The relevance of biologically effective dose for pain relief and sensory dysfunction after Gamma Knife radiosurgery for trigeminal neuralgia: an 871-patient multicenter study

Ronald E. Warnick Gamma Knife Center, Jewish Hospital, Mayfield Clinic, Cincinnati, Ohio;

Search for other papers by Ronald E. Warnick in
Current site
Google Scholar
PubMed
Close
 MD
,
Ian Paddick Queen Square Radiosurgery Centre, London, United Kingdom;

Search for other papers by Ian Paddick in
Current site
Google Scholar
PubMed
Close
 MSc
,
David Mathieu Division of Neurosurgery, University of Sherbrooke, CHUS Research Center, Sherbrooke, Québec, Canada;

Search for other papers by David Mathieu in
Current site
Google Scholar
PubMed
Close
 MD
,
Elizabeth Adam Division of Neurosurgery, University of Sherbrooke, CHUS Research Center, Sherbrooke, Québec, Canada;

Search for other papers by Elizabeth Adam in
Current site
Google Scholar
PubMed
Close
 MD
,
Christian Iorio-Morin Division of Neurosurgery, University of Sherbrooke, CHUS Research Center, Sherbrooke, Québec, Canada;

Search for other papers by Christian Iorio-Morin in
Current site
Google Scholar
PubMed
Close
 MD, PhD
,
William Leduc Division of Neurosurgery, University of Sherbrooke, CHUS Research Center, Sherbrooke, Québec, Canada;

Search for other papers by William Leduc in
Current site
Google Scholar
PubMed
Close
 MD
,
Andréanne Hamel Division of Neurosurgery, University of Sherbrooke, CHUS Research Center, Sherbrooke, Québec, Canada;

Search for other papers by Andréanne Hamel in
Current site
Google Scholar
PubMed
Close
 BSc
,
Sarah E. Johnson Neuro-Informatics Laboratory, Mayo Clinic, Rochester, Minnesota;

Search for other papers by Sarah E. Johnson in
Current site
Google Scholar
PubMed
Close
 MBBS
,
Mohamad Bydon Neuro-Informatics Laboratory, Mayo Clinic, Rochester, Minnesota;

Search for other papers by Mohamad Bydon in
Current site
Google Scholar
PubMed
Close
 MD
,
Ajay Niranjan Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;

Search for other papers by Ajay Niranjan in
Current site
Google Scholar
PubMed
Close
 MD, MBA
,
L. Dade Lunsford Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;

Search for other papers by L. Dade Lunsford in
Current site
Google Scholar
PubMed
Close
 MD
,
Zhishuo Wei Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;

Search for other papers by Zhishuo Wei in
Current site
Google Scholar
PubMed
Close
 MSc
,
Kaitlin Waite Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;

Search for other papers by Kaitlin Waite in
Current site
Google Scholar
PubMed
Close
 MSc
,
Shalini Jose Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;

Search for other papers by Shalini Jose in
Current site
Google Scholar
PubMed
Close
 MS
,
Selcuk Peker Department of Neurosurgery, Koç University School of Medicine, Istanbul, Turkey;

Search for other papers by Selcuk Peker in
Current site
Google Scholar
PubMed
Close
 MD
,
Mustafa Yavuz Samanci Department of Neurosurgery, Koç University School of Medicine, Istanbul, Turkey;

Search for other papers by Mustafa Yavuz Samanci in
Current site
Google Scholar
PubMed
Close
 MD
,
Ece Tek Department of Radiation Oncology, Acıbadem Altunizade Hospital, Istanbul, Turkey;

Search for other papers by Ece Tek in
Current site
Google Scholar
PubMed
Close
 MSc
,
Georgios Mantziaris Department of Neurological Surgery, University of Virginia Health System, Charlottesville, Virginia;

Search for other papers by Georgios Mantziaris in
Current site
Google Scholar
PubMed
Close
 MD
,
Stylianos Pikis Department of Neurological Surgery, University of Virginia Health System, Charlottesville, Virginia;

Search for other papers by Stylianos Pikis in
Current site
Google Scholar
PubMed
Close
 MD
,
Jason P. Sheehan Department of Neurological Surgery, University of Virginia Health System, Charlottesville, Virginia;

Search for other papers by Jason P. Sheehan in
Current site
Google Scholar
PubMed
Close
 MD, PhD
,
Manjul Tripathi Departments of Neurosurgery and

Search for other papers by Manjul Tripathi in
Current site
Google Scholar
PubMed
Close
 MCh
,
Narendra Kumar Radiation Oncology, Postgraduate Institute of Medical Education and Research, Chandigarh, India;

Search for other papers by Narendra Kumar in
Current site
Google Scholar
PubMed
Close
 MD
,
Juan Diego Alzate Departments of Neurosurgery and

Search for other papers by Juan Diego Alzate in
Current site
Google Scholar
PubMed
Close
 MD
,
Kenneth Bernstein Radiation Oncology, New York University Langone Medical Center, New York, New York;

Search for other papers by Kenneth Bernstein in
Current site
Google Scholar
PubMed
Close
 MS
,
Peter Ahorukomeye Department of Neurological Surgery, Rose Ella Burkhart Brain Tumor and Neuro-Oncology Center, Neurological Institute, Cleveland Clinic, Cleveland, Ohio;

Search for other papers by Peter Ahorukomeye in
Current site
Google Scholar
PubMed
Close
 BS
,
Varun R. Kshettry Department of Neurological Surgery, Rose Ella Burkhart Brain Tumor and Neuro-Oncology Center, Neurological Institute, Cleveland Clinic, Cleveland, Ohio;

Search for other papers by Varun R. Kshettry in
Current site
Google Scholar
PubMed
Close
 MD
,
Herwin Speckter Dominican Gamma Knife Center and Radiology Department, CEDIMAT, Santo Domingo, Dominican Republic;

Search for other papers by Herwin Speckter in
Current site
Google Scholar
PubMed
Close
 MSc
,
Wenceslao Hernandez Dominican Gamma Knife Center and Radiology Department, CEDIMAT, Santo Domingo, Dominican Republic;

Search for other papers by Wenceslao Hernandez in
Current site
Google Scholar
PubMed
Close
 MD
,
Dušan Urgošík Department of Stereotactic and Radiation Neurosurgery, Na Homolce Hospital, Prague, Czech Republic;

Search for other papers by Dušan Urgošík in
Current site
Google Scholar
PubMed
Close
 MD, PhD
,
Roman Liščák Department of Stereotactic and Radiation Neurosurgery, Na Homolce Hospital, Prague, Czech Republic;

Search for other papers by Roman Liščák in
Current site
Google Scholar
PubMed
Close
 MD, PhD
,
Andrew I. Yang Department of Neurosurgery, University of Pennsylvania, Philadelphia, Pennsylvania;

Search for other papers by Andrew I. Yang in
Current site
Google Scholar
PubMed
Close
 MD, MS
,
John Y. K. Lee Department of Neurosurgery, University of Pennsylvania, Philadelphia, Pennsylvania;

Search for other papers by John Y. K. Lee in
Current site
Google Scholar
PubMed
Close
 MD
,
Samir Patel Division of Radiation Oncology, University of Alberta, Edmonton, Alberta, Canada; and

Search for other papers by Samir Patel in
Current site
Google Scholar
PubMed
Close
 MD
,
Dorian M. Kusyk Department of Neurosurgery, Allegheny Health Network, Pittsburgh, Pennsylvania

Search for other papers by Dorian M. Kusyk in
Current site
Google Scholar
PubMed
Close
 MD, MA
,
Matthew J. Shepard Department of Neurosurgery, Allegheny Health Network, Pittsburgh, Pennsylvania

Search for other papers by Matthew J. Shepard in
Current site
Google Scholar
PubMed
Close
 MD, MA
, and
Douglas Kondziolka Departments of Neurosurgery and

Search for other papers by Douglas Kondziolka in
Current site
Google Scholar
PubMed
Close
 MD, MSc
Open access

OBJECTIVE

Recent studies have suggested that biologically effective dose (BED) is an important correlate of pain relief and sensory dysfunction after Gamma Knife radiosurgery (GKRS) for trigeminal neuralgia (TN). The goal of this study was to determine if BED is superior to prescription dose in predicting outcomes in TN patients undergoing GKRS as a first procedure.

METHODS

This was a retrospective study of 871 patients with type 1 TN from 13 GKRS centers. Patient demographics, pain characteristics, treatment parameters, and outcomes were reviewed. BED was compared with prescription dose and other dosimetric factors for their predictive value.

RESULTS

The median age of the patients was 68 years, and 60% were female. Nearly 70% of patients experienced pain in the V2 and/or V3 dermatomes, predominantly on the right side (60%). Most patients had modified BNI Pain Intensity Scale grade IV or V pain (89.2%) and were taking 1 or 2 pain medications (74.1%). The median prescription dose was 80 Gy (range 62.5–95 Gy). The proximal trigeminal nerve was targeted in 77.9% of cases, and the median follow-up was 21 months (range 6–156 months). Initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa) was noted in 81.8% of evaluable patients at a median of 30 days. Of 709 patients who achieved initial pain relief, 42.3% experienced at least one pain recurrence after GKRS at a median of 44 months, with 49.0% of these patients undergoing a second procedure. New-onset facial numbness occurred in 25.3% of patients after a median of 8 months. Age ≥ 63 years was associated with a higher probability of both initial pain relief and maintaining pain relief. A distal target location was associated with a higher probability of initial and long-term pain relief, but also a higher incidence of sensory dysfunction. BED ≥ 2100 Gy2.47 was predictive of pain relief at 30 days and 1 year for the distal target, whereas physical dose ≥ 85 Gy was significant for the proximal target, but the restricted range of BED values in this subgroup could be a confounding factor. A maximum brainstem point dose ≥ 29.5 Gy was associated with a higher probability of bothersome facial numbness.

CONCLUSIONS

BED and physical dose were both predictive of pain relief and could be used as treatment planning goals for distal and proximal targets, respectively, while considering maximum brainstem point dose < 29.5 Gy as a potential constraint for bothersome numbness.

ABBREVIATIONS

AUC = area under the ROC curve; BED = biologically effective dose; BNI = Barrow Neurological Institute; GKRS = Gamma Knife radiosurgery; REZ = root entry zone; ROC = receiver operating characteristic; TN = trigeminal neuralgia.

OBJECTIVE

Recent studies have suggested that biologically effective dose (BED) is an important correlate of pain relief and sensory dysfunction after Gamma Knife radiosurgery (GKRS) for trigeminal neuralgia (TN). The goal of this study was to determine if BED is superior to prescription dose in predicting outcomes in TN patients undergoing GKRS as a first procedure.

METHODS

This was a retrospective study of 871 patients with type 1 TN from 13 GKRS centers. Patient demographics, pain characteristics, treatment parameters, and outcomes were reviewed. BED was compared with prescription dose and other dosimetric factors for their predictive value.

RESULTS

The median age of the patients was 68 years, and 60% were female. Nearly 70% of patients experienced pain in the V2 and/or V3 dermatomes, predominantly on the right side (60%). Most patients had modified BNI Pain Intensity Scale grade IV or V pain (89.2%) and were taking 1 or 2 pain medications (74.1%). The median prescription dose was 80 Gy (range 62.5–95 Gy). The proximal trigeminal nerve was targeted in 77.9% of cases, and the median follow-up was 21 months (range 6–156 months). Initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa) was noted in 81.8% of evaluable patients at a median of 30 days. Of 709 patients who achieved initial pain relief, 42.3% experienced at least one pain recurrence after GKRS at a median of 44 months, with 49.0% of these patients undergoing a second procedure. New-onset facial numbness occurred in 25.3% of patients after a median of 8 months. Age ≥ 63 years was associated with a higher probability of both initial pain relief and maintaining pain relief. A distal target location was associated with a higher probability of initial and long-term pain relief, but also a higher incidence of sensory dysfunction. BED ≥ 2100 Gy2.47 was predictive of pain relief at 30 days and 1 year for the distal target, whereas physical dose ≥ 85 Gy was significant for the proximal target, but the restricted range of BED values in this subgroup could be a confounding factor. A maximum brainstem point dose ≥ 29.5 Gy was associated with a higher probability of bothersome facial numbness.

CONCLUSIONS

BED and physical dose were both predictive of pain relief and could be used as treatment planning goals for distal and proximal targets, respectively, while considering maximum brainstem point dose < 29.5 Gy as a potential constraint for bothersome numbness.

In Brief

Is biologically effective dose (BED) superior to prescription dose in predicting pain relief and facial numbness in trigeminal neuralgia patients undergoing Gamma Knife radiosurgery as a first procedure? No. Instead, the authors found an interplay of factors: 1) a distal target may be optimal, 2) BED was predictive for the distal target, 3) the physical dose was predictive for the proximal target, and 4) the maximum brainstem dose was associated with sensory dysfunction.

Stereotactic radiosurgery is an effective treatment option for trigeminal neuralgia (TN). It is often used for patients when microvascular decompression is contraindicated because of advanced age and/or medical comorbidities, or absence of neurovascular compression. The International Stereotactic Radiosurgery Society practice guidelines highlight factors predictive of a higher probability of pain relief (no prior surgery, stereotactic radiosurgery within 3 years of pain onset, and single dose ≥ 80 Gy) and a higher incidence of facial numbness (> 1 isocenter, proximal nerve target, and dose ≥ 90 Gy).1

The biologically effective dose (BED) is dependent on tissue factors (α/β ratio and repair kinetics for sublethal radiation damage to DNA), prescribed dose, and overall treatment time. The latter is influenced by the activity of the 60Co sources, beam blocking, and individual patient geometry in the case of single-isocenter Gamma Knife radiosurgery (GKRS). Unscheduled breaks during treatment must be considered because ignoring sublethal repair associated with gaps in treatment leads to the BED being overestimated. Tuleasca et al. analyzed 408 patients who underwent GKRS (maximum 75–97.9 Gy) for type 1 TN.2 There was a significant association between BED and the incidence of facial numbness, but not the probability of pain relief, which was saturated at > 90% irrespective of BED values calculated, making the impact of BED difficult to evaluate. Yang et al. studied 192 patients with type 1 or 2 TN and found that BED was predictive of short-term pain relief (median 39 days) and facial numbness.3

The purpose of this study was to determine if BED is superior to prescription dose in predicting pain relief and facial numbness in a large cohort of type 1 TN patients undergoing GKRS as the first procedure. A secondary goal was to evaluate additional patient and dosimetric factors that may also predict patient outcome.

We embarked on this multicenter investigation knowing that TN is challenging to study. The endpoints are subjective (unlike tumor volume), and even TN grading scales have subjective aspects. There is also a lack of standardization of treatment planning, medication management, and timing of follow-up when multiple investigators are involved.

Methods

Study Design

Thirteen medical centers and 33 clinical investigators participated in this retrospective study of 871 patients with type 1 TN who underwent GKRS as their first procedure. Participating centers obtained IRB approval to share de-identified data with the International Radiosurgery Research Foundation.

Inclusion and Exclusion Criteria

Patients were included in this study if they were adults (≥ 18 years of age) with type 1 TN according to the classification of Eller et al. and had not undergone prior radiosurgery or an open or percutaneous surgical procedure.4 Patients experienced either failure of medical therapy, medication intolerance, or both. Patients must have been treated with a single isocenter without breaks using either the distal or proximal target with a minimum follow-up of 6 months.

Patients were excluded if type 2 TN was the predominant symptom or if TN was secondary to multiple sclerosis, tumor, or vertebrobasilar artery compression. Treatment plans based solely on CT were also excluded.

GKRS Technique

The Leksell G Frame (Elekta Instrument AB) was secured to the patient’s head using local anesthesia with or without sedation. High-resolution (1 mm) stereotactic MRI scans were acquired, including a 3D T1-weighted sequence (e.g., MPRAGE/FSPGR) with or without contrast and a T2-weighted sequence (e.g., CISS/FIESTA) without contrast. A stereotactic CT scan was obtained in some patients to correct for image distortion. A single 4-mm isocenter was centered on the trigeminal nerve, either proximally at the root entry zone (REZ) or distally along the cisternal segment, as selected by the treating physician. Sector blocking was used in some Perfexion/Icon cases to reduce the brainstem dose, but no guidelines were followed between centers. Radiosurgery plans were approved by a multidisciplinary team composed of a neurosurgeon, a radiation oncologist, and a medical physicist.

Data Collection

Investigators submitted data including 1) clinical presentation (date of diagnosis, side and distribution of pain, and medications), 2) treatment parameters from GammaPlan (Elekta Instrument AB) (dose, isodose, degree of sector blocking, distal or proximal target, treatment time, length of trigeminal nerve within 50% isodose lines, and maximum brainstem point dose), 3) treatment response (time to pain relief, duration of pain relief, and change in medications), 4) adverse events (sensory dysfunction: timing and severity), and 5) additional surgical procedures (type and timing). Central review of imaging and treatment plans was not performed, and therefore determination of the degree of neurovascular compression and target location was a function of local interpretation.

Outcome Measures

Initial follow-up assessment was performed 1 week to 6 months after GKRS. Subsequent evaluations were typically performed at 6- to 12-month intervals. Efficacy was determined according to the modified Barrow Neurological Institute (BNI) Pain Intensity Scale,29 and sensory dysfunction was reported using the BNI Facial Hypesthesia Scale30 (Supplemental Table 1). Initial pain relief was considered as complete pain relief with or without medications (modified BNI Pain Intensity Scale grades I–IIIa). Pain recurrence was considered as a change to a lower modified BNI Pain Intensity Scale grade (IIIb–V) in patients who achieved initial pain relief. Sensory dysfunction was considered as the appearance of new BNI Facial Hypesthesia Scale grade II–IV facial numbness.

Calculation of BED

BED was calculated using an equation described by Hopewell et al.:5

fd1

where DT represents the physical prescription dose, delivered as a single isocenter. The α/β ratio is a tissue-specific constant; in line with previous studies, a value of 2.47 Gy was used for a nerve originating from the brainstem (CNS white matter). Half-times for repair of sublethal irradiation damage were 0.19 hours (ln2/μ1) and 2.16 hours (ln2/μ2) with a partition coefficient (c) of 0.98.6,7 A full description of the BED formula is shown in Supplemental Methods. The terms “prescription dose” and “physical dose” are used interchangeably in all comparisons with BED.

Statistical Analysis

Descriptive statistics were reported as medians (range) or means (SD) for continuous variables or frequencies and percentages for categorical variables. The independent-samples Student t-test and Pearson chi-square test were performed for univariate analyses of continuous and categorical variables, respectively. Analysis of variance was performed when investigating the association between categorical and continuous variables. For the evaluation of outcomes such as initial pain relief, recurrence, and sensory dysfunction, the time to event was estimated and graphically represented by Kaplan-Meier survival curves. Patients who underwent secondary procedures or did not have a date for the onset of facial numbness were excluded from the sensory dysfunction analysis. In accordance with guidance by Perneger and Rothman, statistical corrections for multiple comparisons were not performed.8,9 Multivariable Cox proportional hazards regression models were generated and hazard ratios (HRs), 95% confidence intervals, and p values were reported for each covariate to allow readers to judge clinical and statistical significance. BED- and physical dose–related changes in the incidence of sensory dysfunction and pain-free rates, acutely (30 days) and at 1 year after radiosurgery, were evaluated using weighted probit analysis and interpreted using marginal effects. Separate multivariable Cox analyses were performed for BED and physical dose with treatment time excluded. Subgroup analyses based on target location were conducted because of postulated differences in radiosensitivity between proximal nerve fibers (ensheathed with central myelin) and distal nerve fibers (Schwann cell–generated myelin). The optimal numeric cutoff was estimated according to the Youden index based on a receiver operating characteristic (ROC) curve, and the area under the ROC curve (AUC) was calculated to determine accuracy using the cutpointr package.10 Results were considered statistically significant at p < 0.05. Statistical analyses were performed using R software (version 4.2.2, R Foundation for Statistical Computing).

Results

Baseline Patient Characteristics

A total of 871 patients from 13 institutions met the inclusion criteria (Table 1). The median age was 68 years, and 60% were female. Nearly 70% of patients experienced pain in the V2 and/or V3 dermatomes predominantly on the right side (60%). Most patients had modified BNI Pain Intensity Scale grade IV or V pain (89.2%) and were taking 1 or 2 pain medications (74.1%). The most common indication for GKRS was uncontrolled pain on medication (55.7%); however, 14% of patients were treated because of medication intolerance and 30% of patients had both pain and intolerance. A vessel was detected at the nerve in 51.4% of patients, and 11.4% had imaging evidence of nerve distortion by a vessel. Because central review was not performed, we were unable to categorize patients according to the International Classification of Headache Disorders.11

TABLE 1.

Clinical preoperative and demographic data

VariableValue
Sex
 Male349 (40.1)
 Female522 (59.9)
Age at GKRS, yrs68 (23–91)
Duration since diagnosis, mos (n = 867)57 (0–597)
Side of pain
 Rt526 (60.3)
 Lt345 (39.6)
Pain distribution (n = 867)
 V133 (3.8)
 V1+V2118 (13.6)
 V1+V2+V3106 (12.2)
 V1+V33 (0.3)
 V2204 (23.5)
 V2+V3264 (30.4)
 V3139 (16.0)
Vascular compression
 All cases (n = 860)
  No vessel at nerve320 (37.2)
  Vessel at nerve442 (51.4)
  Vessel w/ nerve distortion98 (11.4)
 Distal (n = 188)*
  No vessel at nerve72 (38.3)
  Vessel at nerve78 (41.5)
  Vessel w/ nerve distortion38 (20.2)
 Proximal (n = 653)*
  No vessel at nerve236 (36.1)
  Vessel at nerve360 (55.2)
  Vessel w/ nerve distortion57 (8.7)
Indication for GKRS (n = 869)
 Medically refractory484 (55.7)
 Intolerance to medications124 (14.3)
 Both261 (30.0)
Modified BNI Pain Intensity Scale grade at GKRS
 I0 (0.0)
 II2 (0.2)
 IIIa12 (1.4)
 IIIb80 (9.2)
 IV563 (64.6)
 V214 (24.6)
No. of TN drugs at GKRS
 039 (4.5)
 1348 (40.0)
 2297 (34.1)
 3149 (17.1)
 ≥438 (4.4)

Values are given as the number of patients (%) or median (range).

Target location was not available/excluded for 19 patients.

Radiosurgery Parameters

Most patients (72.4%) underwent radiosurgery with either the Perfexion or Icon Gamma Knife unit (Table 2). The proximal nerve was targeted in 77.9% of cases. There was a trend toward a distal target, started in 2014, with an even distribution between proximal and distal target cases over the last 4 years of the study. Sector blocking was used in 36.9% of procedures and more common with proximal targets (p = 0.01). The median prescription dose was 80 Gy (range 62.5–95 Gy) delivered over a median beam-on time of 43.2 minutes (range 25–170.5 minutes). The median BED was 1973.9 Gy2.47 (range 1253.1–2530.2 Gy2.47). The mean prescription dose and BED were significantly higher for distal targets (85.8 Gy and 2104.8 Gy2.47, respectively) than for proximal targets (80.8 Gy and 1958.0 Gy2.47, respectively) (p < 0.01). BED variation, as a function of treatment time and dose, for unblocked cases is shown in Supplemental Fig. 1. The mean prescription dose and BED were also significantly higher for unblocked cases (82.6 Gy and 2040.1 Gy2.47, respectively) than for blocked cases (80.8 Gy and 1901.5 Gy2.47, respectively) (p < 0.01). Table 3 shows the subgroups for prescription dose and associated BED ranges.

TABLE 2.

Radiosurgery parameters

VariableValue
Gamma Knife model
 U0 (0)
 B/C/4C240 (27.6)
 Perfexion/Icon631 (72.4)
1 isocenter871 (100.0)
Trigeminal target (n = 852)
 Distal188 (22.1)
 Proximal664 (77.9)
Sector blocking
 All cases
  No550 (63.1)
  Yes321 (36.9)
 Distal (n = 188)*
  No131 (69.7)
  Yes57 (30.3)
 Proximal (n = 664)*
  No400 (60.2)
  Yes264 (39.8)
Prescription dose, Gy
 All cases80/82 (62.5–95)
 Distal (n = 188)*86/86 (70–95)
 Proximal (n = 664)*80/81 (62.5–90)
Median beam-on time, mins (range)43.2 (25.0–170.5)
BED, Gy2.47
 All cases1973.9/1989.0 (1253.1–2530.2)
 Distal (n = 188)*2108.3/2104.8 (1394.1–2530.2)
 Proximal (n = 664)*1957.8/1958.0 (1253.1–2419.0)
Maximum brainstem dose, Gy
 All cases (n = 855)22.4/24.2 (1.7–75.8)
 Distal (n = 188)*16.0/16.0 (2.6–32.0)
 Proximal (n = 664)*24.6/26.6 (9.7–75.8)

Values are given as the number of patients (%) or median/mean (range) unless otherwise indicated.

Target location was not available/excluded for 19 patients.

TABLE 3.

Number of patients treated in 6 different 100% prescription dose ranges with the associated 9 BED ranges

VariableAll CasesDistalProximal
Prescription dose, Gy
 <70303
 70–7415114
 75–79927
 80–8456662504
 85–8916539126
 ≥90948410
BED range, Gy2.47
 1250–1392.5303
 1392.5–153512210
 1535–1677.523815
 1677.5–18201001684
 1820–1962.525721236
 1962.5–210525444210
 2105–2247.5993960
 2247.5–2390753045
 2390–2535.528271

The mean maximum brainstem point dose was 24.2 Gy (SD 10.7, range 1.7–75.8 Gy) and was significantly lower for distal targets (16.0 Gy, SD 5.79, range 2.6–32.0 Gy) compared with proximal targets (26.6 Gy, SD 10.51, range 9.7–75.8 Gy) (p < 0.01). The mean nerve length between the 50% isodose lines was 5.92 mm (SD 0.47), and it increased for up to 4 blocked sectors, after which it decreased (Supplemental Table 2).

Initial Pain Relief and Medication Use

The median follow-up was 21 months (range 6–156 months) (Table 4). Initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa) was achieved in 709 of 867 (81.8%) evaluable patients at a median of 30 days (range 0–365 days) (Fig. 1A). The actuarial probability of becoming pain free reached a plateau of 87% at 6.5 months. At the first follow-up, 60.8% of patients had modified BNI Pain Intensity Scale grade I–IIIa pain and 18.9% were pain free without medication (modified BNI Pain Intensity Scale grade I).

TABLE 4.

Postoperative assessment

VariableValue
FU, mos21.0 (6.0–156.0)
Initial pain relief, modified BNI Pain Intensity Scale grades I–IIIa (n = 867)709 (81.8)
Time to initial pain relief, days (n = 688)30 (0–365)
Modified BNI Pain Intensity Scale grade at initial FU (n = 867)
 I164 (18.9)
 II29 (3.3)
 IIIa335 (38.6)
 IIIb229 (26.4)
 IV89 (10.3)
 V21 (2.4)
No. of TN drugs at initial FU (n = 867)
 0199 (22.9)
 1314 (36.2)
 2262 (30.2)
 378 (9.0)
 ≥414 (1.6)
Recurrence of pain (n = 709)300 (42.3)
 Minor recurrence, no additional procedure*151 (51.0)
 Major recurrence, additional procedure*145 (49.0)
Time to recurrence of pain, mos (n = 381)44.2 (0–171.0)
Additional procedures after GKRS (n = 870)
 No656 (75.4)
 Yes214 (24.6)
Time to additional procedure, mos (n = 214)24.3 (2.3–172.0)
Type of procedure (n = 214)
 Microvascular decompression65 (30.4)
 Radiofrequency3 (1.4)
 Repeat radiosurgery112 (52.3)
 Glycerol injection26 (12.1)
 Balloon compression5 (2.3)
 Nerve lesioning3 (1.4)
New sensory dysfunction220 (25.3)
Time to sensory dysfunction, mos (n = 152)8.0 (0.5–85.4)
BNI Facial Hypesthesia Scale grade
 I649 (74.5)
 II117 (13.4)
 III76 (8.7)
 IV29 (3.3)

FU = follow-up.

Values are given as the number of patients (%) or median (range).

Information regarding additional procedures was only available for 296 patients.

Median time derived from the Kaplan-Meier curve in Fig. 1C.

FIG. 1.
FIG. 1.

A: Probability of initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa) reached a plateau at 6.5 months with rate of freedom from pain of 87%. B: Probability of initial pain relief stratified by target location (distal vs proximal). The 90-day probability of initial pain relief was 88.8% (95% CI 83.2%–92.5%) for distal and 69.2% (95% CI 65.5%–72.5%) for proximal targets (p < 0.0001). C: Probability of maintaining pain relief after achieving an initial response had a median time to recurrence of 44.2 months (95% CI 38.6–52.1 months). D: Probability of maintaining pain relief stratified by target location (distal vs proximal). The estimated median time to recurrence was 74.5 months (95% CI 57.4 months–not reached) for distal and 39 months (95% CI 33–47 months) for proximal targets (p < 0.01). E: Probability of new sensory dysfunction (all cases) had a 25.4% incidence of facial numbness at 5 years. F: Probability of new sensory dysfunction stratified by target location (distal vs proximal). The estimated risk of facial numbness by the 1st year was 19.5% (95% CI 12.9%–25.5%) for distal and 12.0% (95% CI 8.9%–15.0%) for proximal targets (p = 0.0013).

Prior to radiosurgery, the mean number of medications was 1.80 (SD 1.05), which decreased to 1.31 (SD 0.98) at the first follow-up visit and 1.11 (SD 0.98) at the last follow-up visit (both p < 0.01). This trend was similar for both distal and proximal subgroups. Earlier intervention was associated with a higher probability of medication reduction (p = 0.01); however, the ROC analysis to determine the optimal cutoff was of low accuracy (AUC 0.54). The probability of achieving pain-free status without the need for medication (modified BNI Pain Intensity Scale grade I) was inversely related to the number of medications prior to GKRS: 1 medication (49.4%), 2 medications (29.0%), 3 medications (11.1%), and ≥ 4 medications (3.4%).

The impact of target location on the likelihood of initial pain relief is shown in Fig. 1B. The median time to initial pain relief was 29 days (95% CI 28–30 days) for a distal target and 48 days (95% CI 48–59 days) for a proximal target. The 90-day probability of initial pain relief was 88.8% (95% CI 83.2%–92.5%) for a distal target and 69.2% (95% CI 65.5%–72.5%) for a proximal target (p < 0.0001).

Univariate probit analyses conducted for either BED or physical dose and pain-free incidence at 30 days did not yield significance for the distal or proximal subgroup. On multivariable Cox proportional hazards analysis, factors associated with a higher probability of 30-day pain relief were physical dose ≥ 85 Gy (HR 1.76, 95% CI 1.08–2.85 Gy; p = 0.02), age 63–73 years (HR 1.28, 95% CI 1.07–1.54 years; p < 0.01), age > 73 years (HR 1.34, 95% CI 1.11–1.62 years; p < 0.01), and distal target (HR 1.48, 95% CI 1.21–1.81; p < 0.01). BED ≥ 2100 Gy2.47 trended toward significance (HR 1.19, 95% CI 0.99–1.43; p = 0.057). The presence of a vessel at the nerve (without distortion) was correlated with a lower likelihood of initial pain relief (HR 0.82, 95% CI 0.70–0.97; p = 0.02). On subgroup analysis, BED ≥ 2100 Gy2.47 was a significant predictor of initial pain relief for the distal target (HR 1.46, 95% CI 1.05–2.03; p = 0.03) and physical dose ≥ 85 Gy was a significant predictor for the proximal target (HR 1.79, 95% CI 1.05–3.05; p = 0.03).

Maintenance of Pain Relief and Additional Procedures

Of the 709 patients who achieved initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa), 300 patients (42.3%) experienced at least one pain recurrence (modified BNI Pain Intensity Scale grades IIIb–V) after GKRS at a median actuarial time of 44 months (95% CI 38.6–52.1 months) (Table 4, Fig. 1C). The estimated risk of recurrence by the 1st year was 22.6% (95% CI 19.2%–25.8%). Patients who experienced initial pain relief within 30 days of GKRS were significantly less likely to have pain recurrence (35.5%) than patients who responded after 30 days (45.5%) (p = 0.01). Of the 300 patients with recurrent pain, a second procedure was required in 49.0%, including repeat GKRS (52.3%), microvascular decompression (30.4%), glycerol injection (12.1%), and others (5.1%). Early responders (≤ 30 days) were less likely to require a second procedure for pain recurrence (p = 0.01).

Maintenance of pain relief by target location is shown in Fig. 1D. The risk of recurrence at 1 year was 18% (95% CI 11.8%–23.7%) for distal and 23.9% (95% CI 19.9%–27.7%) for proximal cases. The estimated median times to recurrence were 74.5 months (95% CI 57.4–not reached months) for distal and 39 months (95% CI 33–47) for proximal cases (p < 0.01).

Univariate probit analysis showed a correlation between physical dose and 1-year pain relief (p = 0.01), whereas BED was not a significant predictor for distal, proximal, or all cases combined. On multivariable Cox proportional hazards analysis, factors associated with a higher probability of pain relief at 1 year were physical dose ≥ 85 Gy (HR 2.02, 95% CI 1.21–3.37 Gy; p < 0.01), BED ≥ 2100 Gy2.47 (HR 1.20, 95% CI 1.00–1.44 Gy2.47; p = 0.049), age 63–73 years (HR 1.30, 95% CI 1.08–1.56 years; p < 0.01), age > 73 years (HR 1.31, 95% CI 1.08–1.59 years; p < 0.01), and distal target (HR 1.64, 95% CI 1.34–2.01; p < 0.01). The presence of a vessel at the nerve (without distortion) was correlated with a lower likelihood of 1-year pain relief (HR 0.83, 95% CI 0.70–0.98; p = 0.03). On subgroup analysis, BED ≥ 2100 Gy2.47 was a significant predictor of 1-year pain relief for the distal target (HR 1.63, 95% CI 1.16–2.27; p < 0.01) and physical dose ≥ 85 Gy was a significant predictor for the proximal target (HR 2.05, 95% CI 1.17–3.61; p = 0.01).

There was significant overlap between the ≥ 85 Gy and BED ≥ 2100 Gy2.47 groups as shown in Fig. 2 and Supplemental Fig. 1. When considering all cases (n = 871), of the 212 patients with a BED ≥ 2100 Gy2.47, 199 (93.9%) had a physical dose ≥ 85 Gy; the additional cases (6.1%) received 80 or 84 Gy. Of the 267 patients receiving a physical dose ≥ 85 Gy, 199 (74.5%) also had a BED ≥ 2100 Gy2.47 (i.e., 25.5% of ≥ 85-Gy cases had a BED < 2100 Gy2.47).

FIG. 2.
FIG. 2.

Scatterplots of physical dose versus BED. The dotted red lines indicate the threshold values ≥ 85 Gy and BED ≥ 2100 Gy2.47. The right upper quadrants (orange) represent cases that overlap with both ≥ 85 Gy (red columns) and BED ≥ 2100 Gy2.47 (yellow rows). A: All cases. Of the 212 patients with a BED ≥ 2100 Gy2.47, 199 (93.9%) had a physical dose ≥ 85 Gy. Of the 267 patients with a physical dose ≥ 85 Gy, 199 (74.5%) also had a BED ≥ 2100 Gy2.47. B: Unblocked cases. Of the 198 patients with a BED ≥ 2100 Gy2.47, 185 (93.4%) had a physical dose ≥ 85 Gy. Of the 229 patients with a physical dose ≥ 85 Gy, 185 (80.8%) also had a BED ≥ 2100 Gy2.47. C: Blocked cases. Of the 14 patients with a BED ≥ 2100 Gy2.47, 14 (100%) had a physical dose ≥ 85 Gy. Of the 38 patients with a physical dose ≥ 85 Gy, 14 (36.8%) also had a BED ≥ 2100 Gy2.47.

Sensory Dysfunction

Patients who underwent second procedures were censored at that time with regard to facial numbness. New onset of facial numbness occurred in 25.3% of patients at a median time of 8.0 months (0.5–85.4 months) (Table 4, Fig. 1E). Nonbothersome numbness (BNI Facial Hypesthesia Scale grade II) was noted in 13.4% of patients, whereas 12% of patients experienced somewhat bothersome or very bothersome numbness (BNI Facial Hypesthesia Scale grade III or IV). Figure 1F shows the probability of new sensory dysfunction based on target location. The estimated risk of facial numbness at 1 year was 19.5% (95% CI 12.9%–25.5%) for distal and 12.0% (95% CI 8.9%–15.0%) for proximal targets (p < 0.01).

Univariate probit analyses found that physical dose, BED, and brainstem dose were not significant factors for the development of sensory dysfunction (BNI Facial Hypesthesia Scale grades II–IV) when all cases were combined. However, when analyzed by target location (distal vs proximal) and the degree of sector blocking (none vs ≥ 1 sector), BED was correlated with the incidence of sensory dysfunction for distal targets (p = 0.02), blocked cases (p = 0.03), and the distal/blocked subgroup (p < 0.01). Maximum brainstem point dose was a significant predictor for unblocked cases (p < 0.01).

On multivariable Cox proportional hazards analysis, factors associated with a higher probability of sensory dysfunction (all grades) were a distal target (HR 1.71, 95% CI 1.08–2.69; p = 0.02) and vessel with nerve distortion (proximal subgroup: HR 2.37, 95% CI 1.16–4.84; p = 0.02). Physical dose, BED, and maximum brainstem point dose were not significant in separate multivariable analyses when all cases were considered. Using a dose quartile approach, maximum brainstem dose ≥ 29.5 Gy was the most significant predictor of bothersome sensory dysfunction (BNI Facial Hypesthesia Scale grades III–IV) when compared with the reference dose quartile (HR 3.01, 95% CI 1.32–6.83; p < 0.01). Neither physical dose nor BED to the nerve was associated with BNI Facial Hypesthesia Scale grade III–IV facial numbness. Several variables demonstrated in prior studies to predict sensory dysfunction were not significant in the present study, including the number of blocked sectors and nerve length within the 50% isodose lines (Flickinger effect).12

Discussion

Key Study Findings

In this multicenter study of 871 type 1 TN patients who were managed with GKRS as a first procedure, we documented initial pain relief in 82% of patients, a 42% incidence of pain recurrence (49% of whom required a second procedure), and a 25% rate of sensory dysfunction. Table 5 highlights significant patient and treatment variables on multivariable analysis. Patient age ≥ 63 years was predictive of a higher probability of both initial pain relief and maintaining pain relief. Distal targeting demonstrated higher initial pain relief and higher maintenance of pain relief, but also higher incidence of sensory dysfunction. BED ≥ 2100 Gy2.47 was predictive of pain relief at 30 days and 1 year for the distal target, whereas physical dose ≥ 85 Gy was a significant predictor for the proximal target. BED was associated with sensory dysfunction for the distal target on univariate probit analysis but not on Cox multivariable analysis. Maximum brainstem point doses ranging from 29.5 to 75.8 Gy were associated with a higher probability of bothersome facial numbness.

TABLE 5.

Significant patient and treatment factors

OutcomePatient FactorsTreatment Factors
AgeVessel at Nerve ± DistortionDistal TargetDoseBEDBrainstem Point Dose
30-day pain relief↑ for ≥63 yrs↑ for ≥85 Gy at proximal target↑ for ≥2100 Gy2.47 at distal targetNS
1-yr pain relief↑ for ≥63 yrs↑ for ≥85 Gy at proximal target↑ for ≥2100 Gy2.47 at distal targetNS
Facial numbness, all grades of BNI Facial Hypesthesia ScaleNSNSNS*NS
Facial numbness, grades III–IV of BNI Facial Hypesthesia ScaleNSNSNSNS↑ for ≥29.5 Gy

NS = not significant; ↑ = higher probability; ↓ = lower probability.

Significance was determined at p < 0.05 on multivariable Cox proportional hazards analysis.

Higher probability of facial numbness for distal target (univariate probit, p = 0.02).

Benchmarking Key Outcome Measures

Tuleasca et al. reviewed 45 studies (5687 patients) and documented a high rate of initial pain relief after GKRS for TN (range 68.6%–100%, mean 84.8%, median 85.6%).1 In a single-institution study by Régis et al., 34.4% of patients who were initially pain free experienced at least one episode of recurrent pain and 22.5% required a second procedure.13 In the present study, initial pain relief was found in 81.8% of patients after a median of 30 days, but they had a higher rate of recurrence (42.3%) than that reported by Régis et al.13 The presence of a vessel at the nerve was associated with a lower likelihood of initial pain relief; however, the degree of neurovascular compression may have been open to interpretation since central review was not used. We hypothesize that vascular compression may lead to hypoxia, which reduces radiosensitivity of the nerve. Alternatively, patients with vascular compression may have a more severe etiology of TN that is less responsive to GKRS, although this is speculative. Importantly, 51.0% of our patients with recurrent pain were managed by changing their medication regimen, whereas 49.0% underwent a second procedure.

While the goal of GKRS for TN should be to achieve pain-free status without the need for medication (modified BNI Pain Intensity Scale grade I), this may not be feasible in patients who are taking multiple medications. Of those patients on a single medication prior to GKRS, 49.4% were pain free without medication, whereas for those on 3 medications, only 11.1% achieved modified BNI Pain Intensity Scale grade I status. Therefore, a secondary goal should be to reduce medication use in patients who cannot be tapered off medications. Our study shows that the mean number of medications significantly decreased after GKRS (baseline 1.80, last follow-up 1.11) and early GKRS after diagnosis was associated with a higher probability of medication reduction. These results are consistent with the study by Mureb et al. in which there was a similar reduction in medication use (mean 1.98 to 0.90) and a superior response in patients treated within 4 years of TN diagnosis.14

The most common adverse event after GKRS for TN is new or worsening facial numbness, which has been reported in 20%–25% of patients with a mean onset of 6–36 months.1 In the present study, 25.3% of patients developed new sensory dysfunction at a median of 8 months. The mean nerve length between the 50% isodose lines was 5.92 mm (5.73–6.92 mm), but no correlation was found between nerve length and risk of sensory dysfunction on multivariable Cox proportional hazards analysis (p = 0.77). This differs from the results of a randomized double-blind study by Flickinger et al. comparing one versus two isocenters targeting the distal nerve.12 In that study, the two-isocenter cohort had a significantly higher incidence of facial numbness than the single-isocenter group (32.5% vs 16.8%, p = 0.018), which was attributed to a longer irradiated nerve length (mean 8.7 vs 5.4 mm). Our lack of confirmation may be explained by a significantly longer irradiated nerve length in the study by Flickinger et al. (mean 7.06 mm) compared with the current study (mean 5.72 mm) as well as the more complex dosimetry of dual (Flickinger study) versus single (our study) isocenters.

Impact of Target Location

Isocenter location along the trigeminal nerve has undergone considerable evolution. The initial target for GKRS was the gasserian ganglion as reported by Leksell and later by Lindquist et al.15,16 In 1993, Rand et al. described outcomes of 12 TN patients treated with GKRS using a variety of targets including the gasserian ganglion, distal nerve (retrogasserian), and proximal REZ.17 The REZ was thought to be an attractive target for GKRS because it comprises the transition from central myelin to Schwann cell–generated myelin and has been postulated to be more sensitive to various forms of injury (compression, stretching, and irradiation).18 A 2016 study by Régis et al. described 497 TN patients treated with GKRS using a retrogasserian target to facilitate dose escalation while minimizing sensory dysfunction.13 The transition toward a distal target is evident in the present multicenter study. The REZ was the predominant target from 2001 to 2017, followed by a relatively even distribution between proximal and distal targets over the last 4 years of the study.

The literature is replete with studies on the impact of target location. Tuleasca et al. pooled studies from Matsuda et al., Park et al., and Xu et al. and concluded that a distal target was favored based on similar initial pain relief (level II evidence), higher long-term pain relief (level II and III evidence), and lower sensory dysfunction (level II evidence).1,1921 The present study suggests that a distal target provided superior pain relief (both initial and maintenance) but was associated with a higher probability of sensory dysfunction (BNI Facial Hypesthesia Scale grades II–IV). The median prescription dose was significantly higher for distal than proximal targets (86 Gy vs 80 Gy), which could be a confounding factor. However, with the inclusion of prescription dose and BED in separate Cox multivariable analyses, a distal target remained a significant predictor of pain relief and facial numbness. This evidence suggests that the retrogasserian portion of the trigeminal nerve may be more radiosensitive than the REZ. Our conclusion stands in contrast with conventional wisdom in the TN literature that the REZ is more sensitive to irradiation. The biological basis for this phenomenon awaits further study.

In light of our findings, a more distal target is a consideration, based on the higher probability of both initial and long-term pain relief. However, this advantage must be balanced with the higher incidence of facial numbness (BNI Facial Hypesthesia Scale grades II–IV) associated with a distal target. Target location was not a significant predictor of bothersome facial numbness (BNI Facial Hypesthesia Scale grades III–IV) in our study, which should be reassuring to clinicians when selecting a distal target for GKRS. Finally, the choice of an optimal target must be tempered with the realization that the reported target location in this study may have been open to some interpretation. We requested a binary choice (distal vs proximal), whereas the target location in reality was a continuous variable.

Predictive Value of Physical Dose and BED for Key Outcome Measures

A prescription dose of 70–90 Gy is generally considered effective for TN pain relief after GKRS.1 Kondziolka et al. demonstrated that a prescription dose ≥ 70 Gy was associated with a higher probability of pain relief, whereas Longhi et al. found superior pain-free outcomes for patients treated with a prescription dose ≥ 80 Gy.22,23 A recent study by Kotecha et al. showed a stepwise increase in excellent or good pain response (equivalent to modified BNI Pain Intensity Scale grades I–IIIa) for ≤ 82 Gy, 83–86 Gy, and ≥ 90 Gy.24 Kim et al. used prescription doses of either 80 Gy or 85 Gy with similar rates of pain relief, but faster onset of pain relief at 85 Gy.25 The escalation of prescription doses over the last 3 decades must be weighed against the increasing incidence of sensory dysfunction associated with higher dose thresholds as reported by Pollock (≥ 90 Gy), Longhi (≥ 90 Gy), Kim (85 Gy), and Kotecha (≥ 90 Gy).2326 Mureb et al. and Tuleasca et al. failed to show a correlation between prescription dose and risk of facial numbness, whereas Tuleasca et al. found a strong correlation with BED.2,14

In the present study, a prescription dose ≥ 85 Gy was associated with a higher probability of 30-day and 1-year pain relief (overall and proximal subgroup), which is similar to the thresholds demonstrated in the studies by Kim et al. and Kotecha et al.24,25 However, it is important to note that these cases were dominated by associated BED values ≥ 2100 Gy2.47 in unblocked cases. A surprising finding was the lack of correlation between prescription dose and sensory dysfunction in our study compared with prior studies,2326 but our finding was similar to those of Mureb et al. and Tuleasca et al.2,14 This may reflect a limitation of our multicenter study in which target location was more heterogeneous and prescription dose was less likely to be associated with maximum brainstem dose (median distal dose of 16.0 Gy vs median proximal dose of 24.6 Gy). A maximum brainstem point dose ≥ 29.5 Gy was found to be a significant predictor of BNI Facial Hypesthesia Scale grade III–IV sensory dysfunction. This underscores the importance of balancing prescription dose and brainstem exposure to maximize durable pain relief while avoiding bothersome facial numbness. An unanswered question in our data is the location of the maximum point dose within the brainstem because this value was not measured at a standardized anatomical point. In addition, we collected point doses rather than a dose volume (e.g., 10 mm3), which may be more informative.

The 60Co sources in GKRS decay with a half-life of 5.26 years and thus a 50% reduction in dose rate and doubling of treatment time by one half-life (assuming a constant prescription dose and the same treatment geometry). The BED decreases with prolonged treatment times as the result of greater sublethal DNA repair (half-times of 0.19 hours and 2.16 hours) during treatment when single-strand breaks can be repaired prior to the creation of other local damage that otherwise might lead to a lethal double-strand break when that cell attempts to divide.27 Lee et al. analyzed the GKRS reference dose rate (relevant to the dose inside a calibration phantom) in TN patients treated with a prescription dose of 80 Gy without beam blocking, thus ignoring individual patient geometry.28 A higher dose rate was associated with a higher probability of both initial and long-term pain relief. Tuleasca et al. performed a retrospective analysis of 408 TN patients from the prospective study by Régis et al.2 BED was not associated with pain relief at 1 and 2 years; however, there was a clear relationship between BED and the incidence of facial numbness using univariate probit analysis. The incidence varied from < 5% at a BED of 1800 Gy2.47 to 42% at a BED of 2600 Gy2.47. Yang et al. studied 192 patients with type 1 or type 2 TN, including 133 patients originally reported by Lee et al., and found that BED was associated with short-term pain relief (median 39 days) as measured by the Penn Facial Pain Scale–Revised (BNI grade was not an outcome measure).3,28 The GKRS reference dose rate had a similar predictive value such that a decrease in dose rate of 1.5 Gy/min resulted in a 31.8% lower rate of overall pain relief. A higher BED was associated with an increased incidence of facial numbness at the highest BED quartile.

The results of the present study confirmed the finding of Yang et al. that BED was associated with the probability of initial pain relief (defined as 30-day pain relief in the present study).3 Specifically, it was found that patients treated at the distal target with a BED ≥ 2100 Gy2.47 were more likely to achieve pain relief at both 30 days and 1 year. The association between BED and 30-day pain relief is difficult to explain because the acute response reflects functional changes in cells, largely membrane permeability, which is not related to DNA damage/repair as described by the BED equation. In contrast, 1-year pain relief is from a typical delayed radiation response that reflects DNA damage to target cells, which BED takes into account—the longer the treatment time, the less effective the physical dose delivered.

In the present study, it was also found that BED was associated with the risk of facial numbness for the distal target subgroup using univariate probit analysis (similar to the study by Tuleasca et al.2); however, this relationship was not confirmed on multivariable analysis. A key feature of the study by Tuleasca et al. was the inclusion of patients treated using a prescription dose > 90 Gy and a group with a higher BED associated with a 40% incidence of facial numbness.2 Our cohort lacked a similar high prescription dose group, which might have been informative for the influence of dose and BED on sensory dysfunction. From a statistical perspective, Tuleasca et al. relied on a univariate probit analysis, whereas both univariate probit and Cox multivariable methods (including subgroup analyses) were used in the present study. Probit is a dose/BED response methodology in which all variables are meant to be controlled except for dose. It is an accepted approach for detecting a correlation between values with threshold relationships that are a hallmark of clinical effects precipitated by radiation. However, probit is a univariate analysis that does not take into account a variety of patient and treatment factors. In contrast, the Cox linear proportional hazards regression analysis simultaneously assesses multiple variables, although it is not ideally suited for threshold relationships. Both statistical approaches were used in the present study in a complementary fashion—probit analysis to detect relationships and Cox analysis to manage potential confounding variables.

Implications for Treatment Planning

The goal of the present study was to determine if BED is superior to prescription dose in predicting pain relief and facial numbness in TN patients undergoing GKRS as the first procedure. This hypothesis was not confirmed; however, several factors were associated with superior outcomes that could be considered during treatment planning (Table 5): 1) distal targeting was associated with a higher probability of initial pain relief and higher maintenance of pain relief (must be balanced with the higher incidence of facial numbness); 2) a BED ≥ 2100 Gy2.47 was predictive of pain relief for the distal target, whereas a physical dose ≥ 85 Gy was predictive of pain relief for the proximal target (but the BED range in this group was restricted); and 3) a brainstem point dose < 29.5 Gy was a potential constraint for bothersome facial numbness. The selection of prescription dose (or BED) must be evaluated in the context of isocenter placement and brainstem tolerance to achieve the twin goals of pain relief and avoidance of sensory loss.

Tuleasca et al. provided a table of prescription dose and treatment time combinations equivalent to a BED of 1820 Gy2.47 and BED of 2105 Gy2.47; the BED values are associated with approximately 5% and 10% incidences of facial numbness, respectively.2 A selected part of this lookup table is included in Fig. 3, which can be used to prescribe a physical dose with associated treatment time to achieve a BED of 2100 Gy2.47 (within 0.2% of 2105 Gy2.47). Use of this table when the distal nerve is targeted may help achieve the dual outcomes of high probability of 30-day and 1-year pain relief as reported in the current study and a 10% risk of facial numbness as reported by Tuleasca et al.2

FIG. 3.
FIG. 3.

Steps for integrating BED into TN treatment planning. The lookup table provides the prescription dose and treatment time combinations required for a given patient-specific dose rate that would result in the delivery of a BED of 2100 Gy2.47. Reprinted from World Neurosurgery, 134, Tuleasca C, Paddick I, Hopewell JW, Jones B, Millar WT, Hamdi H, Porcheron D, Levivier M, Régis J, Establishment of a therapeutic ratio for Gamma Knife radiosurgery of trigeminal neuralgia: the critical importance of biologically effective dose versus physical dose, e204-e213, © 2020, with permission from Elsevier.

Limitations

TN is a challenging disease to study, especially in a multicenter setting. The two clinical endpoints (pain relief and facial numbness) are both subjective and may not be optimally categorized using the existing grading scales. This is a retrospective study comprising 13 institutions and 33 clinical investigators, which makes it subject to selection bias, temporal bias, and diverse institutional treatment practices. There was a lack of standardization of treatment planning (target location, prescription dose, and beam blocking), medication management, and timing of follow-up. At some centers, the date of pain relief may have defaulted to the date of the first follow-up visit (30 vs 45 days), which would affect the rate of 30-day pain relief and subsequent statistical analyses (Fig. 1). Treatment plans were not centrally reviewed, so it was not possible to further investigate secondary hypotheses related to target location, nerve length, and brainstem dosimetry.

For a procedure that superficially appears to be ideal to study in relation to BED and clinical outcomes, a number of potential confounding factors were identified including 1) sector blocking, which alters the dose distribution along the length of the nerve; 2) target location, which can affect both the radiosensitivity of the nerve and the width of the nerve at the isocenter; and 3) peak brainstem dose, which may not always be at the same anatomical location. In this regard, proximal targeting leads to a maximum dose at the point of nerve emergence from the brainstem, while distal targeting can result in a maximum dose at a significantly more anterior and variable location. All these factors create difficulty in analyzing the clinical effects of physical dose and BED. While it was ideal to divide the cohort into proximal and distal target groups because of potential differences in radiosensitivity along the nerve, it became apparent over the period of the analysis that the definitions of these two targets were not clear-cut. In reality, a spectrum of target positions along the length of the nerve were delivered. This can be seen in the median brainstem doses, which were significantly lower for distal targets (16.0 Gy, range 2.6–32.0 Gy) than proximal targets (24.6 Gy, range 9.7–75.7 Gy) but demonstrated significant overlap, suggesting that subjective clinician definition of target position was somewhat arbitrary. It is also recognized that there are changes in radiosensitivity with the oxygen tension within the tissues during irradiation. The 51.4% incidence of a vessel present in this cohort, which was associated with a lower incidence of acute pain relief, and the 11.4% incidence of nerve distortion by a vessel, which may lead to a degree of hypoxia over the period of irradiation, may add bias to the study. Lastly, initial pain relief (30 days) mainly reflects membrane damage and is not subject to DNA repair kinetics, so results from this acute reaction should be treated with caution.

The confounding factors of target location, beam blocking, brainstem dose, and prescription dose cannot be completely eliminated even with multivariable analysis. Brainstem dose analysis via quartile division by patient numbers yielded unequal dose ranges, which may underestimate the true point dose tolerance of the brainstem. Differences in patient populations and methodology between this study and prior BED studies must be considered when comparing the results. Despite these limitations, we believe that the findings of this large multicenter study reflect the broad practice of GKRS for TN, and therefore the conclusions should be generalizable to type 1 TN patients undergoing GKRS as a first procedure.

Conclusions

The initial hypothesis that BED would better predict key outcomes revealed a more nuanced interplay of variables. BED was predictive of pain relief for the distal target, whereas physical dose was the significant predictor of pain relief for the proximal target. The maximum brainstem dose was associated with bothersome sensory dysfunction. The GKRS treatment plan for TN is a delicate balance between these interrelated parameters. We provide treatment planning considerations and a BED lookup table to assist neurosurgeons and radiation oncologists in formulating treatment plans for patients with type 1 TN. Clearly, more work is necessary to evaluate these relationships further.

Acknowledgments

We thank Karim R. Nathani, MBBS, for his statistical support. We also thank Martha Headworth and Tonya Hines for medical illustration, Laura Mancini for medical editing, and Parinaz Mahbod, MSc, for assistance with data analysis.

Disclosures

Mr. Paddick reported personal fees from Elekta, Zap Surgical, and Varian Medical Systems outside the submitted work. Dr. Leduc reported grants from Sherbrooke University outside the submitted work. Dr. Lunsford reported stock ownership in Elekta outside the submitted work. Mr. Bernstein reported financial support from NeuroPoint Alliance. Dr. Kshettry reported consulting fees from Stryker and Integra outside the submitted work. Dr. Liščák reported consulting fees from Elekta outside the submitted work. Dr. Kondziolka reported grants from Brainlab and Elekta outside the submitted work.

Author Contributions

Conception and design: Warnick, Paddick, Sheehan, Kondziolka. Acquisition of data: Warnick, Paddick, Mathieu, Adam, Iorio-Morin, Leduc, Niranjan, Wei, Waite, Peker, Samanci, Tek, Mantziaris, Pikis, Sheehan, Tripathi, Kumar, Alzate, Bernstein, Ahorukomeye, Kshettry, Speckter, Hernandez, Urgošík, Liščák, Yang, Lee, Patel, Kusyk, Shepard, Kondziolka. Analysis and interpretation of data: Warnick, Paddick, Iorio-Morin, Johnson, Bydon, Waite, Sheehan, Kumar, Ahorukomeye, Kshettry, Urgošík, Patel, Kondziolka. Drafting the article: Warnick, Paddick, Johnson, Bydon, Kondziolka. Critically revising the article: Warnick, Paddick, Mathieu, Iorio-Morin, Johnson, Bydon, Niranjan, Lunsford, Samanci, Mantziaris, Pikis, Sheehan, Tripathi, Kumar, Alzate, Ahorukomeye, Kshettry, Speckter, Liščák, Patel, Shepard, Kondziolka. Reviewed submitted version of manuscript: Warnick, Paddick, Mathieu, Adam, Iorio-Morin, Johnson, Bydon, Niranjan, Jose, Peker, Samanci, Mantziaris, Pikis, Tripathi, Kshettry, Urgošík, Lee, Patel, Kusyk, Shepard, Kondziolka. Approved the final version of the manuscript on behalf of all authors: Warnick. Statistical analysis: Johnson, Bydon. Administrative/technical/material support: Hamel, Yang, Shepard. Study supervision: Warnick, Paddick, Bydon, Sheehan, Kondziolka.

Supplemental Information

Online-Only Content

Supplemental material is available with the online version of the article.

Previous Presentations

Presented at the 5th Biennial Scientific Meeting of the International Radiosurgery Research Foundation, Miami, Florida, October 29, 2023.

References

  • 1

    Tuleasca C, Régis J, Sahgal A, et al. Stereotactic radiosurgery for trigeminal neuralgia: a systematic review. J Neurosurg. 2018;130(3):733757.

  • 2

    Tuleasca C, Paddick I, Hopewell JW, et al. Establishment of a therapeutic ratio for Gamma Knife radiosurgery of trigeminal neuralgia: the critical importance of biologically effective dose (BED) versus physical dose. World Neurosurg. 2020;134:e204e213.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Yang AI, Mensah-Brown KG, Shekhtman EF, et al. Gamma Knife radiosurgery for trigeminal neuralgia provides greater pain relief at higher dose rates. J Radiosurg SBRT. 2022;8(2):117125.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Eller JL, Raslan AM, Burchiel KJ. Trigeminal neuralgia: definition and classification. Neurosurg Focus. 2005;18(5):E3.

  • 5

    Hopewell JW, Millar WT, Lindquist C. Radiobiological principles: their application to Gamma Knife therapy. Prog Neurol Surg. 2012;25:3954.

  • 6

    Ang KK, Jiang GL, Guttenberger R, et al. Impact of spinal cord repair kinetics on the practice of altered fractionation schedules. Radiother Oncol. 1992;25(4):287294.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Pop LA, van der Plas M, Ruifrok AC, Schalkwijk LJ, Hanssen AE, van der Kogel AJ. Tolerance of rat spinal cord to continuous interstitial irradiation. Int J Radiat Oncol Biol Phys. 1998;40(3):681689.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Perneger TV. What’s wrong with Bonferroni adjustments. BMJ. 1998;316(7139):12361238.

  • 9

    Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):4346.

  • 10

    Thiele C, Hirschfeld G. cutpointr: improved estimation and validation of optimal cutpoints in R. J Stat Softw. 2021;98(11):127.

  • 11

    Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition. Cephalalgia. 2018;38(1):1211.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Flickinger JC, Pollock BE, Kondziolka D, et al. Does increased nerve length within the treatment volume improve trigeminal neuralgia radiosurgery? A prospective double-blind, randomized study. Int J Radiat Oncol Biol Phys. 2001;51(2):449454.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Régis J, Tuleasca C, Resseguier N, et al. Long-term safety and efficacy of Gamma Knife surgery in classical trigeminal neuralgia: a 497-patient historical cohort study. J Neurosurg. 2016;124(4):10791087.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Mureb M, Golub D, Benjamin C, et al. Earlier radiosurgery leads to better pain relief and less medication usage for trigeminal neuralgia patients: an international multicenter study. J Neurosurg. 2020;135(1):237244.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand. 1971;137:311314.

  • 16

    Lindquist C, Kihlström L, Hellstrand E. Functional neurosurgery—a future for the gamma knife? Stereotact Funct Neurosurg. 1991;57(1-2):7281.

  • 17

    Rand RW, Jacques DB, Melbye RW, Copcutt BG, Levenick MN, Fisher MR. Leksell Gamma Knife treatment of tic douloureux. Stereotact Funct Neurosurg. 1993;61(suppl 1):93-102.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Gorgulho A. Radiation mechanisms of pain control in classical trigeminal neuralgia. Surg Neurol Int. 2012;3(suppl 1):S17-S25.

  • 19

    Matsuda S, Serizawa T, Nagano O, Ono J. Comparison of the results of 2 targeting methods in Gamma Knife surgery for trigeminal neuralgia. J Neurosurg. 2008;109(suppl):185189.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Park SH, Hwang SK, Kang DH, Park J, Hwang JH, Sung JK. The retrogasserian zone versus dorsal root entry zone: comparison of two targeting techniques of Gamma Knife radiosurgery for trigeminal neuralgia. Acta Neurochir (Wien). 2010;152(7):11651170.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Xu Z, Schlesinger D, Moldovan K, et al. Impact of target location on the response of trigeminal neuralgia to stereotactic radiosurgery. J Neurosurg. 2014;120(3):716724.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22

    Kondziolka D, Flickinger JC, Lunsford LD, Habeck M. Trigeminal neuralgia radiosurgery: the University of Pittsburgh experience. Stereotact Funct Neurosurg. 1996;66(suppl 1):343-348.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Longhi M, Rizzo P, Nicolato A, Foroni R, Reggio M, Gerosa M. Gamma Knife radiosurgery for trigeminal neuralgia: results and potentially predictive parameters—part I: idiopathic trigeminal neuralgia. Neurosurgery. 2007;61(6):12541261.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Kotecha R, Kotecha R, Modugula S, et al. Trigeminal neuralgia treated with stereotactic radiosurgery: the effect of dose escalation on pain control and treatment outcomes. Int J Radiat Oncol Biol Phys. 2016;96(1):142148.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    Kim YH, Kim DG, Kim JW, et al. Is it effective to raise the irradiation dose from 80 to 85 Gy in Gamma Knife radiosurgery for trigeminal neuralgia? Stereotact Funct Neurosurg. 2010;88(3):169176.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26

    Pollock BE, Phuong LK, Foote RL, Stafford SL, Gorman DA. High-dose trigeminal neuralgia radiosurgery associated with increased risk of trigeminal nerve dysfunction. Neurosurgery. 2001;49(1):5864.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Fowler JF, Welsh JS, Howard SP. Loss of biological effect in prolonged fraction delivery. Int J Radiat Oncol Biol Phys. 2004;59(1):242249.

  • 28

    Lee JY, Sandhu S, Miller D, Solberg T, Dorsey JF, Alonso-Basanta M. Higher dose rate Gamma Knife radiosurgery may provide earlier and longer-lasting pain relief for patients with trigeminal neuralgia. J Neurosurg. 2015;123(4):961968.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29

    Shetter AG, Rogers CL, Ponce F, Fiedler JA, Smith K, Speiser BL. Gamma knife radiosurgery for recurrent trigeminal neuralgia. J Neurosurg. 2002;97(5 suppl):536538.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 30

    Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han PP, Speiser BL. Gamma Knife radiosurgery for trigeminal neuralgia: the initial experience of the Barrow Neurological Institute. Int J Radiat Oncol Biol Phys. 2000;47(4):10131019.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • Collapse
  • Expand
  • FIG. 1.

    A: Probability of initial pain relief (modified BNI Pain Intensity Scale grades I–IIIa) reached a plateau at 6.5 months with rate of freedom from pain of 87%. B: Probability of initial pain relief stratified by target location (distal vs proximal). The 90-day probability of initial pain relief was 88.8% (95% CI 83.2%–92.5%) for distal and 69.2% (95% CI 65.5%–72.5%) for proximal targets (p < 0.0001). C: Probability of maintaining pain relief after achieving an initial response had a median time to recurrence of 44.2 months (95% CI 38.6–52.1 months). D: Probability of maintaining pain relief stratified by target location (distal vs proximal). The estimated median time to recurrence was 74.5 months (95% CI 57.4 months–not reached) for distal and 39 months (95% CI 33–47 months) for proximal targets (p < 0.01). E: Probability of new sensory dysfunction (all cases) had a 25.4% incidence of facial numbness at 5 years. F: Probability of new sensory dysfunction stratified by target location (distal vs proximal). The estimated risk of facial numbness by the 1st year was 19.5% (95% CI 12.9%–25.5%) for distal and 12.0% (95% CI 8.9%–15.0%) for proximal targets (p = 0.0013).

  • FIG. 2.

    Scatterplots of physical dose versus BED. The dotted red lines indicate the threshold values ≥ 85 Gy and BED ≥ 2100 Gy2.47. The right upper quadrants (orange) represent cases that overlap with both ≥ 85 Gy (red columns) and BED ≥ 2100 Gy2.47 (yellow rows). A: All cases. Of the 212 patients with a BED ≥ 2100 Gy2.47, 199 (93.9%) had a physical dose ≥ 85 Gy. Of the 267 patients with a physical dose ≥ 85 Gy, 199 (74.5%) also had a BED ≥ 2100 Gy2.47. B: Unblocked cases. Of the 198 patients with a BED ≥ 2100 Gy2.47, 185 (93.4%) had a physical dose ≥ 85 Gy. Of the 229 patients with a physical dose ≥ 85 Gy, 185 (80.8%) also had a BED ≥ 2100 Gy2.47. C: Blocked cases. Of the 14 patients with a BED ≥ 2100 Gy2.47, 14 (100%) had a physical dose ≥ 85 Gy. Of the 38 patients with a physical dose ≥ 85 Gy, 14 (36.8%) also had a BED ≥ 2100 Gy2.47.

  • FIG. 3.

    Steps for integrating BED into TN treatment planning. The lookup table provides the prescription dose and treatment time combinations required for a given patient-specific dose rate that would result in the delivery of a BED of 2100 Gy2.47. Reprinted from World Neurosurgery, 134, Tuleasca C, Paddick I, Hopewell JW, Jones B, Millar WT, Hamdi H, Porcheron D, Levivier M, Régis J, Establishment of a therapeutic ratio for Gamma Knife radiosurgery of trigeminal neuralgia: the critical importance of biologically effective dose versus physical dose, e204-e213, © 2020, with permission from Elsevier.

  • 1

    Tuleasca C, Régis J, Sahgal A, et al. Stereotactic radiosurgery for trigeminal neuralgia: a systematic review. J Neurosurg. 2018;130(3):733757.

  • 2

    Tuleasca C, Paddick I, Hopewell JW, et al. Establishment of a therapeutic ratio for Gamma Knife radiosurgery of trigeminal neuralgia: the critical importance of biologically effective dose (BED) versus physical dose. World Neurosurg. 2020;134:e204e213.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Yang AI, Mensah-Brown KG, Shekhtman EF, et al. Gamma Knife radiosurgery for trigeminal neuralgia provides greater pain relief at higher dose rates. J Radiosurg SBRT. 2022;8(2):117125.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Eller JL, Raslan AM, Burchiel KJ. Trigeminal neuralgia: definition and classification. Neurosurg Focus. 2005;18(5):E3.

  • 5

    Hopewell JW, Millar WT, Lindquist C. Radiobiological principles: their application to Gamma Knife therapy. Prog Neurol Surg. 2012;25:3954.

  • 6

    Ang KK, Jiang GL, Guttenberger R, et al. Impact of spinal cord repair kinetics on the practice of altered fractionation schedules. Radiother Oncol. 1992;25(4):287294.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Pop LA, van der Plas M, Ruifrok AC, Schalkwijk LJ, Hanssen AE, van der Kogel AJ. Tolerance of rat spinal cord to continuous interstitial irradiation. Int J Radiat Oncol Biol Phys. 1998;40(3):681689.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Perneger TV. What’s wrong with Bonferroni adjustments. BMJ. 1998;316(7139):12361238.

  • 9

    Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):4346.

  • 10

    Thiele C, Hirschfeld G. cutpointr: improved estimation and validation of optimal cutpoints in R. J Stat Softw. 2021;98(11):127.

  • 11

    Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition. Cephalalgia. 2018;38(1):1211.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Flickinger JC, Pollock BE, Kondziolka D, et al. Does increased nerve length within the treatment volume improve trigeminal neuralgia radiosurgery? A prospective double-blind, randomized study. Int J Radiat Oncol Biol Phys. 2001;51(2):449454.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Régis J, Tuleasca C, Resseguier N, et al. Long-term safety and efficacy of Gamma Knife surgery in classical trigeminal neuralgia: a 497-patient historical cohort study. J Neurosurg. 2016;124(4):10791087.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Mureb M, Golub D, Benjamin C, et al. Earlier radiosurgery leads to better pain relief and less medication usage for trigeminal neuralgia patients: an international multicenter study. J Neurosurg. 2020;135(1):237244.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand. 1971;137:311314.

  • 16

    Lindquist C, Kihlström L, Hellstrand E. Functional neurosurgery—a future for the gamma knife? Stereotact Funct Neurosurg. 1991;57(1-2):7281.

  • 17

    Rand RW, Jacques DB, Melbye RW, Copcutt BG, Levenick MN, Fisher MR. Leksell Gamma Knife treatment of tic douloureux. Stereotact Funct Neurosurg. 1993;61(suppl 1):93-102.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Gorgulho A. Radiation mechanisms of pain control in classical trigeminal neuralgia. Surg Neurol Int. 2012;3(suppl 1):S17-S25.

  • 19

    Matsuda S, Serizawa T, Nagano O, Ono J. Comparison of the results of 2 targeting methods in Gamma Knife surgery for trigeminal neuralgia. J Neurosurg. 2008;109(suppl):185189.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Park SH, Hwang SK, Kang DH, Park J, Hwang JH, Sung JK. The retrogasserian zone versus dorsal root entry zone: comparison of two targeting techniques of Gamma Knife radiosurgery for trigeminal neuralgia. Acta Neurochir (Wien). 2010;152(7):11651170.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Xu Z, Schlesinger D, Moldovan K, et al. Impact of target location on the response of trigeminal neuralgia to stereotactic radiosurgery. J Neurosurg. 2014;120(3):716724.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22

    Kondziolka D, Flickinger JC, Lunsford LD, Habeck M. Trigeminal neuralgia radiosurgery: the University of Pittsburgh experience. Stereotact Funct Neurosurg. 1996;66(suppl 1):343-348.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Longhi M, Rizzo P, Nicolato A, Foroni R, Reggio M, Gerosa M. Gamma Knife radiosurgery for trigeminal neuralgia: results and potentially predictive parameters—part I: idiopathic trigeminal neuralgia. Neurosurgery. 2007;61(6):12541261.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Kotecha R, Kotecha R, Modugula S, et al. Trigeminal neuralgia treated with stereotactic radiosurgery: the effect of dose escalation on pain control and treatment outcomes. Int J Radiat Oncol Biol Phys. 2016;96(1):142148.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    Kim YH, Kim DG, Kim JW, et al. Is it effective to raise the irradiation dose from 80 to 85 Gy in Gamma Knife radiosurgery for trigeminal neuralgia? Stereotact Funct Neurosurg. 2010;88(3):169176.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26

    Pollock BE, Phuong LK, Foote RL, Stafford SL, Gorman DA. High-dose trigeminal neuralgia radiosurgery associated with increased risk of trigeminal nerve dysfunction. Neurosurgery. 2001;49(1):5864.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Fowler JF, Welsh JS, Howard SP. Loss of biological effect in prolonged fraction delivery. Int J Radiat Oncol Biol Phys. 2004;59(1):242249.

  • 28

    Lee JY, Sandhu S, Miller D, Solberg T, Dorsey JF, Alonso-Basanta M. Higher dose rate Gamma Knife radiosurgery may provide earlier and longer-lasting pain relief for patients with trigeminal neuralgia. J Neurosurg. 2015;123(4):961968.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29

    Shetter AG, Rogers CL, Ponce F, Fiedler JA, Smith K, Speiser BL. Gamma knife radiosurgery for recurrent trigeminal neuralgia. J Neurosurg. 2002;97(5 suppl):536538.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 30

    Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han PP, Speiser BL. Gamma Knife radiosurgery for trigeminal neuralgia: the initial experience of the Barrow Neurological Institute. Int J Radiat Oncol Biol Phys. 2000;47(4):10131019.

    • PubMed
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1252 1252 141
PDF Downloads 1181 1181 59
EPUB Downloads 0 0 0