Browse the corpus

Characteristics of Pain During MRI-Guided Focused Ultrasound Thalamotomy. BACKGROUND: Magnetic resonance imaging-guided focused ultrasound (MRgFUS) has become popular as an incisionless mode of neurosurgical treatment. However, head pain during sonication is common and its pathophysiology remains poorly understood. OBJECTIVE: To explore the characteristics of head pain occurring during MRgFUS thalamotomy. METHODS: Our study comprised 59 patients who answered questions about the pain they experienced during unilateral MRgFUS thalamotomy. The location and features of pain were investigated using a questionnaire including the numerical rating scale (NRS) to estimate maximum pain intensity and the Japanese version of the Short Form of McGill Pain Questionnaire 2 to evaluate the quantitative and qualitative dimensions of pain. Several clinical factors were investigated for possible correlation with pain intensity. RESULTS: Forty-eight patients (81%) reported sonication-related head pain, and the degree of pain was severe (NRS score ≥ 7) in 39 patients (66%). The distribution of sonication-related pain was "localized" in 29 (49%) and "diffuse" in 16 (27%); the most frequent location was the "occipital" region. The pain features most frequently reported were those in the "affective" subscale of the Short Form of McGill Pain Questionnaire 2. Patients with diffuse pain had a higher NRS score and lower skull density ratio than did patients with localized pain. The NRS score negatively correlated with tremor improvement at 6 months post-treatment. CONCLUSION: Most patients in our cohort experienced pain during MRgFUS. The distribution and intensity of pain varied according to the skull density ratio, indicating that the pain might have had different origins. Our results may contribute to the improvement of pain management during MRgFUS.

This study was approved by the Ethics Committee of our institution and performed in accordance with the Declaration of Helsinki. Patient consent was waived because of the retrospective design of this study; however, an option for patients to opt out of the study was provided. Seventy-nine patients with medication-refractory tremors underwent MRgFUS thalamotomy of the unilateral VIM nucleus between December 2015 and April 2022 using the Exablate Neuro system. The characteristics of pain occurring during MRgFUS were investigated using a questionnaire for 59 patients (Figure 1). The answers were obtained in the outpatient clinic for 22 patients who underwent MRgFUS before June 2020 and on the same day for 37 consecutive patients after July 2020. We considered that it would be better to have the patient answer the questionnaire in a relaxed state, and the data were obtained in the outpatient clinic. However, it became uncertain whether patients would be able to return to the clinic because of the COVID-19 pandemic, and the data were obtained on the same day after the treatment after July 2020. For the latter group, the questionnaire was administered after the post-treatment procedures (such as moving out from the MRI room and removal of the frame or membrane), and routine neurological examinations were completed. At this time, the patients were relaxed and free from sonication-related pain. Study flowchart. CRST, clinical rating scale for tremor.

This study was approved by the Ethics Committee of our institution and performed in accordance with the Declaration of Helsinki. Patient consent was waived because of the retrospective design of this study; however, an option for patients to opt out of the study was provided. Seventy-nine patients with medication-refractory tremors underwent MRgFUS thalamotomy of the unilateral VIM nucleus between December 2015 and April 2022 using the Exablate Neuro system. The characteristics of pain occurring during MRgFUS were investigated using a questionnaire for 59 patients (Figure 1). The answers were obtained in the outpatient clinic for 22 patients who underwent MRgFUS before June 2020 and on the same day for 37 consecutive patients after July 2020. We considered that it would be better to have the patient answer the questionnaire in a relaxed state, and the data were obtained in the outpatient clinic. However, it became uncertain whether patients would be able to return to the clinic because of the COVID-19 pandemic, and the data were obtained on the same day after the treatment after July 2020. For the latter group, the questionnaire was administered after the post-treatment procedures (such as moving out from the MRI room and removal of the frame or membrane), and routine neurological examinations were completed. At this time, the patients were relaxed and free from sonication-related pain. Study flowchart. CRST, clinical rating scale for tremor. The patients were asked whether they had felt pain during the procedure and when they started to feel pain, indicated the location and distribution of pain they transiently felt during sonication on a prepared illustration of the head, and described the features of pain freely in their own words. The intensity of the worst pain was expressed using the numerical rating scale (NRS). On this scale, 0 means no pain, 1 to 3 mean “slight” pain, 4 to 6 mean “moderate” pain, and 7 to 10 mean “severe” pain, with a score of 10 indicating the worst imaginable pain.27

scribed the features of pain freely in their own words. The intensity of the worst pain was expressed using the numerical rating scale (NRS). On this scale, 0 means no pain, 1 to 3 mean “slight” pain, 4 to 6 mean “moderate” pain, and 7 to 10 mean “severe” pain, with a score of 10 indicating the worst imaginable pain.27 In addition, the features of pain were assessed using the Japanese version of the Short Form of McGill Pain Questionnaire 2 (SF-MPQ2), which is a revised version of the McGill Pain Questionnaire. The SF-MPQ2 is composed of 22 items and has been validated for use with Japanese-speaking patients.28 For each item, pain is scored as 0 to 10, with a lower score indicating less pain. The items are divided among 4 subscales of pain characteristics: continuous, intermittent, neuropathic, and affective.28 To determine the dominant subscale of pain during MRgFUS, the number of patients who selected a particular item and its scores were assessed.

ored as 0 to 10, with a lower score indicating less pain. The items are divided among 4 subscales of pain characteristics: continuous, intermittent, neuropathic, and affective.28 To determine the dominant subscale of pain during MRgFUS, the number of patients who selected a particular item and its scores were assessed. The following clinical factors and treatment parameters were assessed for their possible relation to the pain characteristics: age, sex, treatment side, tremor etiology, skull density ratio (SDR), number of maximum active transducer elements, number of sonications, number of targets, time in the MRI scanner, treatment time, maximum power, maximum energy, maximum peak temperature, and maximum average temperature. Treatment time was defined as the duration between the beginning of the first sonication and the ending of the last sonication. Clinical rating scale for tremor29 (CRST) at baseline and 1 and 6 months after procedure was collected to evaluate the effect of pain on treatment efficacy. The CRST improvement rate (%) was defined as ([baseline CRST − 1- or 6-month CRST]/baseline CRST) × 100%, and we investigated whether it was affected by pain.

ion. Clinical rating scale for tremor29 (CRST) at baseline and 1 and 6 months after procedure was collected to evaluate the effect of pain on treatment efficacy. The CRST improvement rate (%) was defined as ([baseline CRST − 1- or 6-month CRST]/baseline CRST) × 100%, and we investigated whether it was affected by pain. Patient demographics and clinical and treatment parameters were characterized using descriptive statistics, either mean ± SD or median (range) for continuous variables or number (%) for categorical variables. Differences in NRS and SDR between pain locations were compared by the Mann–Whitney test. Friedman and Dunn's post hoc tests were used to analyze differences between the SF-MPQ2 subscales. To assess the association of the NRS score with the CRST improvement rate, each clinical and treatment factors, Spearman's rank correlation coefficient was calculated for continuous variables and the Mann–Whitney test was used for categorical values. Missing values were not imputed. A P value < .05 was considered statistically significant. All statistical analyses were performed using JMP Pro 16.2.0 (SAS Institute Inc) and GraphPad Prism version 9.4.0 for MacOS (GraphPad Inc; www.graphpad.com).

The patients were asked whether they had felt pain during the procedure and when they started to feel pain, indicated the location and distribution of pain they transiently felt during sonication on a prepared illustration of the head, and described the features of pain freely in their own words. The intensity of the worst pain was expressed using the numerical rating scale (NRS). On this scale, 0 means no pain, 1 to 3 mean “slight” pain, 4 to 6 mean “moderate” pain, and 7 to 10 mean “severe” pain, with a score of 10 indicating the worst imaginable pain.27 In addition, the features of pain were assessed using the Japanese version of the Short Form of McGill Pain Questionnaire 2 (SF-MPQ2), which is a revised version of the McGill Pain Questionnaire. The SF-MPQ2 is composed of 22 items and has been validated for use with Japanese-speaking patients.28 For each item, pain is scored as 0 to 10, with a lower score indicating less pain. The items are divided among 4 subscales of pain characteristics: continuous, intermittent, neuropathic, and affective.28 To determine the dominant subscale of pain during MRgFUS, the number of patients who selected a particular item and its scores were assessed.

The following clinical factors and treatment parameters were assessed for their possible relation to the pain characteristics: age, sex, treatment side, tremor etiology, skull density ratio (SDR), number of maximum active transducer elements, number of sonications, number of targets, time in the MRI scanner, treatment time, maximum power, maximum energy, maximum peak temperature, and maximum average temperature. Treatment time was defined as the duration between the beginning of the first sonication and the ending of the last sonication. Clinical rating scale for tremor29 (CRST) at baseline and 1 and 6 months after procedure was collected to evaluate the effect of pain on treatment efficacy. The CRST improvement rate (%) was defined as ([baseline CRST − 1- or 6-month CRST]/baseline CRST) × 100%, and we investigated whether it was affected by pain.

Patient demographics and clinical and treatment parameters were characterized using descriptive statistics, either mean ± SD or median (range) for continuous variables or number (%) for categorical variables. Differences in NRS and SDR between pain locations were compared by the Mann–Whitney test. Friedman and Dunn's post hoc tests were used to analyze differences between the SF-MPQ2 subscales. To assess the association of the NRS score with the CRST improvement rate, each clinical and treatment factors, Spearman's rank correlation coefficient was calculated for continuous variables and the Mann–Whitney test was used for categorical values. Missing values were not imputed. A P value < .05 was considered statistically significant. All statistical analyses were performed using JMP Pro 16.2.0 (SAS Institute Inc) and GraphPad Prism version 9.4.0 for MacOS (GraphPad Inc; www.graphpad.com).

All procedures were completed without severe complications such as mortality or permanent severe neurological morbidity. The clinical characteristics of the 59 patients in our study are shown in Table 1. The median age of the patients was 72 (38-84) years. Most patients were men (n = 47, 80%), and the target was the left VIM nucleus in 48 patients (81%). The median baseline CRST was 47 (19-86), and the tremor etiologies were essential tremors (n = 51, 86%) and Parkinson's tremors (n = 8, 14%). The mean SDR was 0.48 ± 0.08. Clinical Characteristics CRST, clinical rating scale for tremor; SDR, skull density ratio; VIM, ventral intermediate. Data collected from all patients. Data collected from 52 patients (88%). Values are presented as median (range), number of patients (%), or mean ± SD. The treatment-related parameters are listed in Table 2. All 59 patients were followed up for at least 1 month. Four patients were lost to follow-up at 6 months owing to the COVID-19 pandemic, and 3 patients were less than 6 months post-treatment; hence, 52 patients (88%) completed the 6-month follow-up (Figure 1). The median CRST was 18 (2-65) at 1 month and 25 (4-77) at 6 months, which corresponded to 54% ± 26% and 40% ± 33% improvement from the baseline CRST, respectively. Regarding local complications, moderate scalp swelling was observed in 4 patients with an SDR ranging from 0.31 to 0.38. Treatment Parameters Values are presented as median (range) or mean ± SD.

The treatment-related parameters are listed in Table 2. All 59 patients were followed up for at least 1 month. Four patients were lost to follow-up at 6 months owing to the COVID-19 pandemic, and 3 patients were less than 6 months post-treatment; hence, 52 patients (88%) completed the 6-month follow-up (Figure 1). The median CRST was 18 (2-65) at 1 month and 25 (4-77) at 6 months, which corresponded to 54% ± 26% and 40% ± 33% improvement from the baseline CRST, respectively. Regarding local complications, moderate scalp swelling was observed in 4 patients with an SDR ranging from 0.31 to 0.38. Treatment Parameters Values are presented as median (range) or mean ± SD. Forty-eight of the 59 patients (81%) complained of head pain. The pain was “severe” (NRS 7-10) in 39 patients (66%) and “moderate” (NRS 4-6) in 9 patients (15%) (Figure 2). During the procedure, nonsteroidal anti-inflammatory drugs were administered to 30 patients (51%) and non-narcotic centrally acting analgesics to 5 patients (8%). Despite this, 21 patients (36%) reported pain with an NRS score of 10, indicating the worst imaginable pain, and 14 (24%) pushed the emergency stop button during sonication because of unbearable pain. Distribution of numerical rating scale scores.

Forty-eight of the 59 patients (81%) complained of head pain. The pain was “severe” (NRS 7-10) in 39 patients (66%) and “moderate” (NRS 4-6) in 9 patients (15%) (Figure 2). During the procedure, nonsteroidal anti-inflammatory drugs were administered to 30 patients (51%) and non-narcotic centrally acting analgesics to 5 patients (8%). Despite this, 21 patients (36%) reported pain with an NRS score of 10, indicating the worst imaginable pain, and 14 (24%) pushed the emergency stop button during sonication because of unbearable pain. Distribution of numerical rating scale scores. Pain during sonication was classified as localized or diffuse. Twenty-nine patients reported localized pain, which was described as “occipital” (n = 13, 29%), “temple” (n = 8, 18%), or “convexity” (frontal to parietal) (n = 8, 18%). Sixteen patients reported diffuse pain, either “whole head” (n = 10, 17%) or “center of head” (n = 6, 10%). An additional 3 patients (5%) complained of a continuous ring-shaped (circular) distribution of pain that was considered to be frame-related owing to a pin or tight membrane rather than sonication-related; these patients were excluded from further analysis. The 16 patients with diffuse pain had a higher NRS score (NRS score: 10 [8-10] vs 9 [4-10], P = .02, Mann–Whitney U-test, Figure 3A) and a lower SDR (SDR: 0.42 [0.31-0.49] vs 0.49 [0.31-0.64], P = .001, Mann–Whitney U-test, Figure 3B) than did the 29 patients with localized pain.

ts were excluded from further analysis. The 16 patients with diffuse pain had a higher NRS score (NRS score: 10 [8-10] vs 9 [4-10], P = .02, Mann–Whitney U-test, Figure 3A) and a lower SDR (SDR: 0.42 [0.31-0.49] vs 0.49 [0.31-0.64], P = .001, Mann–Whitney U-test, Figure 3B) than did the 29 patients with localized pain. Pain intensity and SDR according to pain localization. A, The NRS score was higher in patients with diffuse pain than in patients with localized pain (P = .02, Mann–Whitney U-test). B, The mean SDR was lower in patients with diffuse pain than in patients with localized pain (P = .001, Mann–Whitney U-test). The data are expressed as median with IQR. NRS, numerical rating scale; SDR, skull density ratio. *P < .05, **P < .01. The qualities of pain reported freely by the 59 patients were “heat” (n = 37, 63%), “faintness” (n = 18, 31%), “pressure” (n = 11, 19%), “piercing” (n = 11, 19%), and “lightheadedness” (n = 10, 17%). However, many patients reported that such pain had never been experienced before and was difficult to describe. In addition, some answers might not have been directly related to pain; for example, the term “hotness” included variations ranging from “warm” to “painfully burning.” Essentially, it was common for the patients to experience some type of heat and a fainting sensation during high-energy ultrasound sonication.

describe. In addition, some answers might not have been directly related to pain; for example, the term “hotness” included variations ranging from “warm” to “painfully burning.” Essentially, it was common for the patients to experience some type of heat and a fainting sensation during high-energy ultrasound sonication. The SF-MPQ2 item receiving the highest score was “punishing-cruel,” followed in order by “hot-burning,” “fearful,” “sharp pain,” and “sickening” (Figure 4). The most frequently chosen items were “hot-burning” (n = 33, 56%), “sickening” (n = 29, 49%), “fearful” (n = 28, 47%), “punishing-cruel” (n = 27, 46%), “sharp pain” (n = 24, 41%), “piercing” (n = 24, 41%), and “aching” (n = 22, 37%). The score for each item and the average scores in the 4 subscales of the Short Form of McGill Pain Questionnaire 2. The average scores in the 4 subscales are enclosed by a rectangle. The data are expressed as median with IQR. The median scores in the 4 subscales of the SF-MPQ2 differed significantly: affective, 1.8; intermittent, 1.5; neuropathic, 1.2; and continuous, 0.3 (P = .0009, Friedman test). The affective score was significantly greater than the neuropathic score (P = .03) and continuous score (P = .01, Dunn's multiple comparisons test). The scores for “shooting pain,” “stabbing pain,” “electric shock pain,” and “piercing” tended to be higher in patients with diffuse vs localized pain.

009, Friedman test). The affective score was significantly greater than the neuropathic score (P = .03) and continuous score (P = .01, Dunn's multiple comparisons test). The scores for “shooting pain,” “stabbing pain,” “electric shock pain,” and “piercing” tended to be higher in patients with diffuse vs localized pain. The NRS score correlated positively with the number of maximum active transducer elements, treatment time, and maximum energy and power (Supplemental Digital Content, Figure, http://links.lww.com/NEU/D682), but negatively with the SDR and maximum peak/average temperature (Table 3). There were no significant relationships between the NRS and the following variables: sex, age, treatment side, number of sonications, and time in the MRI scanner. The median SDR was higher in the 11 patients who did not complain of pain (0.59 [0.47-0.64]) than in the 48 patients who complained of moderate or severe pain (0.46 [0.31-0.64], P < .0001, Mann–Whitney U test). Association Between the Numerical Rating Scale and Clinical and Treatment Variables CRST, clinical rating scale for tremor; SDR, skull density ratio; VIM, ventral intermediate. The Mann–Whitney U test was used. P < .05. There was no difference in score of NRS and SF-MPQ2 between the 22 patients who underwent questionnaire in the outpatient clinic and the 37 patients who did it on the same day after treatment. The detailed comparison including other factors is given in the Text and Table of Supplemental Digital Content, http://links.lww.com/NEU/D682.

fference in score of NRS and SF-MPQ2 between the 22 patients who underwent questionnaire in the outpatient clinic and the 37 patients who did it on the same day after treatment. The detailed comparison including other factors is given in the Text and Table of Supplemental Digital Content, http://links.lww.com/NEU/D682. In a Spearman correlation analysis, there was a significant negative correlation between the NRS score and the CRST improvement rate at 6 months (ρ = −0.34, P = .02) although not at 1 month (ρ = −0.20, P = .14) (Table 3).

Forty-eight of the 59 patients (81%) complained of head pain. The pain was “severe” (NRS 7-10) in 39 patients (66%) and “moderate” (NRS 4-6) in 9 patients (15%) (Figure 2). During the procedure, nonsteroidal anti-inflammatory drugs were administered to 30 patients (51%) and non-narcotic centrally acting analgesics to 5 patients (8%). Despite this, 21 patients (36%) reported pain with an NRS score of 10, indicating the worst imaginable pain, and 14 (24%) pushed the emergency stop button during sonication because of unbearable pain. Distribution of numerical rating scale scores.

Pain during sonication was classified as localized or diffuse. Twenty-nine patients reported localized pain, which was described as “occipital” (n = 13, 29%), “temple” (n = 8, 18%), or “convexity” (frontal to parietal) (n = 8, 18%). Sixteen patients reported diffuse pain, either “whole head” (n = 10, 17%) or “center of head” (n = 6, 10%). An additional 3 patients (5%) complained of a continuous ring-shaped (circular) distribution of pain that was considered to be frame-related owing to a pin or tight membrane rather than sonication-related; these patients were excluded from further analysis. The 16 patients with diffuse pain had a higher NRS score (NRS score: 10 [8-10] vs 9 [4-10], P = .02, Mann–Whitney U-test, Figure 3A) and a lower SDR (SDR: 0.42 [0.31-0.49] vs 0.49 [0.31-0.64], P = .001, Mann–Whitney U-test, Figure 3B) than did the 29 patients with localized pain. Pain intensity and SDR according to pain localization. A, The NRS score was higher in patients with diffuse pain than in patients with localized pain (P = .02, Mann–Whitney U-test). B, The mean SDR was lower in patients with diffuse pain than in patients with localized pain (P = .001, Mann–Whitney U-test). The data are expressed as median with IQR. NRS, numerical rating scale; SDR, skull density ratio. *P < .05, **P < .01.

The qualities of pain reported freely by the 59 patients were “heat” (n = 37, 63%), “faintness” (n = 18, 31%), “pressure” (n = 11, 19%), “piercing” (n = 11, 19%), and “lightheadedness” (n = 10, 17%). However, many patients reported that such pain had never been experienced before and was difficult to describe. In addition, some answers might not have been directly related to pain; for example, the term “hotness” included variations ranging from “warm” to “painfully burning.” Essentially, it was common for the patients to experience some type of heat and a fainting sensation during high-energy ultrasound sonication.

The SF-MPQ2 item receiving the highest score was “punishing-cruel,” followed in order by “hot-burning,” “fearful,” “sharp pain,” and “sickening” (Figure 4). The most frequently chosen items were “hot-burning” (n = 33, 56%), “sickening” (n = 29, 49%), “fearful” (n = 28, 47%), “punishing-cruel” (n = 27, 46%), “sharp pain” (n = 24, 41%), “piercing” (n = 24, 41%), and “aching” (n = 22, 37%). The score for each item and the average scores in the 4 subscales of the Short Form of McGill Pain Questionnaire 2. The average scores in the 4 subscales are enclosed by a rectangle. The data are expressed as median with IQR. The median scores in the 4 subscales of the SF-MPQ2 differed significantly: affective, 1.8; intermittent, 1.5; neuropathic, 1.2; and continuous, 0.3 (P = .0009, Friedman test). The affective score was significantly greater than the neuropathic score (P = .03) and continuous score (P = .01, Dunn's multiple comparisons test). The scores for “shooting pain,” “stabbing pain,” “electric shock pain,” and “piercing” tended to be higher in patients with diffuse vs localized pain.

The NRS score correlated positively with the number of maximum active transducer elements, treatment time, and maximum energy and power (Supplemental Digital Content, Figure, http://links.lww.com/NEU/D682), but negatively with the SDR and maximum peak/average temperature (Table 3). There were no significant relationships between the NRS and the following variables: sex, age, treatment side, number of sonications, and time in the MRI scanner. The median SDR was higher in the 11 patients who did not complain of pain (0.59 [0.47-0.64]) than in the 48 patients who complained of moderate or severe pain (0.46 [0.31-0.64], P < .0001, Mann–Whitney U test). Association Between the Numerical Rating Scale and Clinical and Treatment Variables CRST, clinical rating scale for tremor; SDR, skull density ratio; VIM, ventral intermediate. The Mann–Whitney U test was used. P < .05. There was no difference in score of NRS and SF-MPQ2 between the 22 patients who underwent questionnaire in the outpatient clinic and the 37 patients who did it on the same day after treatment. The detailed comparison including other factors is given in the Text and Table of Supplemental Digital Content, http://links.lww.com/NEU/D682.

In a Spearman correlation analysis, there was a significant negative correlation between the NRS score and the CRST improvement rate at 6 months (ρ = −0.34, P = .02) although not at 1 month (ρ = −0.20, P = .14) (Table 3).

The prevalence of head pain during MRgFUS ranges from 2% to 80%.1,2,5,6,12,15,16,20,30 In this study, a large percentage of patients (81%) experienced moderate to severe pain during MRgFUS thalamotomy. The pain was often intermittent and accompanied by some type of heat, pressure, or faintness. We found that the intensity and distribution of pain were related to the SDR. When we began performing MRgFUS in 2015, we used an SDR of 0.30 as the threshold for treatment. However, after experiencing poor treatment outcomes in patients with a low SDR, we gradually raised the threshold to 0.40. Currently, we fully inform all patients with an SDR of <0.50 about the efforts that might be required during the procedure. In this study, 9 patients (15%) were recruited before the threshold was raised and had an SDR of <0.40. Ultrasound absorption in the skull31 and skin32 has been observed in animal experiments. In humans, MRI changes potentially caused by MRgFUS have been reported on the brain surface and in the overlying skull, muscle, and scalp.26,33-35 In patients with a relatively low SDR, energy transmission across the skull tends to be insufficient.36,37 Therefore, to achieve the desired temperature (≥55°C) in the target, a higher ablation energy is required. Disproportional extra energy spreads to the tissues adjacent to the skull (eg, the dura, periosteum, muscles, and scalp) where nociceptors are present and thus presumed to be the source of pain.

insufficient.36,37 Therefore, to achieve the desired temperature (≥55°C) in the target, a higher ablation energy is required. Disproportional extra energy spreads to the tissues adjacent to the skull (eg, the dura, periosteum, muscles, and scalp) where nociceptors are present and thus presumed to be the source of pain. Ultrasonic energy transmission varies throughout the head. We speculate that pain originates in the area where energy transmission is insufficient; consequently, patients with a relatively high SDR are able to distinguish the pain location (eg, temple or convexity). For patients with a low SDR, a higher energy will be required in the later stage of treatment; however, because energy transmission efficiency decreases across the entire head, it is assumed that ultrasound energy that spreads in an overlapping manner rather than penetrates will result in more intense and nonlocalized distributed pain that patient feels as being over the entire or center part of the head. The detailed analysis of various data displayed in the transducer map including delivered energy, SDR, and angle of array will contribute to speculating the mechanism.

er than penetrates will result in more intense and nonlocalized distributed pain that patient feels as being over the entire or center part of the head. The detailed analysis of various data displayed in the transducer map including delivered energy, SDR, and angle of array will contribute to speculating the mechanism. Intriguingly, 13 patients (22%) in our study reported occipital pain although this location was outside the sonication path. The median NRS score for occipital pain was 8 (4-10), which was relatively low in this study. We suggest that this type of pain has a different origin than do other types of head pain. Transient occipital pain is encountered clinically in conditions where intracranial pressure fluctuates rapidly; such conditions include occipital headache in patients with Chiari malformation type 1 exacerbated by cough or Valsalva maneuver38 and intracranial hypotension. We speculate that higher-energy ultrasound causes occipital pain by inducing changes in intracranial pressure through pulsation or by altering cerebral blood in response to mild temperature elevation. Faintness is commonly reported by patients during sonication and is thought to reflect the impact of high-energy sonication.

eculate that higher-energy ultrasound causes occipital pain by inducing changes in intracranial pressure through pulsation or by altering cerebral blood in response to mild temperature elevation. Faintness is commonly reported by patients during sonication and is thought to reflect the impact of high-energy sonication. Controlling pain affects the treatment outcome, and pushing of the stop button during sonication affects the CRST.36 In our study, the NRS score negatively correlated with the CRST improvement rate at 6 months post-treatment, rather than at 1 month. We suspect that patients with intense pain hesitated to undergo additional sonication to ensure lesion formation or creation of an additional lesion, which worsened the long-term outcome. Pain management during MRgFUS is controversial, especially because the mechanisms underlying sonication-related pain have not been clarified. The intensity of pain in the scalp can be reduced by administration of a nerve block; however, the pain itself is a reliable indicator of scalp damage because of heat, and the safety of high energy application under anesthesia is questionable. It is notable that 3 patients in our cohort complained of pain because of a tight membrane or frame pin and care should be taken to prevent this kind of pain. In addition, the patients were required to strictly lie still in a narrow MRI scanner in a cold water circulation system for several hours. We should attempt to make these conditions more comfortable in the future.

d of pain because of a tight membrane or frame pin and care should be taken to prevent this kind of pain. In addition, the patients were required to strictly lie still in a narrow MRI scanner in a cold water circulation system for several hours. We should attempt to make these conditions more comfortable in the future. We acknowledge the limitations of this study. First, the study was retrospective and conducted at a single center; thus, the applicability of our findings may be limited to our patient population and treatment protocol. The pain management during MRgFUS was not standardized for each patient. Second, the CRST at 6 months was not available in 7 patients. However, because this was mainly due to the COVID-19 pandemic and unrelated to treatment, it likely had little impact on the results. Third, we used the SF-MPQ2 for pain characterization; however, the Japanese version of this questionnaire has been validated for chronic but not acute pain. Therefore, some items require careful interpretation. For example, many patients categorized their pain as “burning” because they felt hotness during MRgFUS. Although “burning” is a neuropathic item in the SF-MPQ2, the neuropathic subscale had a low score in our study. Classifying sonication-related head pain according to current diagnostic criteria is challenging. Forth, we included 2 groups in which the time of pain questionnaire was different. The “treatment time” or “maximum power” was smaller in the more recently treated group, presumably because we shifted to select the higher SDR as a threshold for treatment. However, there was no difference in NRS and SF-MPQ2 scores between the 2 groups, and many patients indicated that the pain during MRgFUS was unique and hard to forget. Therefore, we consider that the duration between administrations of the questionnaire and treatment would have little impact on our main results regarding the location, feature, and severity of the pain reported during MRgFUS.

We acknowledge the limitations of this study. First, the study was retrospective and conducted at a single center; thus, the applicability of our findings may be limited to our patient population and treatment protocol. The pain management during MRgFUS was not standardized for each patient. Second, the CRST at 6 months was not available in 7 patients. However, because this was mainly due to the COVID-19 pandemic and unrelated to treatment, it likely had little impact on the results. Third, we used the SF-MPQ2 for pain characterization; however, the Japanese version of this questionnaire has been validated for chronic but not acute pain. Therefore, some items require careful interpretation. For example, many patients categorized their pain as “burning” because they felt hotness during MRgFUS. Although “burning” is a neuropathic item in the SF-MPQ2, the neuropathic subscale had a low score in our study. Classifying sonication-related head pain according to current diagnostic criteria is challenging. Forth, we included 2 groups in which the time of pain questionnaire was different. The “treatment time” or “maximum power” was smaller in the more recently treated group, presumably because we shifted to select the higher SDR as a threshold for treatment. However, there was no difference in NRS and SF-MPQ2 scores between the 2 groups, and many patients indicated that the pain during MRgFUS was unique and hard to forget. Therefore, we consider that the duration between administrations of the questionnaire and treatment would have little impact on our main results regarding the location, feature, and severity of the pain reported during MRgFUS.

To our knowledge, this is the first report of the characteristics of sonication-related head pain during MRgFUS. Most of the patients in our cohort experienced pain during MRgFUS, and pain intensity affected the treatment outcome. The distribution and intensity of pain varied according to the SDR, indicating that the pain had different origins. We hope that the findings of this study will help improve pain management during MRgFUS.