knowledge about the new technique, master the methodand the principle of ASL;2Conbine the histopathology and the MR imaging of the brain tumorsand further master the relative knowledge of the tumors;3To evaluate the usefulness and value of pCASL imaging indifferentiating high-grade gliomas from lymphomas.Methods:1We retrospectively reviewed the MR imagings of the high-grade gliomapatients and lymphoma patients, who underwent conventional MR imaging atHandan Central Hospital between2010and2013.2The patient without preoperative pCASL imaging were excluded.Similarly, the lymphoma patients who had undergone the chemotherapy orsteroid therapy before pCASL imaging were excluded.3For each tumor, one neuroradiologist measured the absolute TBF on thepCASL perfusion map at five regions of interest(ROIs) of17.56mm2,randomly selected within the area corresponding to the enhancing portion ofthe tumor. Subsequently,mean tumor blood flow(mTBF) was calculated foreach tumor.4As for the control, five additional ROIs were randomly drawn withinthe ray matter of normal contralateral parenchyma at the same slice levels asthe tumor ROIs, and their mean was also calculated(mBFgm).5To minimize the influence of physiological inter-individual variation inthe basal TBF, rTBF was calculated by the following equation: rTBF=（mTBF/mBFgm）×100.6T test was used to analyze whether mTBF and rTBF values of thehigh-grade gliomas significantly differed from those of the lymphomas.7For both mTBF and rTBF, receiver operating characteristic (ROC)curves were constructed to determin the optimum threshold to differentiatehigh-grade gliomas from lymphomas.8To assess the superiority of rTBF over mTBF in the differentialdiagnosis of the two tumor groups, area under the receiver operating characteristic curve(AUC) values were compared using the method of Delonget al.Results:1The final cohort consisted of31patients with20histopathologicallyconfirmed high-grade gliomas(age range,22-65years; mean age,47.4years)and11lymphomas(age range,38-82years; mean age,60.9years).2The high-grade gliomas included12GBMs,3AAs, and5recurrentGBMs.One patient with recurrent lymphoma refractory to chemotherapy wasincluded in the study.3The mTBF value of the high-grade glioma was significantly higherthan that of the lymphomas (92.1±34.7mL/min/100mg vs.53.6±30.5mL/min/100mg，P=0.008). In addition, the rTBF value was significantlyhigher for the high-grade gliomas than for the lymphomas（182.3±69.5vs.92.5±44.9，P=0.002）.4In ROC analysis, with a cut-off value of57.9mL/min/100mg, mTBFcould differentiate the high-grade gliomas from the lymphomas with asensitivity of90%and a specificity of54.6%. At the same time, with a cut-offvalue of141.1, rTBF could differentiate the high-grade gliomas from thelymphomas with a sensitivity of65.0%and a specificity of100.0%.5mTBF and rTBF did not significantly differ in their AUC values(0.873vs.0.909，P=0.287).Conclusions:1Using the two perfusion parameters--mTBF and rTBF--can effectivelydifferentiate the high-grade gliomas from lymphomas.2The superiority of rTBF over mTBF in the differential diagnosis of thetwo tumor groups has no significant difference.3The pCASL perfusion imaging can be an effective modality fordifferentiating high-grade gliomas from lymphomas. It is the developmenttrend of the ASL technique.