Division Specific Biologically Effective Dose Predicts Gamma Knife Radiosurgery Outcomes in Multiple Sclerosis–Associated Trigeminal Neuralgia

Abstract

BACKGROUND AND OBJECTIVES:

Outcomes after Gamma Knife surgery (GKS) for multiple sclerosis–related trigeminal neuralgia type 1 (MS-TN1) remain variable, and the role of biologically effective dose (BED) distribution across trigeminal nerve subregions is not well defined. This study evaluated whether BED-based volumetric parameters within the trigeminal root, the responsible division (RD), the root entry zone (REZ), and the adjacent brainstem predict pain relief, recurrence, and sensory toxicity after GKS.

METHODS:

We retrospectively analyzed data from 248 patients with medically refractory MS-TN1 who were treated with GKS between 2012 and 2024. Three-dimensional BED-volume histograms were generated for the trigeminal root, RD, REZ, and brainstem. Dosimetric variables included maximum BED and percentage volume receiving ≥1000 Gy₂.₄₇. Outcomes included early pain relief, defined as Barrow Neurological Institute Class I-III at 1 month, pain recurrence, and facial hypoesthesia. Multivariable logistic regression and Cox proportional hazards models identified independent predictors.

RESULTS:

Early pain relief occurred in 156 patients (62.8%). Among responders, 65 (41.7%) experienced recurrence. Facial hypoesthesia developed in 25 patients (9.9%). Higher trigeminal root V%BED1000 was independently associated with improved early pain relief and lower recurrence risk. Greater BED coverage of the RD correlated with more durable pain control, whereas elevated brainstem maximum BED combined with increased REZ V%BED1000 predicted facial hypoesthesia.

CONCLUSION:

Regional BED distribution influences outcomes after GKS for MS-TN1. Greater division-specific BED coverage was associated with more durable pain relief, whereas excessive exposure to the REZ and brainstem increases sensory toxicity. These findings suggest that anatomy-guided, BED-optimized planning may improve patient-specific radiosurgical targeting.

ABBREVIATIONS:

BED

biologically effective dose

BNI

Barrow Neurological Institute

GKS

Gamma Knife surgery

MS

multiple sclerosis

RD

responsible division

REZ

root entry zone

S1

first section of the trigeminal root (proximal, near REZ)

S2

second section of the trigeminal root (intermediate)

S3

third section of the trigeminal root (distal, retrogasserian zone)

V1

ophthalmic nerve

V2

maxillary nerve

V3

mandibular nerve.

Trigeminal neuralgia (TN), particularly type 1 (TN1), is a debilitating neuropathic facial pain syndrome characterized by paroxysmal, electric shock–like attacks involving 1 or more trigeminal divisions and markedly impairing quality of life. In classical TN1, the underlying mechanism is typically neurovascular compression of the trigeminal root, leading to focal demyelination and abnormal nociceptive transmission.^1,2

In multiple sclerosis (MS)-associated TN, demyelinating lesions may affect the trigeminal root entry zone (REZ) and central brainstem pathways, promoting axonal hyperexcitability, ectopic impulse generation, and abnormal nociceptive transmission. Neurovascular compression may coexist and contribute to treatment heterogeneity.³

Gamma Knife surgery (GKS) is an established minimally invasive treatment of medically refractory TN1, but outcomes in MS remain heterogeneous. A recent meta-analysis reported initial pain freedom in 32% (95% CI, 13%-55%), adequate pain relief in 74% (95% CI, 50%-92%), and sustained relief in 65%, with response latency ranging from days to months.¹ Conventional prescription dose and isocenter geometry have shown inconsistent associations with outcomes, whereas 2-isocenter strategies, larger shot size, and higher brainstem dose may improve efficacy at the cost of sensory toxicity.⁴

Despite widespread GKRS use for MS-associated TN, planning remains largely geometric rather than biologically informed. The trigeminal system has somatotopic organization across its divisions.⁵ Adequate coverage of the pain-responsible division may improve efficacy, whereas excess REZ and brainstem exposure may increase toxicity.⁶ Biologically effective dose (BED), which integrates dose and tissue radiobiology, predicts pain relief and sensory dysfunction after GKS, but its 3-dimensional (3D) regional distribution remains insufficiently defined.⁷ Building on previously described division- and section-based BED approaches in idiopathic TN1, the present study evaluates this framework specifically in patients with MS-associated TN1, a clinically and biologically distinct population.

To address this gap, we analyzed data from 248 patients with MS and refractory TN1 who were treated with GKS across a multihospital community healthcare system. We generated 3D BED-volume histograms for the trigeminal root, responsible division [RD], S1, and brainstem, using maximum BED (DMaxBED) and the percentage of volume receiving ≥1000 Gy₂.₄₇ (V%BED1000). We aimed to determine whether these regional BED metrics predict early pain relief, recurrence, and facial hypoesthesia, and thereby identify BED-based correlates of efficacy and toxicity that may refine radiosurgical targeting in MS-related TN1.

METHODS

Study Design and Patient Selection

We performed a retrospective cohort study following Strengthening the Reporting of Observational Studies in Epidemiology guidelines. Consecutive patients with medically refractory MS-associated TN1 who underwent GKS from January 1, 2012, through December 31, 2024, were identified across a network of 3 tertiary hospitals and 3 neurosurgical referral centers under a unified institutional review board. The objective was to determine whether anatomically localized BED distributions along the trigeminal root predict pain relief, recurrence-free survival, and sensory dysfunction.

Inclusion criteria were unilateral MS-associated TN1, age younger than 70 years, and ≥36 months of clinical follow-up. Diagnosis was based on neurological examination, high-resolution MRI, and the Barrow Neurological Institute (BNI) pain scale. Medically refractory disease was defined as persistent pain despite optimized pharmacotherapy or treatment intolerance. Exclusion criteria were bilateral TN, TN type 2, prior TN surgery, unconfirmed MS, or age 70 years and older. Older patients were excluded because MS-related frailty and disability could confound longitudinal pain and sensory assessments. Figure 1 summarizes patient selection.

FIGURE 1.

Patient selection flow diagram. Flow diagram illustrating the identification, screening, and inclusion of patients with multiple sclerosis–associated TN treated with Gamma Knife surgery. Of 312 records screened, 64 patients were excluded because of age 70 years and older (n = 28), previous TN surgery (n = 22), or TN type 2 (n = 14). A total of 248 patients met eligibility criteria and were included in the final analysis. TN, trigeminal neuralgia.

The institutional review board approved the study and waived the requirement for informed consent because of the retrospective design and the use of deidentified data. All study procedures were conducted in accordance with the ethical principles of the Declaration of Helsinki.

Radiosurgical Technique

All patients underwent frame-based stereotactic GKS using the Leksell Gamma Knife Perfexion system (Elekta AB). Treatment planning was performed using high-resolution stereotactic MRI, including contrast-enhanced T1-weighted and constructive interference in steady state sequences, with stereotactic computed tomography fusion when necessary. Treatment planning and anatomic segmentation were performed using conventional stereotactic MRI with standard multiplanar reconstruction and did not require advanced image-processing techniques.

The cisternal segment of the trigeminal nerve, typically located 2 to 4 mm anterior to the brainstem, was targeted while minimizing radiation exposure to adjacent critical structures. Prescription doses ranged from 80 to 90 Gy, with a maximum institutional brainstem dose constraint of ≤12 Gy. Dose selection and isocenter number were individualized according to nerve anatomy, target visualization, proximity to the brainstem, and anticipated sensory risk.

A maximum dose of 90 Gy was generally favored when the cisternal trigeminal segment was well visualized and adequately separated from the brainstem. Conversely, 80 Gy was preferred in patients with short cisternal segments, limited target visualization, or increased concern for sensory morbidity. A single 4-mm isocenter was used in most patients. By contrast, a second isocenter was selectively used to improve conformity to the symptomatic trigeminal division without excessive irradiation of the brainstem or REZ.

Patients were discharged the same day and followed at 1 month, 6 months, and annually thereafter to assess pain outcomes and treatment-related adverse effects.

BED Analysis and Anatomical Segmentation

3D BED analyses were performed using MATLAB R2020a (MathWorks). The trigeminal nerve, adjacent vasculature, and brainstem were manually contoured on fused stereotactic MRI according to established anatomic guidelines.^8,9 BED distributions were calculated using the linear-quadratic model assuming an α/β ratio of 2.47 for late-responding nervous tissue.

BED-volume histograms were generated for the entire trigeminal root, the symptomatic trigeminal division (RD), the REZ-adjacent segment (S1), and the adjacent brainstem. Prespecified dosimetric metrics included the maximum BED (DMaxBED), the absolute volume receiving ≥1000 Gy₂.₄₇ (VBED1000), and the percentage volume receiving ≥1000 Gy₂.₄₇ (V%BED1000). The threshold of 1000 Gy₂.₄₇ was selected a priori based on previously reported associations with radiation-induced neural toxicity and its biological relevance in radiosurgical dose response studies.

Division-specific segmentation was performed using conventional high-resolution 3D constructive interference in steady state MRI with multiplanar reconstruction. The ophthalmic (V1), maxillary (V2), and mandibular (V3) divisions were initially identified at the trigeminal ganglion, where the individual fascicles are anatomically distinguishable. The anatomic cleavage planes separating these divisions were then manually propagated proximally along the cisternal trigeminal root using the nerve centerline as the reference axis, thereby generating patient-specific, division-specific regions of interest throughout the visible trigeminal root, as previously described.¹⁰

This workflow enabled standardized, clinically translatable spatial BED analysis using conventional stereotactic MRI without artificial intelligence, tractography, or specialized processing software. Two fellowship-trained neurosurgeons, 1 fellowship-trained neuroradiologist, and 1 medical physicist independently segmented the anatomy, followed by consensus review. The symptomatic division was designated the RD. The visible root was subdivided into proximal (S1), intermediate (S2), and distal retrogasserian (S3) segments using predefined centerline landmarks (Figures 2-4). Final regions of interest were imported into MATLAB for voxel-wise BED analysis. RD V%BED1000 thresholds of 50% and 60% and RD DMaxBED of 80 Gy₂.₄₇ were prespecified from published literature, biological plausibility, and institutional protocols.

FIGURE 2.

Treatment planning and 3D BED distribution during single-isocenter Gamma Knife surgery for multiple sclerosis–associated trigeminal neuralgia. A, Right lateral 3D reconstruction (head-up orientation) demonstrating the neurovascular anatomy used for stereotactic treatment planning. Arrows identify the cisternal trigeminal nerve (CN V, dark yellow), brainstem (pink), SCA (red), and adjacent vein (blue), illustrating the neurovascular relationships that guide radiosurgical target selection. B, Axial high-resolution MR TOF image fused with stereotactic computed tomography. The 70% prescription isodose line (light yellow contour) encompasses the cisternal trigeminal nerve (CN V, dark yellow) while minimizing overlap with the adjacent SCA (red). The 15-Gy isodose line (green contour) remains tangential to the pontine surface, demonstrating preservation of the institutional brainstem dose constraint. C, Right lateral 3D reconstruction (head-down orientation) demonstrating placement of the single 4-mm collimator shot (cyan) centered on the cisternal trigeminal root. Arrows identify the trigeminal nerve, brainstem, superior cerebellar artery, and adjacent vein, illustrating the anatomic landmarks used for radiosurgical target selection. D, 3D BED reconstruction demonstrating iso-BED surfaces of 40 Gy₂.₄₇ (red), 200 Gy₂.₄₇ (cyan), and 1000 Gy₂.₄₇ (green) surrounding the trigeminal root. The reconstruction illustrates the conformal BED distribution achieved by positioning the 20% prescription isodose line tangential to the pontine surface, maximizing trigeminal nerve coverage while minimizing brainstem irradiation. 3D, 3-dimensional; BED, biologically effective dose; CN V, cranial nerve V; GKS, Gamma Knife surgery; MR, magnetic resonance; NFT, nerve fiber tractography; SCA, superior cerebellar artery; TOF, time of flight.

FIGURE 4.

BED distribution across 3 anatomic sections of the trigeminal nerve (CN V). A, The visible cisternal trigeminal root was defined from the point of nerve emergence at the anterolateral pons to the entrance of the nerve into Meckel's cave. A centerline was generated along the long axis of the nerve, and perpendicular planes divided this axis into 3 equal anteroposterior subsegments: S1, the proximal REZ-adjacent segment; S2, the intermediate cisternal segment; and S3, the distal retrogasserian segment near Meckel's cave, as indicated by green arrows. B, BED-volume histogram for S1. C, BED-volume histogram for S2. D, BED–volume histogram for S3. BED, biologically effective dose; CN, cranial nerve; REZ, root entry zone; S1, proximal REZ-adjacent segment; S2, intermediate cisternal segment; S3, distal retrogasserian segment.

FIGURE 3.

Division-level segmentation and BED distribution across the trigeminal nerve (CN V). A, Division-specific segmentation of the trigeminal nerve was performed using conventional high-resolution CISS MRI with multiplanar reconstruction. The ophthalmic (V1), maxillary (V2), and mandibular (V3) branches were first identified at the trigeminal ganglion (green arrows). The anatomic cleavage planes separating each branch (red arrows) were then manually extended proximally along the cisternal trigeminal root using the nerve centerline as the reference axis, allowing delineation of the corresponding divisions within the visible trigeminal root. The resulting regions of interest were subsequently used for biologically effective dose (BED) analysis. B, BED-volume histogram of the ophthalmic division (V1). C, BED-volume histogram of the maxillary division (V2). D, BED-volume histogram of the mandibular division (V3). BED, biologically effective dose; CISS, constructive interference in steady state; CN, cranial nerve; V1, ophthalmic division; V2, maxillary division; V3, mandibular division.

Outcome Measures and Statistical Analysis

The primary outcome was early pain relief, defined as achieving a BNI Pain Intensity Scale score of Class I-III within 1 month after GKS. Secondary outcomes included pain recurrence and new or worsened facial hypoesthesia. Pain recurrence was defined as deterioration to BNI Class IV-V after an initial treatment response, the need for repeat procedural intervention, or escalation of pharmacological therapy. Facial hypoesthesia was assessed using the BNI Facial Numbness Scale. Clinical outcomes were assessed during scheduled in-person follow-up visits and supplemented by standardized telephone interviews when necessary.

Associations between localized BED metrics and outcomes were evaluated using multivariable logistic and Cox proportional hazards regression. Recurrence analyses included only initial BNI Class I-III responders; recurrence-free survival was estimated by Kaplan-Meier analysis and compared with the log-rank test. Prespecified analyses assessed division and segment-specific BED distributions. Missing data (<2%) were handled by complete-case analysis. Covariates were selected a priori, and models adjusted for age, sex, symptomatic division, symptom duration, prescription dose, and isocenter number. Linearity, multicollinearity, and proportional hazards assumptions were assessed. Two-sided P < .05 indicated significance. Analyses used SPSS 27.0 and R 4.3.2. All patient imaging, anatomical segmentations, BED calculations, statistical analyses, and figures presented in this study were independently derived from the current MS-associated TN1 cohort identified through our Gamma Knife registry.

RESULTS

Patient Demographics and Baseline Characteristics

A total of 248 patients with medically refractory, MS-associated TN1 were included. The mean age was 58.7 ± 10.2 years, and 59.5% were women. Pain was right-sided in 54.0% of cases. The maxillary division (V2) was most commonly affected (78.9%), followed by V3 (17.2%) and V1 (3.9%). Mean symptom duration before treatment was 6.4 ± 3.8 years, with a median follow-up of 42 months.

Pain Relief Outcomes

At 1 month, 156 patients (62.8%) achieved early pain relief corresponding to BNI Class I-III outcomes, whereas 92 patients (37.2%) remained within BNI Class IV-V. Among responders, 65 patients (41.7%) subsequently experienced pain recurrence during follow-up, with a median recurrence-free interval of 270 days (IQR, 180-360). At last follow-up, durable BNI Class I-III pain control remained present in 91 patients (36.7% of the total cohort). Greater regional BED coverage was independently associated with a reduced risk of recurrence (hazard ratio [HR], 0.71; 95% CI, 0.54-0.93; P = .02).

Facial Hypoesthesia

Facial hypoesthesia (≥ BNI Grade III) occurred in 25 patients (9.9%), with a median onset of 180 days. Most sensory changes involved the V2 and V3 territories. Higher brainstem DMaxBED combined with elevated S1 V%BED1000 independently predicted hypoesthesia (odds ratio [OR], 1.84; 95% CI, 1.12-3.04; P = .01).

Associations Between BED Parameters and Outcomes

RD V%BED1000 ≥ 60% was independently associated with higher rates of early pain relief (OR, 2.15; 95% CI, 1.45-3.22; P < .001) and longer recurrence-free survival. By contrast, increased BED exposure to the REZ-adjacent segment (S1) correlated with higher rates of sensory complications (P = .01) (Table 1). These findings suggest that greater division-specific BED coverage may improve pain durability, whereas excessive exposure to the proximal trigeminal root and adjacent brainstem may increase treatment-related toxicity (Table 1).

TABLE 1.

Patient Demographics and Clinical Characteristics (n = 248)

Characteristic	Value
Total patients	248
Mean age	58.7 ± 10.2 y
Sex
Female	148 (59.5%)
Male	100 (40.5%)
Side affected
Right	134 (54.0%)
Left	114 (46.0%)
Division affected
V1 (Ophthalmic)	10 (3.9%)
V2 (Maxillary)	196 (78.9%)
V3 (Mandibular)	42 (17.2%)
Mean duration of symptoms	6.4 ± 3.8 y
Pain type (McGill Pain Questionnaire)
Sharp, shooting, electrical shock-like	244 (98.5%)
Episodic pain with pain-free intervals	248 (100%)
Refractory trigeminal neuralgia definition
Medical treatment ineffective in pain relief	248 (100%)
Medication toxicity unmanageable	57 (22.8%)
Pre-Gamma Knife surgery treatments
Medications (e.g., carbamazepine, gabapentin)	239 (96.4%)
Previous surgical treatments (e.g., microvascular decompression, percutaneous rhizotomy)	0 (7.7%)
Previous radiation therapy	10 (4.0%)
Mean follow-up time	4.5 ± 1.2 y

Brainstem and S1 BED Metrics

Higher brainstem DMaxBED and S1 V%BED1000 were independently associated with facial hypoesthesia (OR, 1.72; 95% CI, 1.05-2.80; P = .03; and OR, 1.95; 95% CI, 1.12-3.40; P = .02, respectively).

Division-Level Analysis

Greater division-specific BED coverage was associated with improved pain control and longer recurrence-free survival (OR, 2.21; 95% CI, 1.48-3.30; P < .001). An RD DMaxBED ≥80 Gy₂.₄₇ was associated with greater early pain relief (68.9% vs 54.3%, P < .001) and lower recurrence (HR, 0.67; 95% CI, 0.48-0.92; P = .02). Similarly, RD V%BED1000 ≥ 50% was associated with improved pain relief and longer recurrence-free survival (P < .01).

Complications and Adverse Effects

The overall complication rate was 9.9%, with facial hypoesthesia representing the most common treatment-related adverse effect. Sensory dysfunction was generally mild and predominantly involved the V2 and V3 distributions, without significant functional impairment in most patients. Increased brainstem DMaxBED and elevated S1 V%BED1000 were independently associated with a higher likelihood of post-treatment sensory dysfunction (P < .05), supporting the importance of minimizing unnecessary irradiation of the REZ-adjacent segment and pontine surface during treatment planning. Importantly, no Grade IV or higher neurological complications, anesthesia dolorosa, brainstem injury, or treatment-related mortality were observed. Overall, these findings suggest that GKRS maintains a favorable safety profile in patients with MS-associated TN when anatomically conformal BED-guided targeting strategies are used.

Multivariable Analysis

In multivariable models adjusted for age, sex, symptom duration, treated trigeminal division, and isocenter number, trigeminal root V%BED1000 and RD DMaxBED remained independent predictors after adjustment for prespecified clinical covariates (P < .001). Greater division-specific BED coverage was also associated with improved durability of response and prolonged recurrence-free survival on Cox proportional hazards analysis (HR, 0.71; 95% CI, 0.55-0.90; P < .01). These findings suggest that regional BED distribution within the symptomatic trigeminal division may influence radiosurgical efficacy more strongly than prescription dose alone, supporting the role of anatomy-guided BED optimization in GKRS planning for MS-associated TN. Overall, the multivariable models demonstrated independent associations between regional BED metrics and clinical outcomes after adjustment for prespecified clinical covariates (Table 2).

TABLE 2.

Association of BED Parameters and Clinical Outcomes in Patients With MS-Associated TN1 Treated With GKRS

Variable	Category	Odds ratio (OR)	95% CI	P-value
V%BED1000 of trigeminal root	≥60% vs <60%	2.45	1.85-3.30	<.001
DMaxBED of RD	≥80 Gy2.47 vs < 80 Gy2.47	1.92	1.42-2.61	<.001
V%BED1000 of RD	≥50% vs <50%	1.63	1.20-2.22	.002
DMaxBED of brainstem	≥60 Gy2.47 vs < 60 Gy2.47	0.74	0.51-1.09	.13
V%BED1000 of root entry zone (S1)	≥50% vs <50%	0.78	0.57-1.07	.13
Follow-up duration (y)	≥4 y vs < 4 y	1.32	1.12-1.56	.001
Age	≥60 y vs < 60 y	1.10	0.89-1.35	.39
Sex	Female vs male	1.18	0.89-1.57	.25

BED, biologically effective dose; GKRS, Gamma Knife radiosurgery; OR, odds ratio; MS-TN, multiple sclerosis–associated trigeminal neuralgia; RD, responsible division.

DISCUSSION

Gamma Knife Radiosurgery Outcomes in MS-Related TN

In this cohort, GKRS achieved early pain relief in 62.8% of patients, consistent with reported rates of 60%-75%.^8,10-12 Among responders, 41.7% recurred after a median of 270 days; this responder-based estimate aligns with the limited durability reported across series, although follow-up and outcome definitions vary.^10,13 Facial hypoesthesia occurred in 9.9%, within the reported 10%-35% range, and was generally mild.¹¹ These findings support GKRS as an effective, relatively safe option for MS-TN, particularly when invasive surgery is unsuitable.

BED-Guided Anatomical Targeting

The principal finding of this study is that the spatial distribution of BED, rather than prescription dose alone, was independently associated with radiosurgical outcomes. Greater BED coverage of the trigeminal root, particularly within the symptomatic trigeminal division, was independently associated with improved early pain relief and lower recurrence rates. Specifically, RD DMaxBED ≥80 Gy₂.₄₇ and RD V%BED1000 ≥ 50% were associated with more durable pain control, whereas increased BED exposure to the REZ-adjacent segment and brainstem correlated with higher rates of sensory dysfunction. These findings suggest that the spatial distribution of BED relative to trigeminal functional anatomy may better explain variability in radiosurgical outcomes than conventional geometric planning parameters alone.^6,7

From a practical radiosurgical perspective, these results support conformal coverage of the symptomatic trigeminal division while minimizing unnecessary irradiation of the REZ and adjacent brainstem. When anatomically feasible, the 4-mm isocenter may be preferentially positioned along the cisternal segment to maximize BED deposition within the symptomatic division while maintaining rapid dose fall-off toward critical structures. In patients with short cisternal nerve segments, limited nerve visualization, or proximity to the pons, dose reduction from 90 Gy to 80 Gy or more conservative target placement may reduce sensory morbidity while preserving institutional brainstem constraints. Conversely, when the cisternal segment is clearly visualized and adequately separated from the brainstem, higher conformal BED coverage may be achieved without substantially increasing pontine exposure.

Distal cisternal targeting may optimize division-specific coverage while limiting REZ and brainstem irradiation. Proximal REZ targeting may improve pain relief but increases facial numbness and sensory complications.^14-17 Distal retrogasserian targeting may reduce toxicity with variable efficacy.^15,16,18,19 Table 3 summarizes the analyzed regions, BED metrics, and clinical relevance. Division-level mapping showed that greater symptomatic-division coverage was associated with durable pain control, whereas greater REZ and brainstem exposure correlated with sensory dysfunction (Table 4). These findings favor alignment of isocenter placement with trigeminal topography rather than increasing irradiated nerve length.^7,14

TABLE 3.

Three-Dimensional BED Analysis Framework

Region of interest (ROI)	Definition/boundaries	Parameters measured	Clinical relevance
Trigeminal root (CN V)	Entire cisternal segment within prescription isodose	DMaxBED, VBED1000, V%BED1000	Global dose coverage and predictor of pain relief/recurrence
Responsible division	Division (V1, V2, or V3) corresponding to patient's symptomatic distribution	DMaxBED, VBED1000, V%BED1000	Division-specific targeting; predictors of durable pain control
S1 (root entry zone)	Proximal segment adjacent to brainstem	DMaxBED, VBED1000, V%BED1000	Predicts sensory complications (e.g., hypoesthesia)
S2 (intermediate)	Mid-cisternal segment	DMaxBED, VBED1000, V%BED1000	Transitional segment; potential balance of efficacy/toxicity
S3 (distal/retrogasserian)	Distal segment near Meckel's cave	DMaxBED, VBED1000, V%BED1000	Target for maximizing efficacy while sparing REZ
Brainstem	Adjacent tissue volume at REZ	DMaxBED, V%BED1000	Excess exposure predicts adverse sensory outcomes
VBED1000 threshold	Volume receiving ≥1000 Gy2.47	Absolute (mm3) and relative (%) values	Reflects biologically relevant iso-BED associated with neural toxicity

BED, biologically effective dose; CN, cranial nerve.

TABLE 4.

Anatomical Targeting Strategies in Gamma Knife Radiosurgery for MS-Associated TN

Target location	Anatomical definition	Efficacy (pain relief)	Risk profile (complications)	Clinical considerations
Proximal/REZ (S1)7,15	Root entry zone, 2–3 mm from brainstem surface	High initial efficacy; rapid pain relief	Higher rates of facial hypoesthesia, corneal anesthesia, and bothersome numbness	Historically common target; may be justified in refractory cases but carries sensory morbidity
Intermediate (S2)6,15	Mid-cisternal segment between REZ and retrogasserian zone	Moderate-to-high efficacy	Balanced risk; potential compromise between relief and toxicity	Suitable when fiber organization suggests mixed division involvement
Distal/Retrogasserian (S3)7,15	Distal cisternal segment near Meckel's cave	Durable efficacy when RD coverage achieved (≥50% V%BED1000, ≥80 Gy2.47 DMaxBED)	Lower incidence of sensory complications	Favored in modern planning to maximize RD coverage while sparing REZ and brainstem
Division-Specific (RD)5,6,14,16	Pain-generating fiber group (V1, V2, or V3)	Quantitative thresholds strongly predict outcomes (RD DMaxBED ≥80 Gy2.47, RD V%BED1000 ≥ 50%)	Risk rises if spillover to S1/brainstem	Supports biologically and anatomically informed targeting; requires precise imaging and contouring

BED, biologically effective dose; MS-TN, multiple sclerosis–associated trigeminal neuralgia; RD, responsible division; REZ, root entry zone; S1, proximal root entry zone; S2, intermediate cisternal segment; S3, distal retrogasserian segment.

These data do not support routine expansion of treatment length or indiscriminate multiple-isocenter use. In a randomized study, 2 isocenters increased irradiated nerve length without improving pain outcomes and increased sensory morbidity.²⁰ Accordingly, 2-isocenter planning was reserved for selected anatomies in which division conformality could be improved without excessive REZ or brainstem exposure (Figures 2-4). The findings support selective, anatomy-guided BED optimization.

Comparison With Alternative Treatments

GKRS remains an important minimally invasive treatment option between pharmacotherapy and microsurgical intervention.²¹ Given the substantial disability and impaired quality of life associated with MS-related TN, improving the durability of pain control while minimizing sensory morbidity remains clinically critical. Although first-line therapy with carbamazepine or oxcarbazepine is often initially effective, long-term tolerability is limited.²² Microvascular decompression remains the preferred surgical treatment of classical TN because of durable outcomes, but seems less effective in MS-associated TN, where central demyelination predominates.^23,24 In this setting, GKRS offers a minimally invasive, repeatable option with a favorable risk–benefit profile, particularly for patients who are poor candidates for microvascular decompression.^8,25 Our findings further suggest that BED-guided, division-specific targeting may help refine radiosurgical planning by optimizing pain control while minimizing sensory morbidity.

Radiosurgical Outcomes and Limitations in Multiple Sclerosis Associated TN

Pain relief after radiosurgery for MS-associated TN is generally less durable than that observed in patients with classical TN.⁸ One retrospective study reported meaningful pain relief in 27.7% of patients with MS-associated TN compared with 57.7% in matched non-MS cohorts.²⁶ Although initial response rates after Gamma Knife radiosurgery may be favorable, long-term durability remains limited, with recurrence rates approaching 78.7% and median pain-free intervals ranging from 10 to 12.7 months across published series.^23-25 These differences likely reflect the complex and multifactorial pathophysiology of MS-associated TN, in which central demyelination and brainstem involvement may contribute to persistent nociceptive dysfunction beyond focal trigeminal root pathology.

Expanded target coverage and selective double-isocenter planning have been explored to improve durability without substantially increasing brainstem dose.²⁶ Radiosurgery nevertheless provides meaningful pain control with low morbidity and remains an important minimally invasive option for refractory MS-TN.^21,27 Our findings suggest that efficacy depends less on nominal dose than on the spatial relationship between BED and trigeminal functional anatomy (Figure 5). Anatomy-guided, division-specific BED optimization may therefore refine planning, but requires prospective multicenter validation.

FIGURE 5.

Anatomy-guided, BED-optimized radiosurgical targeting strategy for MS-associated TN. Conceptual schematic illustrating anatomy-guided, division-specific BED optimization during GKRS for MS-TN. Optimal radiosurgical planning aims to maximize conformal BED coverage of the symptomatic trigeminal division while minimizing unnecessary irradiation of the REZ and adjacent brainstem. Distal cisternal targeting may improve the durability of pain while reducing sensory morbidity. Excessive BED exposure to the REZ-adjacent segment and brainstem is associated with a higher risk of facial hypoesthesia. By contrast, inadequate coverage of the responsible division may contribute to pain recurrence. Collectively, the schematic summarizes the proposed relationship between trigeminal functional anatomy, regional BED distribution, treatment efficacy, and radiosurgical toxicity. BED, biologically effective dose; GKRS, Gamma Knife radiosurgery; MS-TN, multiple sclerosis–associated trigeminal neuralgia; REZ, root entry zone.

Limitations

This study is limited by its retrospective design within a single integrated healthcare system, which may have introduced selection bias and unmeasured confounding. Exclusion of patients aged 70 years and older may limit generalizability to older patients with multiple sclerosis–associated TN. Improvements in imaging and radiosurgical planning over the study period may have introduced temporal heterogeneity. Although anatomic segmentation was standardized and performed with multidisciplinary consensus review, formal interobserver reproducibility was not assessed. Finally, the absence of external validation, patient-reported outcomes, and longer follow-up limits the assessment of late recurrence, delayed toxicity, and broader generalizability. Prospective multicenter studies are warranted to validate these findings.

CONCLUSION

Gamma Knife radiosurgery provided effective pain relief with an acceptable safety profile in patients with multiple sclerosis–associated TN. Greater BED coverage of the symptomatic trigeminal division was associated with improved pain control and longer recurrence-free survival, whereas increased BED exposure to the REZ and adjacent brainstem correlated with a higher risk of sensory dysfunction. These findings suggest that anatomy-guided, division-specific BED optimization may inform patient-specific radiosurgical planning by balancing treatment efficacy and toxicity. Prospective multicenter studies are warranted to validate this approach before routine clinical implementation.

Funding

This study did not receive any funding or financial support.

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.