OBJECTIVE: Different perfusion metrics were investigated to determine the accuracy of Perfusion CT (PCT) in differentiating recurrence of brain metastases from radiation necrosis in patients who previously underwent stereotactic radiation therapy (SRT). PATIENTS AND METHODS: Twenty patients previously treated with SRT underwent PCT examination of the brain. Normalized Cerebral Blood Volume (nCBV) values within the region of interest (ROI) were calculated. Fractional volumes at nCBV values between 1.0 and 2.25, defined as V(1.0)-V(2.25), were derived as alternative metrics and compared to the conventional method based on the mean CBV value within the lesion (nCBV(mean)). The Mann-Whitney test was used to compare the two patient's groups with recurrence and radiation necrosis with respect to the different nCBV metrics. Predictive powers and optimal thresholds for both the nCBV(mean) and the V(1.0)-V(2.25) were evaluated using the Receiver Operating Characteristic Curves. The gold standard was represented either by the histopathological examination or the Magnetic Resonance (MR) imaging follow-up longer than six months. RESULTS AND CONCLUSION: The differences between the patient's group with recurrence and that with radiation necrosis resulted statistically significant for all the metrics, showing the lowest p-value for V(1.75) and V(2). The metrics based on the fractional volumes were found to show higher predictive powers, with the highest value of 0.96 for V(2.0). Quantitative analysis of the CBV map deriving different metrics may potentially improve the diagnostic accuracy of PCT in differentiating brain metastasis recurrence from radiation necrosis.
OBJECTIVE: Different perfusion metrics were investigated to determine the accuracy of Perfusion CT (PCT) in differentiating recurrence of brain metastases from radiation necrosis in patients who previously underwent stereotactic radiation therapy (SRT). PATIENTS AND METHODS: Twenty patients previously treated with SRT underwent PCT examination of the brain. Normalized Cerebral Blood Volume (nCBV) values within the region of interest (ROI) were calculated. Fractional volumes at nCBV values between 1.0 and 2.25, defined as V(1.0)-V(2.25), were derived as alternative metrics and compared to the conventional method based on the mean CBV value within the lesion (nCBV(mean)). The Mann-Whitney test was used to compare the two patient's groups with recurrence and radiation necrosis with respect to the different nCBV metrics. Predictive powers and optimal thresholds for both the nCBV(mean) and the V(1.0)-V(2.25) were evaluated using the Receiver Operating Characteristic Curves. The gold standard was represented either by the histopathological examination or the Magnetic Resonance (MR) imaging follow-up longer than six months. RESULTS AND CONCLUSION: The differences between the patient's group with recurrence and that with radiation necrosis resulted statistically significant for all the metrics, showing the lowest p-value for V(1.75) and V(2). The metrics based on the fractional volumes were found to show higher predictive powers, with the highest value of 0.96 for V(2.0). Quantitative analysis of the CBV map deriving different metrics may potentially improve the diagnostic accuracy of PCT in differentiating brain metastasis recurrence from radiation necrosis.