Semiquantitative Analysis Using Thallium-201 SPECT for Differential Diagnosis Between Tumor Recurrence and Radiation Necrosis After Gamma Knife Surgery for Malignant Brain Tumors

Abstract

Semiquantitative analysis of thallium-201 chloride single photon emission computed tomography (201Tl SPECT) was evaluated for the discrimination between recurrent brain tumor and delayed radiation necrosis after gamma knife surgery (GKS) for metastatic brain tumors and high-grade gliomas. The medical records were reviewed of 75 patients, including 48 patients with metastatic brain tumor and 27 patients with high-grade glioma who underwent GKS in our institution, and had suspected tumor recurrence or radiation necrosis on follow-up neuroimaging and deteriorating clinical status after GKS. Analysis of 201Tl SPECT data used the early ratio (ER) and the delayed ratio (DR) calculated as tumor/normal average counts on the early and delayed images, and the retention index (RI) as the ratio of DR to ER. A total of 107 tumors were analyzed with 201Tl SPECT. Nineteen lesions were removed surgically and histological diagnoses established, and the other lesions were evaluated with follow-up clinical and neuroimaging examinations after GKS. The final diagnosis was considered to be recurrent tumor in 65 lesions and radiation necrosis in 42 lesions. Semiquantitative analysis demonstrated significant differences in DR (P=.002) and RI (P<.0001), but not in ER (P=.372), between the tumor recurrence and radiation necrosis groups, and no significant differences between metastatic brain tumors and high-grade gliomas in all indices (P=.926 for ER, P=.263 for DR, and P=.826 for RI). Receiver operating characteristics analysis indicated that RI was the most informative index with the optimum threshold of 0.775, which provided 82.8% sensitivity, 83.7% specificity, and 82.8% accuracy. Semiquantitative analysis of 201Tl SPECT provides useful information for the differentiation between tumor recurrence and radiation necrosis in metastatic brain tumors and high-grade gliomas after GKS, and the RI may be the most valuable index for this purpose.