Preoperative Assessment Of Vascularisation Of Spinal Tumor: Comparison With Dynamic Contrast-enhanced MRI And DSA

Background: Selective arterial embolization is a safe and effective practice that iscommonly used for surgery of spinal tumors. Studies found that SAE is only meaningfulfor hypervascular tumors and may not be necessary for hypovascular ones. Therefore,evaluation before operation on spinal tumor vascularization has become a critical subjectof clinical practice. Dynamic contrast-enhanced MRI (DCE-MRI) can identify theabnormal tumor vascularity and differentiate hypovacular and hypervascular spinaltumors through providing valuable information before surgery. This study focuses onDCE-MRI parameters of the spinal tumors in patients scheduled for surgery, comparesthem with DSA staining (gold standard), studies the significance of its evaluation onspinal tumor vascularization and evaluates the clinical value of preoperative embolizationin patients with spinal tumors.Part1Evaluation of DCE MR and DSA on Spinal TumorVascularizationPurpose: The purpose is to discuss the value of DCE-MRI on evaluation of spinal tumorvascularization compared with DSA staining.Methods: Between July2012and January2014,40patients with untreated spinal tumors,including23males and17females with an average age of47(16-74years old),underwent DCE MR and conventional MR. Preoperative SAEs were completed andpathological specimens were obtained from surgical resection. DCE curve parameters andDSA staining appearance of the spinal tumor were studied to discuss their relevance.ROC curves were drawn to get AUC referred to DSA staining.Results: The correlation of Slopemaxof the tumor DCE curve and RSlopemax between tumor andnormal vertebral body S1/S0with DSA staining were0.847,0.899(all P<0.01) respectively.The correlation of MR conventional enhancement with DSA staining was0.289, P>0.05.S1/S0and Slopemaxpredicted the ROC AUC of hypovascular spinal tumors were0.971,0.930(P<0.01) respectively, and the cut-off value of S1/S0at this point was1.325(with sensitivity of87.5%, specificity of100%). S1/S0and Slopemaxpredicted the ROC AUC of hypervascularspinal tumors were0.988and0.979(P<0.01) respectively, and the cut-off value of S1/S0at thispoint was1.870(with sensitivity of100%, specificity of96.4%). MR enhancement degreepredicted the ROC AUC of hypovascular spinal tumors and hypervascular ones were0.547and0.697respectively, all P>0.05.Conclusions: Compared with MR conventional enhanced scanning, MR DCE curve canoffer more accurate diagnosis of spinal tumor vascularization, which is a better indicatorwith potential clinical value. Part2Value of Diagnosing the Benign and Malignant Tumorof Spine by DCE MRPurpose: The purpose is to evaluate the clinical value of DCE MRI parameters indifferentiating the benign and malignant tumor of spine.Methods: Between July2012and January2014,40patients with untreated spinal tumors,including23males and17females with an average age of47(16-74years old),underwent DCE MRI and conventional enhanced scanning. Pathological specimens wereobtained from surgical resection. The performance of DCE MRI parameters and MRIenhancement degree in benign and malignant tumor of spine were studied.Results: Statistical differences were found in degree of MR enhancement in benign andmalignant tumor of spine, Z=-4.019and P<0.01; while no significant statistical differences ofDCE curve, Z=-0.556and P=0.578. No significant statistical difference was found in theperformance of indicators like RSlopemax between tumor and normal vertebral body S1/S0,tumor Slopemax,maximum enhancement (Emax), maximum relative enhancement (ERmax) andtime to peak (TTP) in benign and malignant tumor groups, all P>0.05.Conclusions: The diagnostic performance of DCE MRI parameters in benign and malignanttumor of spine showed no significant statistical differences and it could not be usedseparately as the indicator of benign and malignant tumor of spine without morphologychanging Part3Clinical Significance of Preoperative Embolization ofSpinal TumorPurpose:The purpose is to compare the differences of bleeding in surgical resection afterpreoperative embolization, perioperative blood transfusion and complications of spinaltumor and to evaluate the clinical value of preoperative embolization.Methods:Between July2012and January2014,80patients were blanned to undergo spinal tumorsurgical resection, including40patients (24males,16females) with an average age of47.0±15.9underwent preoperative embolization, of which18suffered thoracic lesions,Lumbar spine lesions and13sacral vertebral lesions; and the rest40patients (25males,15females) were the control group (not received preoperative embolization) with anaverage age of46.1±17.9, of which17suffered thoracic lesions,13Lumbar spine lesionsand9sacral vertebral lesions. Differences of intraoperative bleeding, perioperativeblood transfusion and complications between the two groups were compared.Results: Among the40patients of embolization group, radiography practices were cond-ucted on188tumor blood supply vessels, of which embolization of167vessels wereperformed. Intraoperative blood loss of the embolization group was1740±746ml, andthat of the control group was2516±1374ml; significant statistical differences were found(P<0.01). Perioperative blood transfusion of the embolization group was1425±1348ml,and that of the control group was1845±1348ml; no significant statistical differences wasfound (P>0,05). No embolization complication and surgical complication was found inthe embolization group, while2surgical complications were found in the control group.Conclusions: Preoperative embolization of spinal tumor is a safe and effective practicein significantly reducing intraoperative blood loss and complications.