ORIGINAL ARTICLE
Application of Antidromic Conduction Monitoring in Ganglion Radiofrequency Thermocoagulation for Locating Trigeminal Branches in Trigeminal Neuralgia Xiuhua Li, MD; Jianning Yue, MD; Liqiang Yang, MD; Huijie Yang, MD; Shuyue Zheng, MD; Liangliang He, MD; Jiaxiang Ni, MD Department of Pain Management, Xuanwu Hospital of Capital Medical University, Beijing, China
& Abstract Objective: The aim of this study was to investigate whether antidromic conduction monitoring (ACM) can be utilized to map the trigeminal system under sedation as a potential substitute for subjective paresthesia description (SPD) during percutaneous ganglion radiofrequency thermocoagulation (PGRT). Methods: Eighty-two patients with 152 pain divisions of trigeminal neuralgia (TN) were treated by computed tomography (CT)-guided PGRT. After the puncture needle entered the foramen ovale (FO), sensory and motor stimulation were applied to locate the pain division. And the corresponding voltage values were recorded by patients’ SPD. In the following, the proper location was certified by ACM. The corresponding earliest waves and voltage values in the identified trigeminal branch were also recorded to outline a comparison between two methods. Results: The correlation of ACM and patients’ SPD with voltage at ≤ 0.5 V was statistically significant (P < 0.05,
r = 0.159; Spearman’s rank correlation analysis). Although ACM and SPD showed weak correlation, as their interclass correlation coefficient was significant (F = 1.868, P < 0.01) with coefficient of internal consistency. Moreover, the two methods had consistency. Kruskal–Wallis test showed that ophthalmic (V1), maxillary (V2), and mandibular (V3) divisions had significant differences for test sensitivity (H = 15.945, P < 0.01). For comparison of sensitivities with ACM, V3 was most sensitive followed by V2 and then V1. Conclusion: ACM could potentially substitute for SPD of the paresthesias intra-operatively, enabling greater specificity and eliminating the need to interrupt the administration of anesthetic. These improvements would increase patient satisfaction and practitioner efficiency and accuracy. & Key Words: trigeminal neuralgia, radiofrequency thermocoagulation, antidromic conduction, subjective paresthesia, correlation
INTRODUCTION Address correspondence and reprint request to: Jiaxiang Ni, MD, Department of pain management, Xuanwu Hospital of Capital Medical University, No 45 Changchun Street, Xicheng Zone, Beijing 100053, China. E-mail: [email protected]. Submitted: August 8, 2014; Revised November 9, 2014; Revision accepted: November 28, 2014 DOI. 10.1111/papr.12286
© 2015 World Institute of Pain, 1530-7085/16/$15.00 Pain Practice, Volume 16, Issue 3, 2016 305–310
Trigeminal neuralgia (TN) is craniofacial pain involving 1 or more trigeminal nerve branches. It is characterized by mostly shock-like pain.1 Various therapies can be used for TN: medical treatment; microvascular decompression (MVD); gamma knife radiosurgery; and percutaneous treatments including balloon compression; glycerol rhizotomy; and radiofrequency (RF) thermocoagulation.2–4 Different operators have different options
306
LI ET AL.
for TN according to their experience and skills. Generally, RF thermocoagulation is quite easy and safe to perform under computed tomography (CT) guidance with no mortality and low morbidity, especially in elderly patients.5–7 When the puncture needle enters the foramen ovale (FO), sensory (50 Hz) and motor (2 Hz) test stimulation are used to elicit paresthesia in the affected trigeminal distribution area. In the process of the electrical stimulation, patients must keep conscious and cooperative for the subjective paresthesia description (SPD).8 However, much time is wasted when administering intermittent anesthesia, and the repeated anesthesia induction may make the patients’ description somewhat unreliable, not to mention the added procedural pain and discomfort. Neural electrophysiological procedures have emerged to solve drawbacks. It has been suggested that the use of trigeminal evoked potentials (TEPs), which usually entails orthodromic conduction, can be used to detect the needle position. This procedure has produced satisfactory results with TN in the maxillary (V2) division, but it is tedious and time-consuming.9 Jones and Karol have observed that antidromic stimulation could generate wider amplitude and longer latency stage of the negative phase, and the waves were easy to be observed and obtained.10,11 Bendersky et al.12 showed that antidromic nerve conduction monitoring (ACM) was useful for determining the exact innervation of trigeminal nerve divisions in anesthetized patients, making percutaneous ganglion radiofrequency thermocoagulation (PGRT) safer and more comfortable. However, it has not been reported whether ACM can be utilized to map the trigeminal system under sedation as a substitute for SPD during PGRT. It is unknown if different TN divisions can achieve the same sensitivity with ACM. Therefore, this study was conducted to investigate the correlation and consistency of the ACM and the SPD methods, and measure the sensitivity/selectivity of both techniques.
METHODS The institutional research ethics committee approved this study. A series of 82 patients (35 men, 47 women) with classic trigeminal neuralgia were recruited and gave informed consent. They were treated with CT-guided ganglion RF thermocoagulation. In each case, the procedure was performed in a sterile CT examination room with intermittent anesthesia. Each patient was placed in a supine position with their head hanging over the bed. The puncture location at the FO was deter-
mined on the CT scan, and the corresponding percutaneous point was marked. Following sterilization and local anesthesia with 1% lidocaine, a 22 g RF needle with a 5-mm working zone (straight; Cosman, Burlington, MA, USA) was inserted through the marked skin point at the FO under CT guidance. Electrical stimulations were performed to confirm and re-adjust the needle position. After the needle position was identified, anesthesia with propofol and an analgesic was administered intravenously. Intubation was not required. The ganglion was coagulated with consistent radiofrequency at 70 to 75°C for 120 second twice or more depending on the individual’s diversity and the physician’s experience. A multichannel electromyography/evoked potentials (EMG/EP) machine (DAVINCI, Mogliano Veneto, Italy) was used for neurophysiological monitoring. After cleansing the facial skin with 95% alcohol, three pairs of surface electrodes with conductive paste to eliminate skin electrical impedance as much as possible were placed, respectively, on the supraorbital foramen, infraorbital foramen, and mental foramen (negative electrode on the foramen and positive electrode < 2 cm from the negative electrode, ground wire on the ipsilateral earlobe). In previous studies, needle electrodes were also used for recording, but some injures to different extent was induced, as reported by Leandri and Gottlieb9 and Bendersky et al.12 In the present research, we used three pairs of surface electrodes without harming the patients and with their high degree of satisfaction.13 Three channels were connected with the three pairs of electrodes: channel 1—V1, channel 2—V2, and channel 3—V3. Sensory (50 Hz) and motor (2 Hz) test stimulation were applied to adjust the needle position until paresthesia of the corresponding trigeminal branches was elicited, and each minimum stimulation voltage value was recorded, respectively. A proper needle position was defined as a stimulation voltage value of < 0.5 V according to the patient’s SPD. After verification, the electrical stimulation was returned to 0 V again to certify the proper location of needle tip by ACM and began to increase it slowly until the first emerging waveform of the EMG machine in the “needle EMG” state was observed in the corresponding channel. The minimum stimulation voltage was, respectively, recorded at the same time. If a description of paresthesia from the patient for a particular pain division could not be achieved, the apparent, prominent waveforms recorded by ACM were used for identifying the trigeminal distribution.
Neurophysiological Monitoring for TN 307
Statistical analysis was performed using the Statistical Package for the Social Sciences version 19.0 (SPSS, Chicago, IL, USA). The frequency was set for the correlation analysis with Spearman’s ranked data between ACM and SPD. The intraclass correlation coefficient was used to evaluate the coherence between two methods. The Kruskal–Wallis test evaluated the sensitivity among three divisions with ACM. P < 0.05 was considered statistically significant.
RESULTS Eighty-two patients diagnosed with typical TN in 152 pain divisions, including 25 patients with one division, 44 patients with two divisions, and 13 patients with three divisions recruited in the study. Ages ranged from 30 to 84 years. Magnetic resonance imaging results were normal in all patients. The patients’ characteristics were described in detail in Table 1. After verification by patients’ SPD, the waves that achieved the highest amplitude in the corresponding division were recorded through ACM method. When the needle tip was certified in V1, the highest amplitude of waves was recorded in channel 1, and the other channels
demonstrated smaller or no waves (Figure 1A). It was also applied to V2 (Figure 1B) and V3 (Figure 1C). During the process of electrical stimulation, 146 pain divisions were positively detected, six pain divisions (four in V1, one in V2, and one in V3) were negatively detected according to patients’ SPDs, 148 pain divisions were positively detected (≤ 0.5 V), and four pain divisions (detected only in V1) were considered negative (> 0.5 V) by ACM. Although the correlation coefficient was weak, the correlation comparison of ACM and SPD in different ranges of voltage values (≤ 0.5 V) had statistical significance (r = 0.159, P < 0.05). Moreover, the two detection methods had significant coherence (F = 1.868, P < 0.01), indicating that ACM could be potentially used to displace the SPD method (Table 2). Furthermore, the voltage values’ distribution of ACM was significantly lower than that obtained by patients’ SPDs (P < 0.05) (Figure 2). Figure 3 showed that the comparison of the detected voltage values under positive monitoring with ACM had statistical differences for V1, V2, and V3. Among them, the detection for V3 achieved the highest sensitivity (voltage value 0.07 0.04), followed by V2 (voltage value 0.12 0.09) and then V1 (voltage value 0.19 0.16) (H = 15.945, P < 0.01).
DISCUSSION
Table 1. Trigeminal Neuralgia Patients Characteristics Feature Number of patients Age (year), mean age SD Gender (female/male), n (%) Duration of pain symptom SD, (m) Affected side, n (%) Left Right Division of the trigeminal nerve, n (%) V2 V3 V1 + V2 V2 + V3 V1 + V2 + V3
82 30~84 (62.3 11.7) 47/35 (57.3/42.7) 52.6 83.7 35 (42.7) 47 (57.3) 12 13 14 30 13
(14.6) (15.9) (17.0) (36.6) (15.9)
V1 = Ophthalmic division; V2 = maxillary division; V3 = mandibular division.
A
B
Eighty-two patients with TN in 152 trigeminal branches were studied. There were three patients experiencing moderate pain after PGRT, however, a local block improved their pain after procedure. The complete pain relief rate was 96%, which coincided with the present literatures about PGRT applied under imaging guidance.14,15 None of the patients had apparent or severe intracranial or extracranial structural injuries.16,17 Generally, the conventional method for locating trigeminal nerve branches is the patient’s subjective description of his or her paresthesia.11 It has been C
Figure 1. (A) The apparent waveforms in channel 1 showed the needle location in V1 from sensory stimulation with 50 Hz 0.1 millisecond recorded in ganglion by the EMG/EP machine; (B) showed the needle location in V2 and (C) showed the location in V3. All of them were shown with the correct SPD by the patients.
308
LI ET AL.
Table 2. Voltage Value Detection with ACM and SPD in 82 Cases with 152 Different Pain Divisions V1 (27)
V2 (69)
V3 (56)
Trigeminal pain divisions (n) Positive (0 to 0.5 V)/negative (> 0.5 V)
Positive
Negative
Positive
Negative
Positive
Negative
ACM SPD
23 23
4 4
69 68
0 1
56 55
0 1
Figure 2. Showed the different distribution of ACM and SPD in different voltage value.
Figure 3. Showed the different voltage value of ACM in different branches.
reported that antidromic stimulation could generate wider amplitude and longer latency stage of the negative phase, and the waves were easily observed and obtained.10,11 ACM was also useful for determining the exact innervation of trigeminal nerve divisions in anesthetized patients, making PGRT safer and more comfortable.12 In this study, ACM was also used to identify the division location with routine sensory stimulation during CT-guided PGRT. Combined with the collective original data for the analysis, the voltage values were divided into three sectors by ACM: 0 to
0.05 V (78 pain divisions detected at 0.05 V); 0.05 to 0.15 V (53 pain divisions); and 0.15 to 0.5 V (17 pain divisions). Three sectors for the patients’ SPDs: 0 to 0.2 V (46 pain divisions), 0.2 to 0.3 V (76 pain divisions), and 0.3 to 0.5 V (24 pain divisions). Correlation and coherence were analyzed, respectively, between the two detection methods. As the results showed, although the correlation coefficient was weak, the correlation comparison of ACM and SPD in different ranges of voltage values (≤ 0.5 V) had statistical significance (r = 0.159, P < 0.05). Moreover, the two methods had significant coherence (F = 1.868, P < 0.01), indicating that ACM detection could potentially be used instead of the conventional SPD method. Furthermore, the voltage threshold of ACM was obviously lower than that of patients’ SPD, decreasing procedural pain and patient’s discomfort. In our study, the successful detection rate for ACM at 0 to 0.3 V was 74% in V1, 97% in V2, and 100% in V3. At 0 to 0.5 V, it was 85% in V1 and 100% in V2. ACM detection in V3 showed the most sensitivity, whereas the worst detection sensitivity was in V1. We presumed that it related to the anatomical positions of nerve cell bodies in the ganglion. The trigeminal ganglion has a somatotopic arrangement in which V1 is the most craniomedial and V3 the most lateral. Namely, V1 is in the medial third of the FO, V2 in the middle of the FO, and V3 in the lateral third of the FO.15,18 According to anatomical research, V1 is the most deeply located, and operators are extremely cautious when approaching V1 to avoid severe complications, such as cranial nerve injury and intracranial hemorrhage.19,20 Puncture at V1 is the most difficult to deal with and reach. Even with optimal trajectories, V1 continues to be the least selected for targeting (73.3%), topped by V2 (84.6%) and V3 (100%).20 Puncture is relatively easy through the FO into V3, where detection for locating achieves the highest sensitivity. Detection was negative at high voltage (> 0.5 V) in our study, with no waves emerging and no paresthesia reported by the patients. Negative detection in V1 was found in four pain divisions by ACM (waves emerged in V1 at 0.7 to 0.9 V) and SPD (waves emerged in V1 at 0.8
Neurophysiological Monitoring for TN 309
to 1.2 V). For the negative detection in V1, which detected at high voltage, CT guidance can be used to identify the proper needle position in the FO. Considering the mentioned risk,16,20–22 the procedure was undergone with PGRT and the four patients achieved pain relief. Some case reports showed that trigeminal neuralgia was due to vertebrobasilar dolichoectasia and distortion affecting the trigeminal nerve.23,24 In one case, it was uncertain whether or not the negative detection was related to the vascular deformity. The active tip of the radiofrequency needle used in this study was 5 mm, the thermal lesion of which was larger than the 2-mm needle.7 This might explain why the patient achieved pain relief after procedure with negative detection. We also suggested that the large range of 0 to 1.2 V could be used to detect areas in V1, combined with sensory stimulation of 50 Hz for 0.1 millisecond.21 There was no negative detection in V2 or V3 with ACM. Negative detection by SPD was the same in V2 and V3, with one pain division, but prominent waveforms were identified in V2 and V3 by ACM. Based on these findings, we suggested that ACM was more reliable than SPD for identifying the trigeminal nerve divisions, especially V2 and V3. ACM could potentially displace the SPDs in V2 and V3, and in V1 with high voltage (> 0.5 V). In total, 39 patients underwent local nerve block, acupuncture, and percutaneous intervention therapy before this study. Examinations before PGRT showed there were various injuries to peripheral nerves.25 The amplitudes varied among patients at the same level of stimulation (eg, 0.3 V); however, it showed no effect on the division locations. Therefore, we surmised that patients had different nerve conduction, resulting in individual differences. Future research should be taken to explore the lesion level of RF thermocoagulation in individuals using ACM in cases of TN recurrence at the original location.26,27 Although ACM for locating V1 in some cases required a higher voltage for sensory stimulation, this study showed that there was correlation and coherence between ACM and SPD regarding identification of the trigeminal nerve branches’ locations during PGRT. ACM was even more reliable than SPDs in V2 and V3, and most sensitive in V3. Despite this, some limitations of the study required consideration. In particular, the cohort size was not a large number (82 subjects); however, we can draw a conclusion that antidromic conduction monitoring could potentially substitute for the patient’s description of the paresthesias intra-
operatively enabling greater specificity, eliminating the need to interrupt the administration of anesthetic. These improvements would improve patient satisfaction and practitioner efficiency and accuracy.
REFERENCES 1. Headache Classification Committee of the International Headache Society (IHS). The international classification of headache disorders, 3rd edition (beta version). Cephalalgia. 2013;33:629–808. 2. Cheng JS, Lim DA, Chang EF, Barbaro NM. A review of percutaneous treatment for trigeminal neuralgia. Neurosurgery. 2014;10:25–33. 3. Parmar M, Sharma N, Modqill V, Naidu P. Comparative evaluation of surgical procedures for trigeminal neuralgia. J Maxillofac Oral Surg. 2013;12:400–409. 4. Koopman JS, de Vries LM, Dieleman JP, Huygen FJ, Stricker BH, Sturkenboom MC. A nationwide study of three invasive treatments for trigeminal neuralgia. Pain. 2011;152:507–513. 5. Cheng JS, Sanchez-Mejia RO, Limbo M, Ward MM, Barbaro NM. Management of medically refractory trigeminal neuralgia in patients with multiple sclerosis. Neurosurg Focus. 2005;18:e13. 6. Berk C, Constantoyannis C, Honey CR. The treatment of trigeminal neuralgia in patients with multiple sclerosis using percutaneous radiofrequency rhizotomy. Can J Neurol Sci. 2003;30:220–223. 7. van Kleef M, van Genderen WE, Narouze S, et al. Trigeminal neuralgia. Pain Pract. 2009;9:252–259. 8. Hart MG, Nowell M, Coakham HB. Radiofrequency thermocoagulation for trigeminal neuralgia without intra-operative patient waking. Br J Neurosurg. 2012;26: 392–396. 9. Leandri M, Gottlieb A. Trigeminal evoked potentialmonitored thermorhizotomy: a novel approach for relief of trigeminal pain. J Neurosurg. 1996;84:929–939. 10. Jones DL, Wolford LM. Intraoperative recording of trigeminal evoked potentials during orthognathic surgery. Int J Adult Orthodon Orthognath Surg. 1990;5:167–174. 11. Karol EA, Sanz OP, Rey RD. Sensory and motor trigeminal evoked potentials to localize the position of trigeminal electrodes. Acta Neurochir (Wien). 1991;108:110–115. 12. Bendersky M, Hem S, Landriel F, et al. Identifying the trigeminal nerve branches for transovale radiofrequency thermolesion: “no pain, no stress”. Neurosurgery. 2012;70:259– 263. 13. Obermann M, Yoon MS, Ese D. Impaired trigeminal nociceptive processing in patients with trigeminal neuralgia. Neurology. 2007;69:835–841. 14. Kanpolat Y, Savas A, Bekar A, Berk C. Percutaneous controlled radiofrequency trigeminal rhizotomy for the treatment of idiopathic trigeminal neuralgia: 25-year experience with 1,600 patients. Neurosurgery. 2001;48:524–534.
310
LI ET AL.
15. Fraioli MF, Cristino B, Moschettoni L, Cacciotti G, Fraioli C. Validity of percutaneous controlled radiofrequency thermocoagulation in the treatment of isolated third division trigeminal neuralgia. Surg Neurol. 2009;71:180–183. 16. Kosugi S, Shiotani M, Otsuka Y, et al. Long-term outcomes of percutaneous radiofrequency thermocoagulation of gasserian ganglion for 2nd and multiple-division trigeminal neuralgia. Pain Pract. 2014; doi: 10.1111/papr.12163. [Epub ahead of print]. 17. Sanders M, Henny CP. Results of selective percutaneous controlled radiofrequency lesion for treatment of trigeminal neuralgia in 240 patients. Clin J Pain. 1992;8:23–27. 18. Khairnar KB, Bhusari PA. An anatomical study on the foramen ovale and the foramen spinosum. J Clin Diagn Res. 2013;7:427–429. 19. Milne AD, Chui L, Mishra AV, Maxner CE. Unilateral hypoplasia of the trigeminal ganglion. Can J Ophthalmol. 2005;40:772–774. 20. Peris-Celda M, Graziano F, Russo V, Mericle RA, Ulm AJ. Foramen ovale puncture, lesioning accuracy, and avoiding complications: microsurgical anatomy study with clinical implications. J Neurosurg. 2013;119:1176–1193. 21. Huibin Q, Jianxing L, Guangyu H, Dianen F. The treatment of first idiopathic trigeminal neuralgia with radio-
frequency thermocoagulation of the peripheral branches compared to conventional radiofrequency. J Clin Neurosci. 2009;16:1425–1429. 22. Zhang WJ, Wang Y, Chen MJ. Research of the puncture depth in treatment of trigeminal neuralgia (TN) by radiofrequency thermocoagulation under CT location. Shanghai Kou Qiang Yi Xue. 2003;12:94–95. 23. Noma N, Kobayashi A, Kamo H, Imamura Y. Trigeminal neuralgia due to vertebrobasilar dolichoectasia: three case reports. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2009;108:e50–e55. 24. Naphade PS, Keraliya AR. Missing trigeminal nerve found in trigeminal neuralgia. Neurol India. 2014;62:112. 25. Cheshire WP. Trigeminal neuralgia: for one nerve a multitude of treatments. Expert Rev Neurother. 2007;7:1565– 1579. 26. Inoue H, Kondo A, Shimano H, Yasuda S. Recurrent trigeminal neuralgia at 20 years after surgery. Neurol Med Chir. 2013;53:37–39. 27. Zhang LW, Liu YG, Wu CY, Xu SJ, Zhu SG. Radiofrequency thermocoagulation rhizotomy for recurrent trigeminal neuralgia after microvascular decompression. Chin Med J. 2011;124:3726–3730.
Application of Antidromic Conduction Monitoring in Ganglion Radiofrequency Thermocoagulation for Locating Trigeminal Branches in Trigeminal Neuralgia Xiuhua Li, MD; Jianning Yue, MD; Liqiang Yang, MD; Huijie Yang, MD; Shuyue Zheng, MD; Liangliang He, MD; Jiaxiang Ni, MD Department of Pain Management, Xuanwu Hospital of Capital Medical University, Beijing, China
& Abstract Objective: The aim of this study was to investigate whether antidromic conduction monitoring (ACM) can be utilized to map the trigeminal system under sedation as a potential substitute for subjective paresthesia description (SPD) during percutaneous ganglion radiofrequency thermocoagulation (PGRT). Methods: Eighty-two patients with 152 pain divisions of trigeminal neuralgia (TN) were treated by computed tomography (CT)-guided PGRT. After the puncture needle entered the foramen ovale (FO), sensory and motor stimulation were applied to locate the pain division. And the corresponding voltage values were recorded by patients’ SPD. In the following, the proper location was certified by ACM. The corresponding earliest waves and voltage values in the identified trigeminal branch were also recorded to outline a comparison between two methods. Results: The correlation of ACM and patients’ SPD with voltage at ≤ 0.5 V was statistically significant (P < 0.05,
r = 0.159; Spearman’s rank correlation analysis). Although ACM and SPD showed weak correlation, as their interclass correlation coefficient was significant (F = 1.868, P < 0.01) with coefficient of internal consistency. Moreover, the two methods had consistency. Kruskal–Wallis test showed that ophthalmic (V1), maxillary (V2), and mandibular (V3) divisions had significant differences for test sensitivity (H = 15.945, P < 0.01). For comparison of sensitivities with ACM, V3 was most sensitive followed by V2 and then V1. Conclusion: ACM could potentially substitute for SPD of the paresthesias intra-operatively, enabling greater specificity and eliminating the need to interrupt the administration of anesthetic. These improvements would increase patient satisfaction and practitioner efficiency and accuracy. & Key Words: trigeminal neuralgia, radiofrequency thermocoagulation, antidromic conduction, subjective paresthesia, correlation
INTRODUCTION Address correspondence and reprint request to: Jiaxiang Ni, MD, Department of pain management, Xuanwu Hospital of Capital Medical University, No 45 Changchun Street, Xicheng Zone, Beijing 100053, China. E-mail: [email protected]. Submitted: August 8, 2014; Revised November 9, 2014; Revision accepted: November 28, 2014 DOI. 10.1111/papr.12286
© 2015 World Institute of Pain, 1530-7085/16/$15.00 Pain Practice, Volume 16, Issue 3, 2016 305–310
Trigeminal neuralgia (TN) is craniofacial pain involving 1 or more trigeminal nerve branches. It is characterized by mostly shock-like pain.1 Various therapies can be used for TN: medical treatment; microvascular decompression (MVD); gamma knife radiosurgery; and percutaneous treatments including balloon compression; glycerol rhizotomy; and radiofrequency (RF) thermocoagulation.2–4 Different operators have different options
306
LI ET AL.
for TN according to their experience and skills. Generally, RF thermocoagulation is quite easy and safe to perform under computed tomography (CT) guidance with no mortality and low morbidity, especially in elderly patients.5–7 When the puncture needle enters the foramen ovale (FO), sensory (50 Hz) and motor (2 Hz) test stimulation are used to elicit paresthesia in the affected trigeminal distribution area. In the process of the electrical stimulation, patients must keep conscious and cooperative for the subjective paresthesia description (SPD).8 However, much time is wasted when administering intermittent anesthesia, and the repeated anesthesia induction may make the patients’ description somewhat unreliable, not to mention the added procedural pain and discomfort. Neural electrophysiological procedures have emerged to solve drawbacks. It has been suggested that the use of trigeminal evoked potentials (TEPs), which usually entails orthodromic conduction, can be used to detect the needle position. This procedure has produced satisfactory results with TN in the maxillary (V2) division, but it is tedious and time-consuming.9 Jones and Karol have observed that antidromic stimulation could generate wider amplitude and longer latency stage of the negative phase, and the waves were easy to be observed and obtained.10,11 Bendersky et al.12 showed that antidromic nerve conduction monitoring (ACM) was useful for determining the exact innervation of trigeminal nerve divisions in anesthetized patients, making percutaneous ganglion radiofrequency thermocoagulation (PGRT) safer and more comfortable. However, it has not been reported whether ACM can be utilized to map the trigeminal system under sedation as a substitute for SPD during PGRT. It is unknown if different TN divisions can achieve the same sensitivity with ACM. Therefore, this study was conducted to investigate the correlation and consistency of the ACM and the SPD methods, and measure the sensitivity/selectivity of both techniques.
METHODS The institutional research ethics committee approved this study. A series of 82 patients (35 men, 47 women) with classic trigeminal neuralgia were recruited and gave informed consent. They were treated with CT-guided ganglion RF thermocoagulation. In each case, the procedure was performed in a sterile CT examination room with intermittent anesthesia. Each patient was placed in a supine position with their head hanging over the bed. The puncture location at the FO was deter-
mined on the CT scan, and the corresponding percutaneous point was marked. Following sterilization and local anesthesia with 1% lidocaine, a 22 g RF needle with a 5-mm working zone (straight; Cosman, Burlington, MA, USA) was inserted through the marked skin point at the FO under CT guidance. Electrical stimulations were performed to confirm and re-adjust the needle position. After the needle position was identified, anesthesia with propofol and an analgesic was administered intravenously. Intubation was not required. The ganglion was coagulated with consistent radiofrequency at 70 to 75°C for 120 second twice or more depending on the individual’s diversity and the physician’s experience. A multichannel electromyography/evoked potentials (EMG/EP) machine (DAVINCI, Mogliano Veneto, Italy) was used for neurophysiological monitoring. After cleansing the facial skin with 95% alcohol, three pairs of surface electrodes with conductive paste to eliminate skin electrical impedance as much as possible were placed, respectively, on the supraorbital foramen, infraorbital foramen, and mental foramen (negative electrode on the foramen and positive electrode < 2 cm from the negative electrode, ground wire on the ipsilateral earlobe). In previous studies, needle electrodes were also used for recording, but some injures to different extent was induced, as reported by Leandri and Gottlieb9 and Bendersky et al.12 In the present research, we used three pairs of surface electrodes without harming the patients and with their high degree of satisfaction.13 Three channels were connected with the three pairs of electrodes: channel 1—V1, channel 2—V2, and channel 3—V3. Sensory (50 Hz) and motor (2 Hz) test stimulation were applied to adjust the needle position until paresthesia of the corresponding trigeminal branches was elicited, and each minimum stimulation voltage value was recorded, respectively. A proper needle position was defined as a stimulation voltage value of < 0.5 V according to the patient’s SPD. After verification, the electrical stimulation was returned to 0 V again to certify the proper location of needle tip by ACM and began to increase it slowly until the first emerging waveform of the EMG machine in the “needle EMG” state was observed in the corresponding channel. The minimum stimulation voltage was, respectively, recorded at the same time. If a description of paresthesia from the patient for a particular pain division could not be achieved, the apparent, prominent waveforms recorded by ACM were used for identifying the trigeminal distribution.
Neurophysiological Monitoring for TN 307
Statistical analysis was performed using the Statistical Package for the Social Sciences version 19.0 (SPSS, Chicago, IL, USA). The frequency was set for the correlation analysis with Spearman’s ranked data between ACM and SPD. The intraclass correlation coefficient was used to evaluate the coherence between two methods. The Kruskal–Wallis test evaluated the sensitivity among three divisions with ACM. P < 0.05 was considered statistically significant.
RESULTS Eighty-two patients diagnosed with typical TN in 152 pain divisions, including 25 patients with one division, 44 patients with two divisions, and 13 patients with three divisions recruited in the study. Ages ranged from 30 to 84 years. Magnetic resonance imaging results were normal in all patients. The patients’ characteristics were described in detail in Table 1. After verification by patients’ SPD, the waves that achieved the highest amplitude in the corresponding division were recorded through ACM method. When the needle tip was certified in V1, the highest amplitude of waves was recorded in channel 1, and the other channels
demonstrated smaller or no waves (Figure 1A). It was also applied to V2 (Figure 1B) and V3 (Figure 1C). During the process of electrical stimulation, 146 pain divisions were positively detected, six pain divisions (four in V1, one in V2, and one in V3) were negatively detected according to patients’ SPDs, 148 pain divisions were positively detected (≤ 0.5 V), and four pain divisions (detected only in V1) were considered negative (> 0.5 V) by ACM. Although the correlation coefficient was weak, the correlation comparison of ACM and SPD in different ranges of voltage values (≤ 0.5 V) had statistical significance (r = 0.159, P < 0.05). Moreover, the two detection methods had significant coherence (F = 1.868, P < 0.01), indicating that ACM could be potentially used to displace the SPD method (Table 2). Furthermore, the voltage values’ distribution of ACM was significantly lower than that obtained by patients’ SPDs (P < 0.05) (Figure 2). Figure 3 showed that the comparison of the detected voltage values under positive monitoring with ACM had statistical differences for V1, V2, and V3. Among them, the detection for V3 achieved the highest sensitivity (voltage value 0.07 0.04), followed by V2 (voltage value 0.12 0.09) and then V1 (voltage value 0.19 0.16) (H = 15.945, P < 0.01).
DISCUSSION
Table 1. Trigeminal Neuralgia Patients Characteristics Feature Number of patients Age (year), mean age SD Gender (female/male), n (%) Duration of pain symptom SD, (m) Affected side, n (%) Left Right Division of the trigeminal nerve, n (%) V2 V3 V1 + V2 V2 + V3 V1 + V2 + V3
82 30~84 (62.3 11.7) 47/35 (57.3/42.7) 52.6 83.7 35 (42.7) 47 (57.3) 12 13 14 30 13
(14.6) (15.9) (17.0) (36.6) (15.9)
V1 = Ophthalmic division; V2 = maxillary division; V3 = mandibular division.
A
B
Eighty-two patients with TN in 152 trigeminal branches were studied. There were three patients experiencing moderate pain after PGRT, however, a local block improved their pain after procedure. The complete pain relief rate was 96%, which coincided with the present literatures about PGRT applied under imaging guidance.14,15 None of the patients had apparent or severe intracranial or extracranial structural injuries.16,17 Generally, the conventional method for locating trigeminal nerve branches is the patient’s subjective description of his or her paresthesia.11 It has been C
Figure 1. (A) The apparent waveforms in channel 1 showed the needle location in V1 from sensory stimulation with 50 Hz 0.1 millisecond recorded in ganglion by the EMG/EP machine; (B) showed the needle location in V2 and (C) showed the location in V3. All of them were shown with the correct SPD by the patients.
308
LI ET AL.
Table 2. Voltage Value Detection with ACM and SPD in 82 Cases with 152 Different Pain Divisions V1 (27)
V2 (69)
V3 (56)
Trigeminal pain divisions (n) Positive (0 to 0.5 V)/negative (> 0.5 V)
Positive
Negative
Positive
Negative
Positive
Negative
ACM SPD
23 23
4 4
69 68
0 1
56 55
0 1
Figure 2. Showed the different distribution of ACM and SPD in different voltage value.
Figure 3. Showed the different voltage value of ACM in different branches.
reported that antidromic stimulation could generate wider amplitude and longer latency stage of the negative phase, and the waves were easily observed and obtained.10,11 ACM was also useful for determining the exact innervation of trigeminal nerve divisions in anesthetized patients, making PGRT safer and more comfortable.12 In this study, ACM was also used to identify the division location with routine sensory stimulation during CT-guided PGRT. Combined with the collective original data for the analysis, the voltage values were divided into three sectors by ACM: 0 to
0.05 V (78 pain divisions detected at 0.05 V); 0.05 to 0.15 V (53 pain divisions); and 0.15 to 0.5 V (17 pain divisions). Three sectors for the patients’ SPDs: 0 to 0.2 V (46 pain divisions), 0.2 to 0.3 V (76 pain divisions), and 0.3 to 0.5 V (24 pain divisions). Correlation and coherence were analyzed, respectively, between the two detection methods. As the results showed, although the correlation coefficient was weak, the correlation comparison of ACM and SPD in different ranges of voltage values (≤ 0.5 V) had statistical significance (r = 0.159, P < 0.05). Moreover, the two methods had significant coherence (F = 1.868, P < 0.01), indicating that ACM detection could potentially be used instead of the conventional SPD method. Furthermore, the voltage threshold of ACM was obviously lower than that of patients’ SPD, decreasing procedural pain and patient’s discomfort. In our study, the successful detection rate for ACM at 0 to 0.3 V was 74% in V1, 97% in V2, and 100% in V3. At 0 to 0.5 V, it was 85% in V1 and 100% in V2. ACM detection in V3 showed the most sensitivity, whereas the worst detection sensitivity was in V1. We presumed that it related to the anatomical positions of nerve cell bodies in the ganglion. The trigeminal ganglion has a somatotopic arrangement in which V1 is the most craniomedial and V3 the most lateral. Namely, V1 is in the medial third of the FO, V2 in the middle of the FO, and V3 in the lateral third of the FO.15,18 According to anatomical research, V1 is the most deeply located, and operators are extremely cautious when approaching V1 to avoid severe complications, such as cranial nerve injury and intracranial hemorrhage.19,20 Puncture at V1 is the most difficult to deal with and reach. Even with optimal trajectories, V1 continues to be the least selected for targeting (73.3%), topped by V2 (84.6%) and V3 (100%).20 Puncture is relatively easy through the FO into V3, where detection for locating achieves the highest sensitivity. Detection was negative at high voltage (> 0.5 V) in our study, with no waves emerging and no paresthesia reported by the patients. Negative detection in V1 was found in four pain divisions by ACM (waves emerged in V1 at 0.7 to 0.9 V) and SPD (waves emerged in V1 at 0.8
Neurophysiological Monitoring for TN 309
to 1.2 V). For the negative detection in V1, which detected at high voltage, CT guidance can be used to identify the proper needle position in the FO. Considering the mentioned risk,16,20–22 the procedure was undergone with PGRT and the four patients achieved pain relief. Some case reports showed that trigeminal neuralgia was due to vertebrobasilar dolichoectasia and distortion affecting the trigeminal nerve.23,24 In one case, it was uncertain whether or not the negative detection was related to the vascular deformity. The active tip of the radiofrequency needle used in this study was 5 mm, the thermal lesion of which was larger than the 2-mm needle.7 This might explain why the patient achieved pain relief after procedure with negative detection. We also suggested that the large range of 0 to 1.2 V could be used to detect areas in V1, combined with sensory stimulation of 50 Hz for 0.1 millisecond.21 There was no negative detection in V2 or V3 with ACM. Negative detection by SPD was the same in V2 and V3, with one pain division, but prominent waveforms were identified in V2 and V3 by ACM. Based on these findings, we suggested that ACM was more reliable than SPD for identifying the trigeminal nerve divisions, especially V2 and V3. ACM could potentially displace the SPDs in V2 and V3, and in V1 with high voltage (> 0.5 V). In total, 39 patients underwent local nerve block, acupuncture, and percutaneous intervention therapy before this study. Examinations before PGRT showed there were various injuries to peripheral nerves.25 The amplitudes varied among patients at the same level of stimulation (eg, 0.3 V); however, it showed no effect on the division locations. Therefore, we surmised that patients had different nerve conduction, resulting in individual differences. Future research should be taken to explore the lesion level of RF thermocoagulation in individuals using ACM in cases of TN recurrence at the original location.26,27 Although ACM for locating V1 in some cases required a higher voltage for sensory stimulation, this study showed that there was correlation and coherence between ACM and SPD regarding identification of the trigeminal nerve branches’ locations during PGRT. ACM was even more reliable than SPDs in V2 and V3, and most sensitive in V3. Despite this, some limitations of the study required consideration. In particular, the cohort size was not a large number (82 subjects); however, we can draw a conclusion that antidromic conduction monitoring could potentially substitute for the patient’s description of the paresthesias intra-
operatively enabling greater specificity, eliminating the need to interrupt the administration of anesthetic. These improvements would improve patient satisfaction and practitioner efficiency and accuracy.
REFERENCES 1. Headache Classification Committee of the International Headache Society (IHS). The international classification of headache disorders, 3rd edition (beta version). Cephalalgia. 2013;33:629–808. 2. Cheng JS, Lim DA, Chang EF, Barbaro NM. A review of percutaneous treatment for trigeminal neuralgia. Neurosurgery. 2014;10:25–33. 3. Parmar M, Sharma N, Modqill V, Naidu P. Comparative evaluation of surgical procedures for trigeminal neuralgia. J Maxillofac Oral Surg. 2013;12:400–409. 4. Koopman JS, de Vries LM, Dieleman JP, Huygen FJ, Stricker BH, Sturkenboom MC. A nationwide study of three invasive treatments for trigeminal neuralgia. Pain. 2011;152:507–513. 5. Cheng JS, Sanchez-Mejia RO, Limbo M, Ward MM, Barbaro NM. Management of medically refractory trigeminal neuralgia in patients with multiple sclerosis. Neurosurg Focus. 2005;18:e13. 6. Berk C, Constantoyannis C, Honey CR. The treatment of trigeminal neuralgia in patients with multiple sclerosis using percutaneous radiofrequency rhizotomy. Can J Neurol Sci. 2003;30:220–223. 7. van Kleef M, van Genderen WE, Narouze S, et al. Trigeminal neuralgia. Pain Pract. 2009;9:252–259. 8. Hart MG, Nowell M, Coakham HB. Radiofrequency thermocoagulation for trigeminal neuralgia without intra-operative patient waking. Br J Neurosurg. 2012;26: 392–396. 9. Leandri M, Gottlieb A. Trigeminal evoked potentialmonitored thermorhizotomy: a novel approach for relief of trigeminal pain. J Neurosurg. 1996;84:929–939. 10. Jones DL, Wolford LM. Intraoperative recording of trigeminal evoked potentials during orthognathic surgery. Int J Adult Orthodon Orthognath Surg. 1990;5:167–174. 11. Karol EA, Sanz OP, Rey RD. Sensory and motor trigeminal evoked potentials to localize the position of trigeminal electrodes. Acta Neurochir (Wien). 1991;108:110–115. 12. Bendersky M, Hem S, Landriel F, et al. Identifying the trigeminal nerve branches for transovale radiofrequency thermolesion: “no pain, no stress”. Neurosurgery. 2012;70:259– 263. 13. Obermann M, Yoon MS, Ese D. Impaired trigeminal nociceptive processing in patients with trigeminal neuralgia. Neurology. 2007;69:835–841. 14. Kanpolat Y, Savas A, Bekar A, Berk C. Percutaneous controlled radiofrequency trigeminal rhizotomy for the treatment of idiopathic trigeminal neuralgia: 25-year experience with 1,600 patients. Neurosurgery. 2001;48:524–534.
310
LI ET AL.
15. Fraioli MF, Cristino B, Moschettoni L, Cacciotti G, Fraioli C. Validity of percutaneous controlled radiofrequency thermocoagulation in the treatment of isolated third division trigeminal neuralgia. Surg Neurol. 2009;71:180–183. 16. Kosugi S, Shiotani M, Otsuka Y, et al. Long-term outcomes of percutaneous radiofrequency thermocoagulation of gasserian ganglion for 2nd and multiple-division trigeminal neuralgia. Pain Pract. 2014; doi: 10.1111/papr.12163. [Epub ahead of print]. 17. Sanders M, Henny CP. Results of selective percutaneous controlled radiofrequency lesion for treatment of trigeminal neuralgia in 240 patients. Clin J Pain. 1992;8:23–27. 18. Khairnar KB, Bhusari PA. An anatomical study on the foramen ovale and the foramen spinosum. J Clin Diagn Res. 2013;7:427–429. 19. Milne AD, Chui L, Mishra AV, Maxner CE. Unilateral hypoplasia of the trigeminal ganglion. Can J Ophthalmol. 2005;40:772–774. 20. Peris-Celda M, Graziano F, Russo V, Mericle RA, Ulm AJ. Foramen ovale puncture, lesioning accuracy, and avoiding complications: microsurgical anatomy study with clinical implications. J Neurosurg. 2013;119:1176–1193. 21. Huibin Q, Jianxing L, Guangyu H, Dianen F. The treatment of first idiopathic trigeminal neuralgia with radio-
frequency thermocoagulation of the peripheral branches compared to conventional radiofrequency. J Clin Neurosci. 2009;16:1425–1429. 22. Zhang WJ, Wang Y, Chen MJ. Research of the puncture depth in treatment of trigeminal neuralgia (TN) by radiofrequency thermocoagulation under CT location. Shanghai Kou Qiang Yi Xue. 2003;12:94–95. 23. Noma N, Kobayashi A, Kamo H, Imamura Y. Trigeminal neuralgia due to vertebrobasilar dolichoectasia: three case reports. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2009;108:e50–e55. 24. Naphade PS, Keraliya AR. Missing trigeminal nerve found in trigeminal neuralgia. Neurol India. 2014;62:112. 25. Cheshire WP. Trigeminal neuralgia: for one nerve a multitude of treatments. Expert Rev Neurother. 2007;7:1565– 1579. 26. Inoue H, Kondo A, Shimano H, Yasuda S. Recurrent trigeminal neuralgia at 20 years after surgery. Neurol Med Chir. 2013;53:37–39. 27. Zhang LW, Liu YG, Wu CY, Xu SJ, Zhu SG. Radiofrequency thermocoagulation rhizotomy for recurrent trigeminal neuralgia after microvascular decompression. Chin Med J. 2011;124:3726–3730.
