Clin Neuroradiol DOI 10.1007/s00062-015-0382-2
O r i g i n a l A rt i c l e
Preoperative Evaluation of Patients with Hemifacial Spasm by Three-dimensional Time-of-Flight (3D-TOF) and Threedimensional Constructive Interference in Steady State (3D-CISS) Sequence J.-m. Jia · H. Guo · W.-j. Huo · S.-w. Hu · F. He · X.-d. Sun · G.-j. Lin
Received: 5 January 2015 / Accepted: 26 February 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose The purpose of this study was to investigate and evaluate the accuracy and the preoperative diagnostic value of high-resolution magnetic resonance imaging (MRI) techniques, three-dimensional time-of-flight (3D-TOF) and three-dimensional constructive interference in steady state (3D-CISS) sequence, solely or in combination for the detection of the relationship between the facial nerve and adjacent vessels in patients with hemifacial spasm (HFS). Methods A total of 95 patients with primary HFS were subject to 3D-TOF and 3D-CISS MRI. The MR images were then used to evaluate the anatomical neurovascular relationships, and the results were compared with the surgical findings. We categorized the neurovascular relationship into three types: compression, contact, and neighboring or distant. Compression and/or contacts between root exit zone (REZ) and vessels were defined as positive, whereas neighboring or distant was considered to be negative. Results 3D-TOF combined with 3D-CISS assessment showed that 94 of 95 patients had artery compression or contact at REZ, whereas the remaining patient had compression at the peripheral branch of the facial nerve but not at REZ. The positive rates and the overall accuracy were 98.95 and 100 %, respectively, for the 3D-TOF combined with 3D-CISS assessment; 92.63 and 93.68 %, respectively, for the 3D-TOF assessment; and 85.26 and 86.32 %, respectively, for the 3D-CISS assessment. The positive rates and
G.-j. Lin, PhD () · J.-m. Jia · H. Guo · W.-j. Huo · S.-w. Hu · F. He · X.-d. Sun Department of Neurosurgery, Aerospace Center Hospital, 15 Yuquan Road, Haidian District, 100049 Beijing, P. R. China e-mail: [email protected]
overall accuracy for the 3D-TOF combined with 3D-CISS assessment was significantly higher than those for the 3DTOF or 3D-CISS assessment. Conclusions Combination of 3D-TOF with 3D-CISS imaging well delineates the relationship between the facial nerve and adjacent vessels in terms of increased positive rates and accuracy. Keywords Hemifacial spasm · Magnetic resonance imaging · Root exit zone Introduction Hemifacial spasm (HFS) is a disorder that is characterized by complex symptoms related to the hyperactive motor dysfunction of the facial nerve [1]. Although HFS is not a life-threatening disease, it can severely reduce patients’ life quality [2]. It is generally believed that HFS is caused by vascular compression of the root exit zone (REZ) of the facial nerve [3–6], except for a limited number of cases resulting from known underlying diseases [7] such as tumors, demyelinating disorders, or infections [8, 9]. Microvascular decompression (MVD), the most effective therapy for HFS, treats the disease by eliminating neurovascular compression [10–12, 4]. Although the decision to perform an MVD of the facial nerve is based on clinical history and assessment, preoperative imaging of the neurovascular relationships at the REZ is essential to exclude a nonvascular effect. In addition, more detailed magnetic resonance imaging (MRI) through delineating the responsible vessel may help the preoperative planning because MVD may be associated with operative complications, e.g., hearing impairment, facial nerve paresis, and meningitis [13]. Moreover, the postoperative recurrence rate is increased when compression is responsible by a
13
2
small artery or vein [14]. If detailed imaging can distinguish these, the patient will be more effectively counseled. Computed tomography is obviously an unsatisfactory investigation under this condition [15]. MRI is more accurate [16], but the routine T1-weighted (T1W) and T2-weighted (T2W) images invariably fail to adequately delineate neurovascular relationships at the REZ of the nerve [17]. Recently, due to the advancement of high-resolution MRI (HR-MRI), three-dimensional time-of-fight (3D-TOF) and three-dimensional constructive interference in steady state (3D-CISS) sequences were used in patients with HFS. Such a procedure has emerged to be a useful tool for assessing neurovascular compression [5, 18]. In the present study, we included 95 patients with HFS at the Aerospace Clinical Medical College of Peking University during September 2012 to July 2013. The aim of this study was to analyze the accuracy and preoperative diagnostic value of HR-MRI techniques by using 3D-TOF and 3D-CISS sequence solely or in combination for the detection of the relationship between the facial nerve and adjacent vessels in patients with HFS. Patients and Methods Patients The study was approved by the ethics committee of Aerospace Center Hospital, China, and written informed consent was obtained from all subjects. A total of 95 patients (20 male and 75 female, with a mean age of 49.2 years) with primary HFS were included during September 2012 to July 2013. The diagnosis of HFS was based on the clinical history and physical examination and then was confirmed with an abnormal motor response on electromyography. The duration of disease varied from 6 months to 23 years, with a mean value of 6.7 years. All the cases were of HFS on one side (left side: 51 cases, right side: 44 cases). For the purposes of this study, we excluded all patients with a confounding diagnosis (such as secondary HFS, tumor, disseminated sclerosis, and facial paralysis caused by trauma and cold). Patients who received botulinum toxin therapy within half a year were also excluded. In addition, patients with poor physical condition who could not be surgically treated were also excluded. MR Examination All patients underwent MR preoperatively using a 1.5-T MR imager (Siemens Magnetom, Erlangen, Germany), including three following sequences: high-spatial-resolution 3D-TOF and 3D-CISS-MRA.
13
J.-m. Jia et al.
3D-TOF MRA (repetition time (TR) = 30 ms, echo time (TE) = 5.0 ms, flip angle = 20°, NEX = 1, section thickness = 0.9 mm, gap = 0, field of view (FOV) = 240 mm, and matrix size = 512 × 512) and 3D-CISS sequences (TR = 10.9 ms, TE = 5.4 ms, flip angle = 70°, NEX = 1, section thickness = 0.8 mm, gap = 0, FOV = 230 mm, and matrix size = 173 × 256) were used in the present imaging study. According to the midsagittal plane of brain stem, transverse scanning of the whole brain stem was performed. Based on the 3D-TOF axial images, multiplanar reformatting in the oblique sagittal (left and right) and coronal planes, parallel and perpendicular to the subarachnoid cisterns, was performed on the workstation. The maximum intensity projection (MIP) and volume rendering were used to reconstruct blood vessels so that the offending vessels of the facial nerve can be displayed. The examination was regarded as positive if the definite neurovascular conflict between the facial nerve and a vessel can be seen by 3D-TOF and/or 3D-CISS sequences (Figs. 1 and 2). Imaging Analysis All the images were evaluated respectively by two experienced radiologists, who were blinded to any knowledge of which side the spasm was located at and the operative findings. Decision was obtained by consensus. We categorized the relationship between REZ and neighboring vessels into three types based on the axial, coronal, and oblique sagittal images: (1) compression, visualized by axial plus coronal and/or oblique sagittal images, defined as the presence of obvious vascular indentation at the REZ of the facial nerve by either vessels/facial nerve displacement or the course of anomalous vessels; (2) contacts, considered to be present if the minimum distance between vessel and the REZ was not more than the maximum diameter of the vessel by axial plus coronal and/or oblique sagittal images; (3) neighboring or distant, considered to be present if the minimum distance between the vessel and the REZ was larger than the maximum diameter of vessels by axial plus coronal and oblique sagittal images. REZ was defined as the point where the facial nerve exited from the pons. Compression and/or contacts between REZ and vessels were defined as positive, whereas neighboring or distant was considered to be negative. Surgical Technique All the procedures were performed by one experienced neurosurgeon via a lateral retrosigmoid suboccipital approach, which has been previously described in the literature [19]. Necessary judgments were made for the identification of offending vessels according to the distribution as well as arrangement.
Preoperative Evaluation of Patients with Hemifacial Spasm by Three-dimensional Time-of-Flight (3D-TOF)
3
Table 1 Comparison of results from preoperative imaging assessment with intraoperative findings (cases, n = 95) Methods Sensitivity Specificity Positive predictive Negative predictive Accuracy Positive rates value value 3D-TOF 88/94 (93.61 %) 1/1 (100 %) 88/88 (100 %) 1/7 (14.29 %) 89/95 (93.68 %)* 92.63 %* 3D-CISS 81/94 (86.17 %) 1/1 (100 %) 81/81 (100 %) 1/14 (7.14 %) 82/95 (86.32 %)** 85.26 %** 3D-TOF + CISS 94/94 (100 %) 1/1 (100 %) 94/94 (100 %) 1/1 (100 %) 95/95 (100 %) 98.95 % All results were obtained by comparison with 3D-TOF + 3D-CISS 3D-TOF three-dimensional time-of-fight, 3D-CISS three-dimensional constructive interference in steady state *p
O r i g i n a l A rt i c l e
Preoperative Evaluation of Patients with Hemifacial Spasm by Three-dimensional Time-of-Flight (3D-TOF) and Threedimensional Constructive Interference in Steady State (3D-CISS) Sequence J.-m. Jia · H. Guo · W.-j. Huo · S.-w. Hu · F. He · X.-d. Sun · G.-j. Lin
Received: 5 January 2015 / Accepted: 26 February 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose The purpose of this study was to investigate and evaluate the accuracy and the preoperative diagnostic value of high-resolution magnetic resonance imaging (MRI) techniques, three-dimensional time-of-flight (3D-TOF) and three-dimensional constructive interference in steady state (3D-CISS) sequence, solely or in combination for the detection of the relationship between the facial nerve and adjacent vessels in patients with hemifacial spasm (HFS). Methods A total of 95 patients with primary HFS were subject to 3D-TOF and 3D-CISS MRI. The MR images were then used to evaluate the anatomical neurovascular relationships, and the results were compared with the surgical findings. We categorized the neurovascular relationship into three types: compression, contact, and neighboring or distant. Compression and/or contacts between root exit zone (REZ) and vessels were defined as positive, whereas neighboring or distant was considered to be negative. Results 3D-TOF combined with 3D-CISS assessment showed that 94 of 95 patients had artery compression or contact at REZ, whereas the remaining patient had compression at the peripheral branch of the facial nerve but not at REZ. The positive rates and the overall accuracy were 98.95 and 100 %, respectively, for the 3D-TOF combined with 3D-CISS assessment; 92.63 and 93.68 %, respectively, for the 3D-TOF assessment; and 85.26 and 86.32 %, respectively, for the 3D-CISS assessment. The positive rates and
G.-j. Lin, PhD () · J.-m. Jia · H. Guo · W.-j. Huo · S.-w. Hu · F. He · X.-d. Sun Department of Neurosurgery, Aerospace Center Hospital, 15 Yuquan Road, Haidian District, 100049 Beijing, P. R. China e-mail: [email protected]
overall accuracy for the 3D-TOF combined with 3D-CISS assessment was significantly higher than those for the 3DTOF or 3D-CISS assessment. Conclusions Combination of 3D-TOF with 3D-CISS imaging well delineates the relationship between the facial nerve and adjacent vessels in terms of increased positive rates and accuracy. Keywords Hemifacial spasm · Magnetic resonance imaging · Root exit zone Introduction Hemifacial spasm (HFS) is a disorder that is characterized by complex symptoms related to the hyperactive motor dysfunction of the facial nerve [1]. Although HFS is not a life-threatening disease, it can severely reduce patients’ life quality [2]. It is generally believed that HFS is caused by vascular compression of the root exit zone (REZ) of the facial nerve [3–6], except for a limited number of cases resulting from known underlying diseases [7] such as tumors, demyelinating disorders, or infections [8, 9]. Microvascular decompression (MVD), the most effective therapy for HFS, treats the disease by eliminating neurovascular compression [10–12, 4]. Although the decision to perform an MVD of the facial nerve is based on clinical history and assessment, preoperative imaging of the neurovascular relationships at the REZ is essential to exclude a nonvascular effect. In addition, more detailed magnetic resonance imaging (MRI) through delineating the responsible vessel may help the preoperative planning because MVD may be associated with operative complications, e.g., hearing impairment, facial nerve paresis, and meningitis [13]. Moreover, the postoperative recurrence rate is increased when compression is responsible by a
13
2
small artery or vein [14]. If detailed imaging can distinguish these, the patient will be more effectively counseled. Computed tomography is obviously an unsatisfactory investigation under this condition [15]. MRI is more accurate [16], but the routine T1-weighted (T1W) and T2-weighted (T2W) images invariably fail to adequately delineate neurovascular relationships at the REZ of the nerve [17]. Recently, due to the advancement of high-resolution MRI (HR-MRI), three-dimensional time-of-fight (3D-TOF) and three-dimensional constructive interference in steady state (3D-CISS) sequences were used in patients with HFS. Such a procedure has emerged to be a useful tool for assessing neurovascular compression [5, 18]. In the present study, we included 95 patients with HFS at the Aerospace Clinical Medical College of Peking University during September 2012 to July 2013. The aim of this study was to analyze the accuracy and preoperative diagnostic value of HR-MRI techniques by using 3D-TOF and 3D-CISS sequence solely or in combination for the detection of the relationship between the facial nerve and adjacent vessels in patients with HFS. Patients and Methods Patients The study was approved by the ethics committee of Aerospace Center Hospital, China, and written informed consent was obtained from all subjects. A total of 95 patients (20 male and 75 female, with a mean age of 49.2 years) with primary HFS were included during September 2012 to July 2013. The diagnosis of HFS was based on the clinical history and physical examination and then was confirmed with an abnormal motor response on electromyography. The duration of disease varied from 6 months to 23 years, with a mean value of 6.7 years. All the cases were of HFS on one side (left side: 51 cases, right side: 44 cases). For the purposes of this study, we excluded all patients with a confounding diagnosis (such as secondary HFS, tumor, disseminated sclerosis, and facial paralysis caused by trauma and cold). Patients who received botulinum toxin therapy within half a year were also excluded. In addition, patients with poor physical condition who could not be surgically treated were also excluded. MR Examination All patients underwent MR preoperatively using a 1.5-T MR imager (Siemens Magnetom, Erlangen, Germany), including three following sequences: high-spatial-resolution 3D-TOF and 3D-CISS-MRA.
13
J.-m. Jia et al.
3D-TOF MRA (repetition time (TR) = 30 ms, echo time (TE) = 5.0 ms, flip angle = 20°, NEX = 1, section thickness = 0.9 mm, gap = 0, field of view (FOV) = 240 mm, and matrix size = 512 × 512) and 3D-CISS sequences (TR = 10.9 ms, TE = 5.4 ms, flip angle = 70°, NEX = 1, section thickness = 0.8 mm, gap = 0, FOV = 230 mm, and matrix size = 173 × 256) were used in the present imaging study. According to the midsagittal plane of brain stem, transverse scanning of the whole brain stem was performed. Based on the 3D-TOF axial images, multiplanar reformatting in the oblique sagittal (left and right) and coronal planes, parallel and perpendicular to the subarachnoid cisterns, was performed on the workstation. The maximum intensity projection (MIP) and volume rendering were used to reconstruct blood vessels so that the offending vessels of the facial nerve can be displayed. The examination was regarded as positive if the definite neurovascular conflict between the facial nerve and a vessel can be seen by 3D-TOF and/or 3D-CISS sequences (Figs. 1 and 2). Imaging Analysis All the images were evaluated respectively by two experienced radiologists, who were blinded to any knowledge of which side the spasm was located at and the operative findings. Decision was obtained by consensus. We categorized the relationship between REZ and neighboring vessels into three types based on the axial, coronal, and oblique sagittal images: (1) compression, visualized by axial plus coronal and/or oblique sagittal images, defined as the presence of obvious vascular indentation at the REZ of the facial nerve by either vessels/facial nerve displacement or the course of anomalous vessels; (2) contacts, considered to be present if the minimum distance between vessel and the REZ was not more than the maximum diameter of the vessel by axial plus coronal and/or oblique sagittal images; (3) neighboring or distant, considered to be present if the minimum distance between the vessel and the REZ was larger than the maximum diameter of vessels by axial plus coronal and oblique sagittal images. REZ was defined as the point where the facial nerve exited from the pons. Compression and/or contacts between REZ and vessels were defined as positive, whereas neighboring or distant was considered to be negative. Surgical Technique All the procedures were performed by one experienced neurosurgeon via a lateral retrosigmoid suboccipital approach, which has been previously described in the literature [19]. Necessary judgments were made for the identification of offending vessels according to the distribution as well as arrangement.
Preoperative Evaluation of Patients with Hemifacial Spasm by Three-dimensional Time-of-Flight (3D-TOF)
3
Table 1 Comparison of results from preoperative imaging assessment with intraoperative findings (cases, n = 95) Methods Sensitivity Specificity Positive predictive Negative predictive Accuracy Positive rates value value 3D-TOF 88/94 (93.61 %) 1/1 (100 %) 88/88 (100 %) 1/7 (14.29 %) 89/95 (93.68 %)* 92.63 %* 3D-CISS 81/94 (86.17 %) 1/1 (100 %) 81/81 (100 %) 1/14 (7.14 %) 82/95 (86.32 %)** 85.26 %** 3D-TOF + CISS 94/94 (100 %) 1/1 (100 %) 94/94 (100 %) 1/1 (100 %) 95/95 (100 %) 98.95 % All results were obtained by comparison with 3D-TOF + 3D-CISS 3D-TOF three-dimensional time-of-fight, 3D-CISS three-dimensional constructive interference in steady state *p
