Wenting Rui1, Yue Wu2, Zengyi Ma3, Yongfei Wang4, Yin Wang5, Xiao Xu6, Junhai Zhang7, Zhenwei Yao8. 1. Department of Radiology, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: wennyrui@126.com. 2. Department of Radiology, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: wuyue0115@foxmail.com. 3. Department of Neurosurgery, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: zengyima@foxmail.com. 4. Department of Neurosurgery, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: eamns@hotmail.com. 5. Department of Neuropathology, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: yinwang88@hotmail.com. 6. GE Healthcare Life Sciences, GE Chinese Science and Technology Park, Huatuo Road, Shanghai, 201203, PR China. Electronic address: adamxu1027@163.com. 7. Department of Radiology, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: jhzhang007@fudan.edu.cn. 8. Department of Radiology, Huashan Hospital, Fudan University, Mid Wulumuqi Road, Shanghai, 200040, PR China. Electronic address: zwyao@fudan.edu.cn.
Abstract
OBJECTIVES: To explore the value of magnetic resonance textural analysis (MRTA) in assessing consistency of pituitary macroadenoma (PMA) based on contrast enhanced (CE) three-dimensional sampling perfection with application-optimized contrasts by using different flip angle evolution (3D-SPACE) images. MATERIALS AND METHODS: Fifty-three patients with PMAs that underwent CE 3D-SPACE scanning by 3.0 T MRI and endoscopic trans-sphenoidal surgery were included in the present study. Consistency levels of PMAs were evaluated intraoperatively by two neurosurgeons. Each resection specimen was stained with H&E and anti-collagen IV. MRTA was conducted and texture features were calculated. An unpaired t-test was used to analyze the differences of texture features between soft and hard PMAs. ROC curves by individual and combined features were used to calculate the diagnostic accuracy of MRTA in predicting PMA consistency. RESULTS: First-order energy and second-order correlation negatively correlated with hard PMAs, while first-order entropy and second-order variance, sum variance, and sum entropy positively correlated with stiffness. All showed significant differences between soft and hard PMAs (P < 0.05). Diagnostic accuracy of combined negative features could achieve an AUC of 0.819, sensitivity of 88.9%, specificity of 61.5%, PPV of 70.6%, NPV of 84.2% and positive features could achieve an AUC of 0.836, sensitivity of 85.2%, specificity of 69.2%, PPV of 74.2%, NPV of 81.8% (P < 0.001). CONCLUSION: MRTA using CE 3D-SPACE images is helpful for assessing PMA consistency preoperatively and noninvasively.
OBJECTIVES: To explore the value of magnetic resonance textural analysis (MRTA) in assessing consistency of pituitary macroadenoma (PMA) based on contrast enhanced (CE) three-dimensional sampling perfection with application-optimized contrasts by using different flip angle evolution (3D-SPACE) images. MATERIALS AND METHODS: Fifty-three patients with PMAs that underwent CE 3D-SPACE scanning by 3.0 T MRI and endoscopic trans-sphenoidal surgery were included in the present study. Consistency levels of PMAs were evaluated intraoperatively by two neurosurgeons. Each resection specimen was stained with H&E and anti-collagen IV. MRTA was conducted and texture features were calculated. An unpaired t-test was used to analyze the differences of texture features between soft and hard PMAs. ROC curves by individual and combined features were used to calculate the diagnostic accuracy of MRTA in predicting PMA consistency. RESULTS: First-order energy and second-order correlation negatively correlated with hard PMAs, while first-order entropy and second-order variance, sum variance, and sum entropy positively correlated with stiffness. All showed significant differences between soft and hard PMAs (P < 0.05). Diagnostic accuracy of combined negative features could achieve an AUC of 0.819, sensitivity of 88.9%, specificity of 61.5%, PPV of 70.6%, NPV of 84.2% and positive features could achieve an AUC of 0.836, sensitivity of 85.2%, specificity of 69.2%, PPV of 74.2%, NPV of 81.8% (P < 0.001). CONCLUSION: MRTA using CE 3D-SPACE images is helpful for assessing PMA consistency preoperatively and noninvasively.
Authors: Paul Windisch; Carole Koechli; Susanne Rogers; Christina Schröder; Robert Förster; Daniel R Zwahlen; Stephan Bodis Journal: Cancers (Basel) Date: 2022-05-27 Impact factor: 6.575