Effects of the MR-DTI Characteristics of the Trigeminal Ganglion Target on Radiofrequency Treatment in Patients with Trigeminal Neuralgia: A Retrospective Observational Clinical Study

Background In the percutaneous treatment of trigeminal neuralgia (TN), the difficulty in accessing the foramen ovale (FO) has been widely recognized. However, the most efficient percutaneous treatment target is the so-called trigeminal ganglion target (TGT). We propose that the TGT in a puncture can be identified by magnetic resonance diffusion tensor imaging (MR-DTI). Objectives To observe the effect of the characteristics of the TGT as detected by MR-DTI on percutaneous stereotactic radiofrequency rhizotomy (PSR) in TN patients. Methods In our observational study, we preoperatively performed MR-DTI and/or 3D-CT for 48 TN patients, analyzed the characteristics of the TGT and/or FO, and designed appropriate surgical schemes for producing an accurate PSR trajectory according to these characteristics. The position and size of the TGT aided in adjusting the puncture angle and guiding the approach. Then, we successfully performed a customized PSR guided by the characteristics of the FO or TGT. During the postoperative and follow-up periods, we assessed the effect of treatment with pain scores and MR-DTI findings. Results The characteristics of the TGT vary from patient to patient. We performed PSR with a single puncture guided by MR-DTI and 3D-CT in 16 patients, and only one patient required three punctures. All three of these punctures reached the FO target, as confirmed by intraoperative C-arm X-ray. We finally reached the TGT successfully after 2 additional attempts, confirming that the probe reached the TGT that accurately covered the pain territory with an electrophysiology test. The characteristics of the TGT were negatively correlated with the number of PSR punctures. Fewer complications occurred for PSRs guided by the TGT than for PSRs guided by the FO. Conclusions The characteristics of the TGT are correlated with the number of punctures in the PSR. The application of MR-DTI for detecting the size of the TGT is an important step in predicting the difficulty of puncture. The PSR approach can be guided by the TGT and FO for TN patients who present with multiple adverse factors and thus may be beneficial in reducing the number of complications.

Te single-puncture rate in radiofrequency treatment has attracted great attention. Te PSR procedure has two key steps, including puncture of the FO and TG [7]. In certain circumstances, the FO is difcult to access in certain circumstances, such as when the trajectory is obstructed by the maxilla or mandible, the distance between the mandible and tooth ridge is too narrow, the FO is covered by the peripheral bones, or the FO is small; in these circumstances, the frst puncture attempt is often unsuccessful, and multiple attempts are needed [8,9], potentially causing unnecessary damage and complications. However, inaccurate localization is the primary reason for puncture failure and is considered a signifcant cause of pain recurrence and complications. Accordingly, accurate puncture is of the utmost importance, as it may decrease the incidence of complications and improve the efects of treatment. Te difculty in accessing the FO can be analyzed and lessened by methods such as three-dimensional computed tomography (3D-CT) imaging.
Te efcient therapeutic target of PSR is a part of the TG, known as the trigeminal ganglion target (TGT). Conventional imaging is unable to identify and locate this part of the TG, which controls the afected CNV branches [4,10]. It has been reported that magnetic resonance difusion tensor imaging (MR-DTI) can detect CNV lesions in TN patients [11][12][13][14][15]. Te success of treatment with radiofrequency is directly related to the area afected by the lesion [6]. Confrmation of probe insertion into the TGT is performed with a combination of electrophysiology tests and MR-DTI. Accordingly, it is vitally important that the location of the lesion in the TG be determined by MR-DTI to ensure adequate puncture for PSR [15]. Consequently, the characteristics of the TGT could serve as a novel indicator of puncture difculty in PSR, as accurate puncture and treatment are important to ensuring the efects of PSR [7].
Here, we investigated the characteristics of the TGT and the number of punctures needed to perform PSR and observed the efect of novel customized PSR approaches guided by combining MR-DTI and 3D-CT skull base reconstruction in TN patients. To our knowledge, this is the frst study on the efect of the characteristics of the TGT on the puncture difculty and treatment efcacy of PSR.

Participants.
From May 2012 to March 2022, 48 patients with TN underwent PSR treatments at the Department of Neurosurgery. In our observational study, 17 patients who underwent PSR guided by both the TGTand FO and 31 patients who underwent PSR guided by only the FO were included. We compared the puncturing difculty and treatment efect of PSR before and after selection of the TGT in guiding the PSR approach. Tis study was approved by the Ethics Committee of the China-Japan Union Hospital of Jilin University, and patients consented to the publication of medical images.

Image Acquisition and Postprocessing.
Te patients underwent preoperative MR-DTI and/or 3D-CT scans (DTI: spin-echo echo-planar imaging sequence, 72 directions; inplane voxel size, 2 × 2; slice thickness, 2 mm; echo time, 95 ms; repetition time, 4100 ms; 3.0 Tesla Siemens MR scanner). Te MR-DTI data were postprocessed in the 3D Slicer V4.11 software package utilities (https://www.slicer. org/). We applied tractography to delineate the bilateral CNVs on DTI and calculated the scalar maps (fractional anisotropy (FA), mean difusivity (MD), radial difusivity (RD), and axial difusivity (AD)) ( Figure 1). Ten, we extracted the difusivity metrics from the 8 manually delineated segments of the CNV tracts according to the slice thickness of the coronal images ( Figure 2). According to previous literature, the FA values of the afected side were lower than those of the contralateral side [12][13][14]. We defned the signifcantly low FA segment of the TG as the therapeutic target of PSR, in other words, the TGT. Te position and size of the TGT are marked on a coronal image of the FA map in Figure 1.
Te 3D reconstruction of the skull base on 3D-CT was used to identify the characteristics of the FO (GE AW 4.6 Image Postprocessing Workstation, GE Healthcare). For an illustrative case, 3D-CT skull base reconstruction images and a 3D-printed model of the patient demonstrated the difculty of puncture: obfuscation of the FO by the pterygoid process ridge (PPR) and the lateral pterygoid plate (LPP) and the small size of the FO (Figure 3).
Morphological MRI at the cerebellopontine angle was used to identify intracranial focal lesions and compression of the CNV root on the afected side by the adjacent vessel [16].

Treatment.
Te operations were performed using Komai's CT stereotactic frame (Mizuho Medical Innovation, Tokyo, Japan). We designed accurate trajectories for PSR according to the characteristics of the TGT and/or FO on the MR-DTI images; in particular, the position and size of the TGT and TG aided in adjusting the arc angle of the probe. Ten, we performed PSR guided by the TGT for 17 TN patients and by the FO for 31 TN patients [8,9,17]. During the operation, we performed radiofrequency treatment and electrophysiology tests. Te voltage of the electrophysiology sensory test ranged from 0.15 to 0.30 V, and the voltage of the motor test ranged from 0.9 to 1.2 V. Next, we performed 120 seconds of continuous radiofrequency ablation with a temperature ranging from 65 to 72°C.

Evaluation of the Treatment Efect of PSR.
Te efect of PSR treatment was evaluated during the postoperative and follow-up periods. Te treatment outcome of the TN patients was assessed by a visual analog scale (VAS), on the MR-DTI, and in terms of complications.

Statistical
Analysis. Summary statistics were computed for all data using SPSS v26.0 (IBM, USA) in this study. Te data were compared between the TGT-and FO-guided PSR groups using the independent-samplet test. Te correlation between the TGT characteristics and the number of PSR punctures was determined with a 2-tailed Pearson's correlation (P < 0.05). In all statistical tests, P < 0.05 was considered statistically signifcant.

Characteristics of TN Patients Who Underwent PSR.
Te baseline characteristics of the 48 TN patients are summarized in Table 1. Compared with that of the contralateral CNV, the FA of the segment of the TG on the afected side was interrupted and therefore lower, identifying an abnormal segment ( Figure 4).

Number of PSR Punctures for the TN Patients.
Among the patients who underwent PSR guided by the TGT, only one patient (the illustrative case) required three   punctures to successfully target the TGT; the other patients only required a single puncture each. Te three punctures in the illustrative case all reached the FO target, as confrmed by intraoperative C-arm X-ray ( Figure 5). Te frst puncture to the TG controlled the V2 region, as confrmed by an electrophysiology test. Ten, we reached the TGT successfully after 2 adjustments and confrmed that the probe covered the pain territory (V1) accurately with another electrophysiology test. Among patients who underwent PSR guided by the FO, four patients required multiple punctures, including two patients who underwent two punctures and another two patients who required three punctures.

Characteristics of the TGTs in TN Patients.
Te size and position of the TGTs within the TG varied with the coronal view of the FA map ( Figure 6; Table 2). Te position of TGTs in TG also difered. In 17 patients who underwent PSR guided by the TGT, 3 TGTs (17.6%, 3/17) were located in the center of the TG, and the other 14 TGTs (82.4%, 14/17) were found in diferent positions, including 1 at the upper right, 2 at the right edge, 5 at the lower right, 1 at the upper left, 4 at the lower left, and 1 at the lower edge.
In the illustrative case, the size of the TGT on the coronal view was the smallest at 2.5 mm 2 , and the short diameter of the TGT was 1.4 mm. Te TGT was located laterally and superiorly in the TG. Te patient was a 74year-old male who presented with right-sided V2 and V1 TN. Te pain mainly spanned from the upper lip and near the nose.

Correlation of the Characteristics of the TGTs with the Number of PSR Punctures.
In analyzing the data of patients who underwent PSR guided by the TGT (Table 3), the size of the TGT was signifcantly negatively correlated with the number of PSR punctures (r � −0.6, P < 0.05), and the short diameter of the TGT was signifcantly negatively correlated with the number of PSR punctures (r � −0.6, P < 0.05).

Te Efect of PSR on TN Patients.
In patients who underwent PSR guided by the TGT, the VAS score decreased from 10 to 1 in 15 patients (88.2%) at 1 day postoperatively and to 4 in the other 2 patients (11.8%) ( Table 4). Te VAS score decreased to 1 for all patients at 10 days postoperatively. Tere were no common complications in any of the 17 patients, and only one patient (5.9%) presented with a rare complication of apnea.
In the illustrative case, there were no postoperative complications. On the frst postoperative day, the VAS score in the V2 control area was 1, while that in the V1 control area was 4. At 2 days postoperatively, the VAS score had decreased to 1, and the FA of the TGT was increased (Figure 2). At a 3-monthfollow-up, the VAS score of the patient was 1, and the medication was suspended. During the 4-month follow-up, the patient underwent an MR-DTI scan, and the FA of the TGT was greater than the preoperative value ( Figure 2).
In patients who underwent PSR guided by the FO, the VAS score decreased from 10 to below 3 in 28 patients (90.3%), to 4 in one patient (3.2%), and to 7 in two patients (6.5%) 1 day after the operation (Table 4). Two patients (6.5%) still had a VAS score greater than 3 at 10 days postoperatively. Seven patients (22.6%) experienced facial numbness after treatment, but this side efect subsided within six months after surgery. One patient (3.2%) experienced a hematoma of the facial soft tissue, and one patient (3.2%) experienced residual facial pain. Four patients (12.9%) experienced pain recurrence 1 to 2 years after surgery.
Tere was no signifcant diference in the VAS score between the two TN patient groups (P > 0.05). However, the rate of complications in the PSR guided by the FO group was higher than that in the PSR guided by the TGT group.

Discussion
We investigated the efect of the characteristics of the TGT on radiofrequency treatment in TN patients in our study. Te results suggest that these characteristics are correlated with the number of punctures required to perform PSR and are benefcial for guiding radiofrequency treatment. Te characteristics of the TGT may be a valuable predictive indicator of the difculty of puncture in PSR. In particular, patients with small TGTs will likely need multiple punctures. Te PSR approach has two key steps for TN treatment: localization of both the FO and TG.
First, the puncture probe passes through the FO; this structure is difcult to access successfully with a single puncture. Guo et al. reported that the puncture failure rates range from 10% to 20% for experienced neurosurgeons [8]. It was previously reported that 2-4% of the FOs of TN patients are difcult to access, and multiple unsuccessful punctures may induce complications [17,18]. Many recent studies have reported on the difculty in puncturing the FO for TN treatment, and thus this difculty has become widely acknowledged. A variety of factors infuencing this difculty have been reported, including (1) whether the trajectory is obstructed by the maxilla or mandible or if the distance between the mandible and tooth ridge is too narrow; (2) whether the FO is covered by the LPP and/or the PPR; and (3) the size of the FO, which, due to anatomic variability, can be small [9,19]. Te problem of the difcult puncture of the FO during the operations is a subject of seemingly constant investigation. To avoid the infuence of soft tissue in easily displacing the probe in the Härtel route, Matthew et al. reported that bony landmarks should be identifed during puncture [20]. Bony landmarks are easy to identify and easily marked on the surface of the face and visualized in Xray projections. Various techniques and applications, such as 3D-CT skull base reconstruction, stereotactic frames, or navigation, have been implemented to resolve these issues [1,7]. Te aforementioned studies suggested that accurate puncture allows for a more efcient operation and a reduced need for repeat punctures. Te operative plan is tailored so that we can determine the design of the trajectory, identify the best operative position, and predict intraoperative difculties.
Second, it is well known that the FO is just the frst target of PSR, as the probe also needs to reach the TG after passing through the FO. Generally, for patients for whom the V1 or V2/ V3 branch is afected, the inner third or half of the transverse diameter of the FO is selected as the puncture target [9]. However, the TGT is the ideal but invisible puncture target in PSR. Specifc parts of the TG control the V1, V2, and V3 branches of CNV, but the location of the actual therapeutic TGT has not been identifed previously. In our novel model, we used MR-DTI to identify the specifc part of the TG responsible for the afected branches and named it the TGT. Ten, we designed the operation scheme with guidance by the TGT and FO, which assisted in accurately performing PSR at the TGT. Especially in the illustrative case, the area of the TGT was 2.4 mm 2 , and its shorter diameter was only 1.4 mm. Nonetheless, the diameter of the puncture probe is 0.9 mm, which makes it challenging to successfully puncture the TGT. Terefore, the characteristics of the TGTare strongly correlated with the number of punctures in the PSR; that is, a small TGT size may be a valuable predictive indicator for the need for multiple punctures. Te success of the TGTpuncture is important to ensure the efect of PSR. Te TG is located at the anterior, inferior, and lateral sides of Meckel's cave in the skull base, and the ganglion is small due to its concave shape [21][22][23]. Te difculty in accessing TGT was assessed by the combination of 3D-CT and MR-DTI. When the probe has passed through the FO, it can be inserted into the part of the TG that projects to the afected branches. Te probe location can infuence the efcacy and complications of the therapy [15,23]. In practice, the parts of the TG that project to the afected branches usually need to be validated by an electrophysiology test [1]. However, a high rate of complications due to radiofrequency treatment, such as facial numbness, facial paresthesia, masseter weakness, decreased corneal refex, and keratitis, has also been observed [1,24]. In our study, the efect of treatment between the two groups was not signifcantly diferent, but postoperative complications were less common in the group that underwent PSR guided by the TGT. Te model primarily highlights the signifcance of puncturing the TGT during PSR for achieving fewer complications.
However, we used MR-DTI to detect the TGT for predicting the difculty of puncture in PSR, and the novel model, combining 3D-CT reconstruction and MR-DTI, improved the accuracy, safety, and efciency of PSR for TN patients.
Limitations: at present, we are not yet able to precisely fuse CT and MRI, making it impossible to measure the precise distance and angle of the TGT puncture.   Additionally, this study is only an exploratory study with a small sample size, and further study with a larger sample size is needed. However, this study also does not involve an overall sampling problem and lacks underlying data, so no sample size calculation can be performed. In our observational study, the sample size was the number of patients who had FO and/or TGT-guided PSR procedures applied.

Conclusions
We found that the characteristics of the TGT were correlated with the number of punctures in the PSR. Te application of MR-DTI for detecting the size of the TGT is an important step in predicting the difculty of puncture. Te PSR approach can be guided by TGT and FO for TN patients who have multiple factors afecting the difculty of puncture, which may be benefcial to producing a smaller number of complications.

Data Availability
Te data used to support the fndings of this study are included within the supplementary information fles.

Conflicts of Interest
Te authors declare that they have no conficts of interest.