- Email: [email protected]
Trigeminal Neuralgia Treated With Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes
Accepted Manuscript Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes Rupesh Kotecha, MD, Ritesh Kotecha, MD, Sujith Modugula, BS, Erin S. Murphy, MD, Mark Jones, MD, Rajesh Kotecha, MD, Chandana A. Reddy, MS, John H. Suh, MD, Gene H. Barnett, MD, MBA, Gennady Neyman, PhD, Andre Machado, MD, PhD, Sean Nagel, MD, Samuel T. Chao, MD PII:
S0360-3016(16)30116-X
DOI:
10.1016/j.ijrobp.2016.04.013
Reference:
ROB 23557
To appear in:
International Journal of Radiation Oncology • Biology • Physics
Received Date: 24 December 2015 Revised Date:
25 March 2016
Accepted Date: 11 April 2016
Please cite this article as: Kotecha R, Kotecha R, Modugula S, Murphy ES, Jones M, Kotecha R, Reddy CA, Suh JH, Barnett GH, Neyman G, Machado A, Nagel S, Chao ST, Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes, International Journal of Radiation Oncology • Biology • Physics (2016), doi: 10.1016/ j.ijrobp.2016.04.013. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes
RI PT
Authors: Rupesh Kotecha MD1, Ritesh Kotecha MD2, Sujith Modugula BS1, Erin S. Murphy MD1,3, Mark Jones MD2, Rajesh Kotecha MD2, Chandana A. Reddy MS1, John H. Suh MD1,3, Gene H. Barnett MD, MBA3,4, Gennady Neyman PhD1,3, Andre Machado MD, PhD4, Sean Nagel MD4, and Samuel T. Chao MD1,3 Affiliations: Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH 2 MidMichigan Medical Center, Midland, MI 3 Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Cleveland Clinic, Cleveland, OH 4 Department of Neurosurgery, Neurological Institute, Cleveland Clinic, Cleveland, OH
SC
1
M AN U
*Corresponding author’s name and complete mailing address:
Phone: (216) 445-7876 Fax: (216) 445-1068 Email: [email protected]
TE D
Samuel T. Chao, MD Associate Professor, Cleveland Clinic Lerner College of Medicine Department of Radiation Oncology Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center Cleveland Clinic Main Campus, Mail Code T28 9500 Euclid Avenue Cleveland, OH 44195
EP
Running Title: Dose Effect on Trigeminal Neuralgia Outcome Note:
AC C
Presented in part at the ASTRO 57th Annual Meeting in San Antonio, TX from October 18-21, 2015. Number of text pages: 11 Number of Tables: 4 (1 Supplemental Table) Number of Figures: 2
Key Words: Trigeminal neuralgia; stereotactic radiosurgery; dose; dose-escalation; pain; numbness
ACCEPTED MANUSCRIPT
Disclosure: Conflicts of Interest:
M AN U
SC
RI PT
R. Kotecha: none R. Kotecha: none S. Modugla: none E.S. Murphy: none M. Jones: none R. Kotecha: none C.A. Reddy: none J.H. Suh: Research support Varian Medical Systems, Travel Elekta G.H. Barnett: none G. Neyman: Consulting for Elekta A. Machado: Ownership interest in Autonomic Technologies, Inc., Cardionomics, and Enspire, and a consultant relationship with Functional Neuromodulation and Spinal Modulation S. Nagel: none S.T. Chao: honorarium from Varian Medical Systems Acknowledgments: The authors greatly appreciate the help of Dennis Ouillette, RN who assisted with data collection.
AC C
EP
TE D
Cite Sources of Support (if applicable): This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
ACCEPTED MANUSCRIPT
Title: Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes
AC C
EP
TE D
M AN U
SC
RI PT
Running Title: Dose Effect on Trigeminal Neuralgia Outcome
1
ACCEPTED MANUSCRIPT
Summary This multi-institutional analysis of over 800 patients with trigeminal neuralgia with a median follow-up of approximately three years demonstrated that durable pain control can be achieved
RI PT
with stereotactic radiosurgery (SRS) doses >82 Gy to the dorsal root entry zone. Reduced efficacy is observed when SRS follows other procedures. Patients should be counseled about the
AC C
EP
TE D
M AN U
SC
risk of treatment-related facial numbness as a result of therapy.
2
ACCEPTED MANUSCRIPT
Abstract Purpose: To analyze the effect of dose-escalation on treatment outcome in patients undergoing
RI PT
stereotactic radiosurgery (SRS) for trigeminal neuralgia (TN).
Methods and Materials: A retrospective review was performed of 870 patients who underwent SRS for a diagnosis of TN from two institutions. Patients were typically treated using a single 4mm isocenter placed at the trigeminal nerve dorsal root entry zone. Patients were divided into groups based on treatment doses: ≤82 Gy (352 patients), 83-86 Gy (85 patients), and ≥90 Gy
SC
(433 patients). Pain response was classified using a categorical scoring system, with fair or poor pain control representing treatment failure. Treatment-related facial numbness was classified
M AN U
using the Barrow Neurological Institute scale. Log-rank tests were performed to test differences in time to pain failure or development of facial numbness for patients treated with different doses.
Results: Median age at first pain onset was 63 years, median age at time of SRS was 71 years, and median follow-up was 36.5 months from the time of SRS. A majority of patients (827, 95%)
TE D
were clinically diagnosed with typical TN. The 4-year rate of excellent to good pain relief was 87% (95% CI: 84-90%). The 4-year rate of pain response was 79%, 82%, and 92% in patients treated to ≤82 Gy, 83-86 Gy, and ≥90 Gy, respectively. Patients treated to doses ≤82 Gy had an increased risk of pain failure following SRS compared to patients treated to ≥90 Gy (HR 2.0,
EP
p=0.0007). Rates of treatment-related facial numbness were similar among patients treated to
AC C
doses ≥ 83 Gy. Nine patients (1%) were diagnosed with anesthesia dolorosa.
Conclusions: Dose-escalation for TN to doses >82 Gy is associated with an improvement in response to treatment and duration of pain relief. Patients treated at these doses, however, should be counseled about the increased risk of treatment-related facial numbness.
3
ACCEPTED MANUSCRIPT
Introduction: Stereotactic radiosurgery (SRS) is an effective minimally-invasive treatment for trigeminal
RI PT
neuralgia (TN) patients that alleviates pain, reduces dependence on medication, and improves quality of life [1]. Kondziolka and colleagues demonstrated improved pain control with SRS doses of ≥70 Gy. Subsequent studies have investigated the effect of the prescription dose on pain
SC
response and facial numbness [2]. Currently, a dose between 80-90 Gy is standard, though
variability exists regarding prescription dose and target location. In this study we analyzed the
M AN U
effect of dose on long-term treatment outcome using data obtained on a large number of patients treated with SRS from two institutions whose dosing parameters varied.
Materials and Methods:
We used an IRB-approved institutional database to identify 417 patients who underwent SRS for
TE D
TN from 1997-2014. We combined this with a comparable dataset from a second institution containing 453 patients treated from 2003-2014. Approval from each institution’s IRB board was
AC C
combined analysis.
EP
obtained as well as permission from each of the institutions for sharing of de-identified data for
Patients treated with Gamma Knife® (GK) (Elekta Instruments AB, Stockholm, Sweden) radiosurgery prior to 2007 were treated with a 201-source 60Co Gamma Knife® system (Model B, C™, and 4C™). After this date, patients were treated with the 192-source unit (Perfexion™) at both hospitals. A single 4-mm isocenter was typically positioned at the emergence of the trigeminal nerve (dorsal root entry zone, DREZ); 47 patients were treated with concentrically aimed and equally weighted 4- and 8-mm isocenters on an institutional protocol [3]. The doses 4
ACCEPTED MANUSCRIPT
varied over this time period from 70-95 Gy prescribed to the 100% isodose line at both institutions. For patients treated prior to September 2000, the prescribed doses were re-calculated (from 75 Gy to 82 Gy) using the corrected output factor [3]. An output factor of 0.87 was utilized
RI PT
after this period up to 2007 on the 201-source unit. After installation of the upgraded 192-source unit, an output factor of 0.814 was used [4].
SC
Basic patient information including age, gender, typical or atypical TN symptoms [5], date of first pain onset, date of SRS, diagnosis of multiple sclerosis (MS), history of prior surgical
M AN U
procedures, and distribution of trigeminal nerve symptomatology was obtained. Radiosurgery treatment data were abstracted. Pertinent follow-up data were recorded including the dates of follow-up, pain response, and treatment-related facial numbness.
TE D
Outcome measures in this study were limited to pain response and facial numbness. Pain response was characterized using the Barrow Neurological Institute (BNI) scale: class I (no trigeminal pain, no medication), class II (occasional pain, not requiring medication), class III (no
EP
pain with continued medication or pain controlled with medication), class IV (some pain, not adequately controlled with medication), and class V (severe pain) at institution #1 and also coded
AC C
to a simplified four group EGFP (Excellent, Good, Fair, Poor) categorical scoring system (coded at institution #2) for combined data analysis [6]. In the four group EGFP categorical system, excellent was scored if the patient was pain-free and off medications, good if they had rare pain or were pain-free on doses of medications not producing side-effects, fair if they had persistent pain but less severe than prior to treatment, or poor if they had no significant response to therapy. For treatment-related facial numbness, both institutions used the BNI facial numbness scale:
5
ACCEPTED MANUSCRIPT
class I (no facial numbness), class II (mild facial numbness, not bothersome), class III (facial numbness, somewhat bothersome), and class IV (facial numbness, very bothersome) [6].
RI PT
Actuarial likelihoods of pain control and facial numbness were evaluated according to the
product-limit estimate (Kaplan-Meier) method. Patients were designated into three dose groups: ≤ 82 Gy, 83-86 Gy, and ≥90 Gy. Patients were classified as having pain failure if their pain was
SC
recorded to be fair or poorly controlled using the four category system. Treatment-related facial numbness was defined as BNI class III or IV. For both pain failure and numbness, patients
M AN U
without the respective endpoint were censored at time of last follow-up. Logrank (Mantel-Cox) tests were performed to test differences in time to pain failure or development of facial numbness for patients treated with different doses. A univariate analysis was performed to determine the effect of age, gender, typical vs. atypical symptoms, MS, history of prior surgical procedures,
TE D
and time between initial onset of TN symptoms and SRS treatment date on treatment outcomes. Statistically significant factors were then analyzed in a multivariable model using the backwards method. Given overlapping variables (prior procedure and number of procedures), two
EP
multivariable models were constructed. Data analysis was performed using SAS version 9.4
AC C
(SAS Institute, Cary, NC). The threshold of statistical significance was set at p<0.05.
Results:
The patient demographics and baseline characteristics are presented in Table 1. A total of 870 patients were treated across the two institutions. The only significant demographic variable differing between the institutions was age at first pain onset (median 60 vs. 66 years, see Supplemental Table 1). The majority of the patients had typical TN (827 patients, 95%); 78
6
ACCEPTED MANUSCRIPT
patients (9%) had a diagnosis of MS. The median age at first onset of symptoms was 63 years (Range [R]: 11-96 years) and the median age at the time of the SRS was 71 years (R: 25-98 years). The median time from first pain episode to SRS treatment was 48.7 months (R: 0-610
RI PT
months). SRS was the primary treatment for 637 patients (73%) while 233 patients (27%) had undergone ≥1 prior procedure.
SC
The 870 patients were divided into three dose groups for comparative analysis: 352 patients (40%) were treated to ≤ 82 Gy, 85 patients (10%) to 83-86 Gy, and 433 patients (50%) to ≥ 90
M AN U
Gy. The median dose prescribed was 86 Gy to the 100% isodose line and the median doses prescribed in the three dose cohorts were 82 Gy (R: 70-82 Gy), 86 Gy (R: 83-86 Gy), and 90 Gy (R: 90-95 Gy). The median follow-up was 36.5 months (R: 0-207.9 months). The median followup time for patients treated to 83-86 Gy was shorter (6.6 months) than the median follow-up for
TE D
patients treated to ≤ 82 Gy (22.4 months) or ≥90 Gy (48.7 months).
In total, 69 patients (8%) did not have a pain response recorded at follow-up and these patients
EP
were omitted from the analysis for pain response, though other variables and outcomes may have been collected. An improvement in pain after SRS was noted in 747 patients (86% of the total
AC C
group, 93% of patients with follow-up data regarding pain response). The 4-year rate of excellent to good pain relief across all patients was 86.7% (95% CI: 83.6-89.8%). There was a stepwise improvement in the 4-year rate of excellent or good pain response with increasing dose: 79.0% (95% CI: 73.1-85.0%), 81.6% (95% CI: 63.4-99.9%), and 92.0% (95% CI: 88.7%-95.4%) in patients treated to ≤82 Gy, 83-86 Gy, and ≥90 Gy, respectively (p=0.0019, Figure 1), and patients treated to ≥90 Gy had longer times to pain failure compared to patients treated to ≤82
7
ACCEPTED MANUSCRIPT
Gy (p=0.0019). For patients with MS or atypical trigeminal neuralgia, the 4-year rate of excellent or good pain response was 71.6% (95% CI: 56.4-86.8%) and 85.3% (95% CI: 72.0-98.7%) for those treated to ≤82 Gy and ≥90 Gy, respectively (p=0.21, Supplemental Figure 1). For patients
RI PT
who had previously undergone at least one procedure, the 4-year rate of excellent or good pain response was 75.6% (95% CI: 65.4-85.8%) and 82.5% (95% CI: 71.4-93.6%) for those treated to
SC
≤82 Gy and ≥90 Gy, respectively (p=0.24).
Univariate analysis revealed that dose, age at time of SRS, and a history of prior procedures were
M AN U
predictors of pain failure following SRS. Specifically, patients treated to lower doses (≤82 Gy) were at higher risk of a having a recurrence of their pain (HR: 2.0, 95% CI: 1.3-3.0) compared to patients treated to ≥90 Gy (p=0.0007). Although patients treated to doses from 83-86 Gy also appeared to experience a higher rate of pain failure (HR: 2.2, 95% CI: 0.8-6.1), this result was
TE D
not statistically significant (p=0.15). Additionally, patients who had a prior surgical procedure were at higher risk for pain failure after SRS (HR: 1.70, 95% CI: 1.12-2.52). When the number of prior procedures (R: 1-6) were analyzed as a continuous variable, they were also associated
EP
with an increased risk of pain failure (HR: 1.28, 95% CI: 1.1-1.5). Multivariate analysis confirmed that prescribed dose influenced pain control outcomes (p=0.008) and in a second
AC C
multivariable model examining the effect of the number of procedures prior to SRS, an increasing number of procedures was also significantly associated with an increased risk of pain failure following SRS (p=0.033) (Table 2). Time between the first onset of symptoms to SRS did not appear to influence pain control.
8
ACCEPTED MANUSCRIPT
Follow-up records for the assessment of facial numbness were available for 802 patients (92%). The worst facial numbness score was class I in 379 patients (43%), class II in 170 patients (20%), class III in 175 patients (20%), and class IV in 78 patients (9%). The 5-year rate of
RI PT
freedom from class III or class IV facial numbness was 58.4% (95% CI: 53.9-62.8%). The 4-year rates of freedom from class III/IV numbness were similar among patients treated to 83-86 Gy (50.7%, 95% CI: 20.7-80.8%) and ≥90 Gy (59.7%, 95% CI: 54.3-65.2%) and were significantly
SC
lower than the 4-year rate for patients treated to the ≤82 Gy (74.9%, 95% CI: 68.8-80.9%)
M AN U
(p=0.0056, Figure 2).
On univariate analysis, patients treated to doses ≥90 Gy were at higher risk of developing a class III/IV facial numbness (HR: 1.53, 95% CI: 1.16-2.04) compared to patients treated to the lowest doses (≤ 82 Gy, p=0.0028). Gender, age at time of GK, MS, or history of a previous surgical
TE D
procedure did not relate to risk of treatment-related facial numbness (see Table 3).
At last follow-up, 539 patients (62%) reported excellent and 162 patients (19%) reported good
EP
pain control. Similarly, at last follow-up, 511 patients (59%), 130 patients (15%), 112 patients (13%), and 45 patients (5%) experienced class I-IV facial numbness, respectively. Nine patients
AC C
(1%) were diagnosed with anesthesia dolorosa, all of whom were treated to doses ≤86 Gy.
Discussion:
In our analysis of over 800 patients with a median follow-up of approximately three years, we concluded that pain control outcomes are best when SRS is the first procedure in the management of TN and doses >82 Gy are prescribed to the DREZ. Moreover, patients can
9
ACCEPTED MANUSCRIPT
expect favorable outcomes with treatment regardless of the length of symptoms prior to radiosurgery. They must be cautioned, however, that efficacy is reduced when SRS follows other procedures. Patients should also be counseled about the risks of treatment-related facial
RI PT
numbness as a result of therapy.
The dose for SRS was first established by a multi-institutional study in which 50 patients were
SC
treated at five centers to doses ranging from 60-90 Gy prescribed to the 100% isodose line using a single 4-mm isocenter targeted at the nerve root entry zone [2]. A dose of ≥70 Gy was
M AN U
associated with a significantly greater chance of pain relief (72% vs. 9%, p=0.0003). Limitations of applicability of this data include lack of control for other factors influencing treatment outcome, such as age or a history of a prior procedure. Moreover, the prescribed doses have gradually increased over time, with some studies suggesting a benefit to doses as high as 90 Gy
TE D
(although caution should be used for doses exceeding 90 Gy) [9,10,11]. In general, there is a significant variation in prescription doses across institutions, as is best illustrated when reviewing the literature for large experiences with comparably long-term treatment outcomes
EP
(Table 4) [12-16]. In general, significant heterogeneity in patient selection, previous neurosurgical procedures, type of trigeminal neuralgia, prescription doses, and target placement
AC C
limit the extent of cross-comparisons. Besides these variations, the types of outcomes data reported, including the scales used, and the management style of post-treatment recurrences and complications are institutionally dependent and this adds considerable variation to the data pool. However, consideration should be given to a larger multi-institutional analysis to determine the effect of dose on treatment outcome using standardized and validated scales, ideally in a prospective manner.
10
ACCEPTED MANUSCRIPT
In addition to a higher chance of pain relief, there are other advantages to dose escalation. Kim et al. reported a more rapid response to treatment in 44 patients treated to a dose of 85 Gy when retrospectively compared to 66 patients treated to a dose of 80 Gy; however, they did not observe
RI PT
a significant difference in pain control rate between the two groups [17]. Moreover, MorbidniGaffney et al. observed a longer duration of pain response in patients treated to doses exceeding 85 Gy [18]. Our results suggest a significant long-term benefit to dose escalation, with step-wise
SC
improvements in rates of pain response at 4-years with increasing dose: from 79% for those patients treated to doses ≤82 Gy compared to 92% in those treated to doses ≥90 Gy. At eight
M AN U
years post SRS, with approximately 50 patients in both the lowest and highest dose cohorts, higher rates of pain relief were still noted in patients treated to ≥90 Gy (78%, 95% CI: 71.484.4%) vs. (67%, 95% CI: 59.0-75.6%) for patients treated to ≤82 Gy. Collectively, these retrospective reports and the results of our study suggest a benefit with regards to dose escalation
TE D
in terms of likelihood of pain response and duration of symptom relief.
There is an ultimate trade-off to dose escalation and multiple series have cautioned about
EP
increased risk for treatment-related morbidity. For example, Marshall and colleagues reported on a series of 448 patients treated to doses ranging from 80 to 97 Gy, where the mean maximum
AC C
prescribed dose was 88 Gy and the majority of patients received 90 Gy [19]. The mean dose to the dorsal root entry zone was 57.6 Gy in those who developed facial numbness compared to 47.3 Gy in those who did not develop numbness (p=0.02). In the present series, patients treated to doses >82 Gy were at increased risk for treatment-related numbness. Interestingly, this risk was not significantly higher in patients treated to doses ≥90 Gy vs. patients treated to 83-86 Gy. Since there was a benefit with dose escalation beyond 83 Gy but no significant increase in facial
11
ACCEPTED MANUSCRIPT
numbness, doses up to 90 Gy to the DREZ may be safe for clinical practice if mindful of the isocenter placement and brainstem tolerance and as long as patients are counseled about the risks of facial numbness. Longer follow-up in patients treated to doses ranging from 83 to 90 Gy may
RI PT
help determine the optimal dose which provides a favorable balance between pain control and treatment-related facial numbness.
SC
Dose escalation to 90 Gy to the DREZ may be considered when treating patients with apriori risk factors for pain failure following radiosurgery. In our series, the total number of prior procedures
M AN U
before SRS was significant on multivariable analysis as a predictor for early pain failure. When using a similar four-category pain response scale, Flickinger and colleagues also observed a similar relationship with rates of complete pain relief of 75%, 70%, and 63% at a median followup of 26 months in patients who had previously undergone 0, 1, and ≥2 prior neurosurgical
TE D
procedures, respectively [8]. On multivariable analysis, we did not observe a statistically significant improvement in freedom from pain failure with dose escalation in patients who had received ≥1 procedure; however, dose escalation may be considered in patients who have had
EP
multiple prior procedures and are at the highest risk for early pain failure. Determining this threshold will be key to improving outcomes for these patients and should be studied in a larger
AC C
multi-institutional analysis with adequate power.
We acknowledge several limitations to our study. Foremost, when performing multi-institutional studies, there always exists a risk for differences in data collection and outcomes assessment between institutions. We tried to limit this by using a four category classification system for pain response, albeit with a tradeoff of reduced sensitivity. The BNI scale is most commonly used in
12
ACCEPTED MANUSCRIPT
evaluation of trigeminal neuralgia patients, however, it has yet to be empirically validated. It contains as many as 12 permutations of outcomes (four pain categories and three medication-use categories) and was thought to be too complicated for a retrospective multi-institutional analysis.
RI PT
Therefore, we used a simplified classification system, in which pain response and medication use could be divided into four categories. Simplified scales such as this have been used in similar multi-institutional and large single institution studies, though we recognize that the best data
SC
would be collected prospectively using a sensitive patient-reported test obtained at baseline and at standard follow-up appointments [8]. Also, use of a categorical system in this series (similar to
M AN U
the BNI) as opposed to a quantifiable pain response scale prevents any quantitative estimations of benefit and limits analysis to dichotomous outcomes (pain relief vs. pain failure). To help improve data coding reliability, we limited coding to a few individuals. Given the lack of a true threshold for initiation of salvage procedures and the heterogeneity in re-treatment approaches,
TE D
we only used pain failure or facial numbness to evaluate SRS treatment efficacy and were not able to evaluate potential benefits of salvage therapies. Furthermore, as this was a retrospective study, a standardized follow-up schedule was not mandated. Considering the clinical complexity
EP
regarding patient symptoms from TN, the primary endpoints are often based on subjective outcomes which make data analysis difficult. Therefore, clinical questions regarding treatment
AC C
delivery techniques or comparisons of treatment parameters would be best addressed by prospectively collected data with quality-of-life endpoints, medication diaries, and patient satisfaction outcomes [20].
Conclusion:
13
ACCEPTED MANUSCRIPT
Our study demonstrates that SRS is an effective treatment option for patients with TN that provides durable pain relief. Patients should be counseled about reduced efficacy if they have undergone multiple prior procedures. In addition, pain control outcomes are best when patients
RI PT
are treated to >82 Gy to the DREZ, however, they must be counseled about the risks of
treatment-related facial numbness. We hope that future studies will analyze additional variables
AC C
EP
TE D
M AN U
SC
to help improve the outcomes in patients with this complex neurological disorder.
14
ACCEPTED MANUSCRIPT
Figure Legends:
dose group: ≤ 82 Gy, 83-86 Gy, and ≥ 90 Gy.
RI PT
Figure 1. Kaplan-Meier curve of freedom from pain failure (fair or poor pain control) based on
Figure 2. Kaplan-Meier curve of freedom from Barrow Neurologic Institute class III or IV
AC C
EP
TE D
M AN U
SC
treatment-related facial numbness based on dose group: ≤ 82 Gy, 83-86 Gy, and ≥ 90 Gy.
15
ACCEPTED MANUSCRIPT
References 1. Regis J, Metellus P, Hayashi M, Roussel P, Donnet A, Bille-Turc F. Prospective controlled trial of gamma knife surgery for essential trigeminal neuralgia. J Neurosurg
RI PT
2006;104:913-924.
2. Kondziolka D, Lunsford LD, Flickinger JC, Young RF, Vermeulen S, Duma CM, Jacques DB, Rand RW, Regis J, Peragut JC, Manera L, Epstein MH, Lindquist C. Stereotactic
SC
radiosurgery for trigeminal neuralgia: A multiinstitutional study using the gamma unit. J
3. XXXXX
M AN U
Neurosurg 1996;84:940-945.
4. Tsai JS, Rivard MJ, Engler MJ, Mignano JE, Wazer DE, Shucart WA. Determination of the 4 mm gamma knife helmet relative output factor using a variety of detectors. Medical physics 2003;30:986-992.
TE D
5. Burchiel KJ. A new classification for facial pain. Neurosurgery 2003;53:1164-1166; discussion 1166-1167.
6. Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han PP, Speiser BL. Gamma knife
EP
radiosurgery for trigeminal neuralgia: The initial experience of the barrow neurological institute. International journal of radiation oncology, biology, physics 2000;47:1013-
AC C
1019.
7. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta chirurgica Scandinavica 1971;137:311-314.
8. Flickinger JC, Pollock BE, Kondziolka D, Phuong LK, Foote RL, Stafford SL, Lunsford LD. Does increased nerve length within the treatment volume improve trigeminal neuralgia
16
ACCEPTED MANUSCRIPT
radiosurgery? A prospective double-blind, randomized study. International journal of radiation oncology, biology, physics 2001;51:449-454. 9. Pollock BE, Phuong LK, Gorman DA, Foote RL, Stafford SL. Stereotactic radiosurgery for
RI PT
idiopathic trigeminal neuralgia. J Neurosurg 2002;97:347-353.
10. Kondziolka D, Lacomis D, Niranjan A, Mori Y, Maesawa S, Fellows W, Lunsford LD. Histological effects of trigeminal nerve radiosurgery in a primate model: Implications for
SC
trigeminal neuralgia radiosurgery. Neurosurgery 2000;46:971-976; discussion 976-977. 11. Longhi M, Rizzo P, Nicolato A, Foroni R, Reggio M, Gerosa M. Gamma knife radiosurgery
M AN U
for trigeminal neuralgia: Results and potentially predictive parameters--part i: Idiopathic trigeminal neuralgia. Neurosurgery 2007;61:1254-1260; discussion 1260-1251. 12. Verheul JB, Hanssens PE, Lie ST, Leenstra S, Piersma H, Beute GN. Gamma knife surgery for trigeminal neuralgia: A review of 450 consecutive cases. J Neurosurg 2010;113
TE D
Suppl:160-167.
13. Kondziolka D, Zorro O, Lobato-Polo J, Kano H, Flannery TJ, Flickinger JC, Lunsford LD. Gamma knife stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg
EP
2010;112:758-765.
14. Young B, Shivazad A, Kryscio RJ, St Clair W, Bush HM. Long-term outcome of high-dose
AC C
gamma knife surgery in treatment of trigeminal neuralgia. J Neurosurg 2013;119:11661175.
15. Lucas JT, Jr., Nida AM, Isom S, Marshall K, Bourland JD, Laxton AW, Tatter SB, Chan MD. Predictive nomogram for the durability of pain relief from gamma knife radiation surgery in the treatment of trigeminal neuralgia. International journal of radiation oncology, biology, physics 2014;89:120-126.
17
ACCEPTED MANUSCRIPT
16. Regis J, Tuleasca C, Resseguier N, Carron R, Donnet A, Gaudart J, Levivier M. Long-term safety and efficacy of gamma knife surgery in classical trigeminal neuralgia: A 497patient historical cohort study. J Neurosurg 2015:1-9.
RI PT
17. Kim YH, Kim DG, Kim JW, Kim YH, Han JH, Chung HT, Paek SH. Is it effective to raise the irradiation dose from 80 to 85 gy in gamma knife radiosurgery for trigeminal neuralgia? Stereotactic and functional neurosurgery 2010;88:169-176.
SC
18. Morbidini-Gaffney S, Chung CT, Alpert TE, Newman N, Hahn SS, Shah H, Mitchell L, Bassano D, Darbar A, Bajwa SA, Hodge C. Doses greater than 85 gy and two isocenters
M AN U
in gamma knife surgery for trigeminal neuralgia: Updated results. J Neurosurg 2006;105 Suppl:107-111.
19. Marshall K, Chan MD, McCoy TP, Aubuchon AC, Bourland JD, McMullen KP, deGuzman AF, Munley MT, Shaw EG, Tatter SB, Ellis TL. Predictive variables for the successful
TE D
treatment of trigeminal neuralgia with gamma knife radiosurgery. Neurosurgery 2012;70:566-572; discussion 572-563.
20. Zakrzewska JM, Lopez BC. Quality of reporting in evaluations of surgical treatment of
EP
trigeminal neuralgia: Recommendations for future reports. Neurosurgery 2003;53:110-
AC C
120; discussion 120-112.
18
ACCEPTED MANUSCRIPT
≤ 82 Gy N (%) 352 (40%)
≥ 90 Gy N (%) 433 (50%)
p-value
60 30-90
RI PT
60 11-89
66 18-96
0.2427
69 30-96
71 25-98
SC
71 28-93
0.4596
83 (98%) 2 (2%)
409 (95%) 24 (6%)
M AN U
335 (95%) 17 (5%) 204 (58%) 148 (42%) 0 (0%) 314 (89%) 38 (11%)
48 (56%) 37 (43%) 0 (0%) 76 (89%) 9 (11%)
0.0650
281 (65%) 149 (34%) 3 (1%) 0.1782
402 (93%) 31 (7%)
211 (60%) 140 (40%) 1 (0.3%)
56 (66%) 29 (34%) 0 (0%)
257 (60%) 176 (41%) 0 (0%)
108 (31%) 255 (73%) 212 (60%)
25 (29%) 67 (79%) 47 (55%)
103 (24%) 311 (72%) 239 (55%)
EP
AC C
Abbreviations NA = not available yrs = years
83-86 Gy N (%) 85 (10%)
<0.0001
TE D
Table 1. Patient characteristics All Patients N (%) Number of Patients 870 Age at first pain onset Median (yrs) 63 Range (yrs) 11-96 Age at stereotactic radiosurgery Median (yrs) 71 Range (yrs) 25-98 Typical vs. Atypical Symptoms Typical 827 (95%) Atypical 43 (5%) Gender Female 533 (61%) Male 334 (38%) NA 3 (0.3%) Multiple Sclerosis No 792 (91%) Yes 78 (9%) Right vs. Left Right 524 (60%) Left 345 (40%) NA 1 (0.1%) Trigeminal nerve distribution I 236 (27%) II 633 (73%) III 498 (57%)
0.6008
0.0855 0.4095 0.0340
ACCEPTED MANUSCRIPT
Table 2. Cox proportional hazards regression analysis of predictors of pain failure following stereotactic radiosurgery
1.755 0.824 2.131 0.988 1.478 -
RI PT
0.001 0.888 0.151 0.399 0.068 0.639 0.126 0.043 0.560 0.011 0.005
95% CI Lower Upper P value Multivariable analysis 1.157 0.290 0.749 0.972 0.976 -
SC
2.986 2.620 6.123 1.791 3.752 2.179 1.003 1.000 1.002 2.529 1.523
Hazard ratio
M AN U
1.336 0.329 0.757 0.793 0.953 0.620 0.975 0.969 0.997 1.127 1.078
EP
Abbreviations: CI = confidence interval TN = trigeminal neuralgia SRS = stereotactic radiosurgery
1.997 0.928 2.152 1.192 1.891 1.163 0.989 0.984 0.999 1.689 1.281
AC C
Dose Group ≤ 82 Gy vs. ≥ 90 Gy ≤ 82 Gy vs. 83-86 Gy 83-86 Gy vs. ≥ 90 Gy Gender (F vs. M) Atypical vs. typical TN Multiple sclerosis Age at first pain onset Age at time of SRS Time from first pain onset to SRS SRS after previous procedure No. of previous procedures prior to SRS
95% CI Lower Upper P value Univariate analysis
TE D
Hazard ratio
2.663 2.339 6.067 1.003 2.239 -
0.008 0.716 0.156 0.118 0.0652 -
Hazard ratio
95% CI Lower Upper P value Multivariable analysis
1.777 0.872 2.039 0.988
1.175 0.308 0.715 0.972
2.687 2.464 5.815 1.003
0.006 0.795 0.183 0.120
1.215
1.016
1.452
0.033
ACCEPTED MANUSCRIPT
95% CI
SC
1.157 0.929 0.442 0.754 0.583 0.778 0.999 0.999 0.999 0.607 0.775
Upper
M AN U
1.534 1.821 0.842 0.972 0.982 1.204 1.008 1.009 1.000 0.813 0.910
TE D
AC C
Abbreviations: CI = confidence interval TN = trigeminal neuralgia SRS = stereotactic radiosurgery
Lower
EP
Univariate analysis covariates Dose Group ≥ 90 Gy vs. ≤ 82 Gy 83-86 Gy vs. ≤ 82 Gy ≥ 90 Gy vs. 83-86 Gy Gender (F vs. M) Atypical vs. typical TN Multiple sclerosis Age at first pain onset Age at time of SRS Time from first pain onset to SRS SRS after previous procedure No. of previous procedures prior to SRS
Hazard ratio
RI PT
Table 3. Cox proportional hazards regression analysis of predictors of facial numbness following stereotactic radiosurgery
2.033 3.571 1.603 1.254 1.655 1.866 1.017 1.020 1.002 1.089 1.067
P value
0.003 0.081 0.601 0.828 0.945 0.405 0.101 0.074 0.831 0.165 0.245
ACCEPTED MANUSCRIPT
Table 4. Selected large institutional experiences of stereotactic radiosurgery for trigeminal neuralgia
Young et al. Lucas et al. Régis et al. Present study
Median FUP (months)
Dose Range
Pain Control
Pain Control Scale
Freedom from facial numbness
Facial numbness scale
2010
365
28.0
80 Gy
58.0%1
BNI
93.5%2
BNI
BNI
89.5%
2
BNI
67.1%
2
WBFS
2010 2013 2014 2015 2015
503 315 446 497 870
24.0
60 - 90 Gy
58.8
90 Gy
21.2
80 - 97 Gy
43.8
70 - 90 Gy
36.5
70 - 95 Gy
46.0%
1
71.4%
2
46.9%
1
64.9%
1
84.0%
1
RI PT
Kondziolka et al.
No. of patients
MMS
SC
Vergeul et al.
Year
M AN U
Series Author
BNI BNI
EGFP
-
-
79.6%
1
BNI
58.4%
1
BNI
AC C
EP
TE D
EGFP = Excellent, Good, Fair, Poor categorical scoring system, BNI = Barrow Neurologic Institute, MMS = Modified Marseille Scale, WBFS = Wong-Baker FACES scale 1 = 5-year rates, 2 = crude-rate, 3 = 4-year rates
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0360-3016(16)30116-X
DOI:
10.1016/j.ijrobp.2016.04.013
Reference:
ROB 23557
To appear in:
International Journal of Radiation Oncology • Biology • Physics
Received Date: 24 December 2015 Revised Date:
25 March 2016
Accepted Date: 11 April 2016
Please cite this article as: Kotecha R, Kotecha R, Modugula S, Murphy ES, Jones M, Kotecha R, Reddy CA, Suh JH, Barnett GH, Neyman G, Machado A, Nagel S, Chao ST, Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes, International Journal of Radiation Oncology • Biology • Physics (2016), doi: 10.1016/ j.ijrobp.2016.04.013. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes
RI PT
Authors: Rupesh Kotecha MD1, Ritesh Kotecha MD2, Sujith Modugula BS1, Erin S. Murphy MD1,3, Mark Jones MD2, Rajesh Kotecha MD2, Chandana A. Reddy MS1, John H. Suh MD1,3, Gene H. Barnett MD, MBA3,4, Gennady Neyman PhD1,3, Andre Machado MD, PhD4, Sean Nagel MD4, and Samuel T. Chao MD1,3 Affiliations: Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH 2 MidMichigan Medical Center, Midland, MI 3 Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Cleveland Clinic, Cleveland, OH 4 Department of Neurosurgery, Neurological Institute, Cleveland Clinic, Cleveland, OH
SC
1
M AN U
*Corresponding author’s name and complete mailing address:
Phone: (216) 445-7876 Fax: (216) 445-1068 Email: [email protected]
TE D
Samuel T. Chao, MD Associate Professor, Cleveland Clinic Lerner College of Medicine Department of Radiation Oncology Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center Cleveland Clinic Main Campus, Mail Code T28 9500 Euclid Avenue Cleveland, OH 44195
EP
Running Title: Dose Effect on Trigeminal Neuralgia Outcome Note:
AC C
Presented in part at the ASTRO 57th Annual Meeting in San Antonio, TX from October 18-21, 2015. Number of text pages: 11 Number of Tables: 4 (1 Supplemental Table) Number of Figures: 2
Key Words: Trigeminal neuralgia; stereotactic radiosurgery; dose; dose-escalation; pain; numbness
ACCEPTED MANUSCRIPT
Disclosure: Conflicts of Interest:
M AN U
SC
RI PT
R. Kotecha: none R. Kotecha: none S. Modugla: none E.S. Murphy: none M. Jones: none R. Kotecha: none C.A. Reddy: none J.H. Suh: Research support Varian Medical Systems, Travel Elekta G.H. Barnett: none G. Neyman: Consulting for Elekta A. Machado: Ownership interest in Autonomic Technologies, Inc., Cardionomics, and Enspire, and a consultant relationship with Functional Neuromodulation and Spinal Modulation S. Nagel: none S.T. Chao: honorarium from Varian Medical Systems Acknowledgments: The authors greatly appreciate the help of Dennis Ouillette, RN who assisted with data collection.
AC C
EP
TE D
Cite Sources of Support (if applicable): This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
ACCEPTED MANUSCRIPT
Title: Trigeminal Neuralgia Treated with Stereotactic Radiosurgery: The Effect of Dose Escalation on Pain Control and Treatment Outcomes
AC C
EP
TE D
M AN U
SC
RI PT
Running Title: Dose Effect on Trigeminal Neuralgia Outcome
1
ACCEPTED MANUSCRIPT
Summary This multi-institutional analysis of over 800 patients with trigeminal neuralgia with a median follow-up of approximately three years demonstrated that durable pain control can be achieved
RI PT
with stereotactic radiosurgery (SRS) doses >82 Gy to the dorsal root entry zone. Reduced efficacy is observed when SRS follows other procedures. Patients should be counseled about the
AC C
EP
TE D
M AN U
SC
risk of treatment-related facial numbness as a result of therapy.
2
ACCEPTED MANUSCRIPT
Abstract Purpose: To analyze the effect of dose-escalation on treatment outcome in patients undergoing
RI PT
stereotactic radiosurgery (SRS) for trigeminal neuralgia (TN).
Methods and Materials: A retrospective review was performed of 870 patients who underwent SRS for a diagnosis of TN from two institutions. Patients were typically treated using a single 4mm isocenter placed at the trigeminal nerve dorsal root entry zone. Patients were divided into groups based on treatment doses: ≤82 Gy (352 patients), 83-86 Gy (85 patients), and ≥90 Gy
SC
(433 patients). Pain response was classified using a categorical scoring system, with fair or poor pain control representing treatment failure. Treatment-related facial numbness was classified
M AN U
using the Barrow Neurological Institute scale. Log-rank tests were performed to test differences in time to pain failure or development of facial numbness for patients treated with different doses.
Results: Median age at first pain onset was 63 years, median age at time of SRS was 71 years, and median follow-up was 36.5 months from the time of SRS. A majority of patients (827, 95%)
TE D
were clinically diagnosed with typical TN. The 4-year rate of excellent to good pain relief was 87% (95% CI: 84-90%). The 4-year rate of pain response was 79%, 82%, and 92% in patients treated to ≤82 Gy, 83-86 Gy, and ≥90 Gy, respectively. Patients treated to doses ≤82 Gy had an increased risk of pain failure following SRS compared to patients treated to ≥90 Gy (HR 2.0,
EP
p=0.0007). Rates of treatment-related facial numbness were similar among patients treated to
AC C
doses ≥ 83 Gy. Nine patients (1%) were diagnosed with anesthesia dolorosa.
Conclusions: Dose-escalation for TN to doses >82 Gy is associated with an improvement in response to treatment and duration of pain relief. Patients treated at these doses, however, should be counseled about the increased risk of treatment-related facial numbness.
3
ACCEPTED MANUSCRIPT
Introduction: Stereotactic radiosurgery (SRS) is an effective minimally-invasive treatment for trigeminal
RI PT
neuralgia (TN) patients that alleviates pain, reduces dependence on medication, and improves quality of life [1]. Kondziolka and colleagues demonstrated improved pain control with SRS doses of ≥70 Gy. Subsequent studies have investigated the effect of the prescription dose on pain
SC
response and facial numbness [2]. Currently, a dose between 80-90 Gy is standard, though
variability exists regarding prescription dose and target location. In this study we analyzed the
M AN U
effect of dose on long-term treatment outcome using data obtained on a large number of patients treated with SRS from two institutions whose dosing parameters varied.
Materials and Methods:
We used an IRB-approved institutional database to identify 417 patients who underwent SRS for
TE D
TN from 1997-2014. We combined this with a comparable dataset from a second institution containing 453 patients treated from 2003-2014. Approval from each institution’s IRB board was
AC C
combined analysis.
EP
obtained as well as permission from each of the institutions for sharing of de-identified data for
Patients treated with Gamma Knife® (GK) (Elekta Instruments AB, Stockholm, Sweden) radiosurgery prior to 2007 were treated with a 201-source 60Co Gamma Knife® system (Model B, C™, and 4C™). After this date, patients were treated with the 192-source unit (Perfexion™) at both hospitals. A single 4-mm isocenter was typically positioned at the emergence of the trigeminal nerve (dorsal root entry zone, DREZ); 47 patients were treated with concentrically aimed and equally weighted 4- and 8-mm isocenters on an institutional protocol [3]. The doses 4
ACCEPTED MANUSCRIPT
varied over this time period from 70-95 Gy prescribed to the 100% isodose line at both institutions. For patients treated prior to September 2000, the prescribed doses were re-calculated (from 75 Gy to 82 Gy) using the corrected output factor [3]. An output factor of 0.87 was utilized
RI PT
after this period up to 2007 on the 201-source unit. After installation of the upgraded 192-source unit, an output factor of 0.814 was used [4].
SC
Basic patient information including age, gender, typical or atypical TN symptoms [5], date of first pain onset, date of SRS, diagnosis of multiple sclerosis (MS), history of prior surgical
M AN U
procedures, and distribution of trigeminal nerve symptomatology was obtained. Radiosurgery treatment data were abstracted. Pertinent follow-up data were recorded including the dates of follow-up, pain response, and treatment-related facial numbness.
TE D
Outcome measures in this study were limited to pain response and facial numbness. Pain response was characterized using the Barrow Neurological Institute (BNI) scale: class I (no trigeminal pain, no medication), class II (occasional pain, not requiring medication), class III (no
EP
pain with continued medication or pain controlled with medication), class IV (some pain, not adequately controlled with medication), and class V (severe pain) at institution #1 and also coded
AC C
to a simplified four group EGFP (Excellent, Good, Fair, Poor) categorical scoring system (coded at institution #2) for combined data analysis [6]. In the four group EGFP categorical system, excellent was scored if the patient was pain-free and off medications, good if they had rare pain or were pain-free on doses of medications not producing side-effects, fair if they had persistent pain but less severe than prior to treatment, or poor if they had no significant response to therapy. For treatment-related facial numbness, both institutions used the BNI facial numbness scale:
5
ACCEPTED MANUSCRIPT
class I (no facial numbness), class II (mild facial numbness, not bothersome), class III (facial numbness, somewhat bothersome), and class IV (facial numbness, very bothersome) [6].
RI PT
Actuarial likelihoods of pain control and facial numbness were evaluated according to the
product-limit estimate (Kaplan-Meier) method. Patients were designated into three dose groups: ≤ 82 Gy, 83-86 Gy, and ≥90 Gy. Patients were classified as having pain failure if their pain was
SC
recorded to be fair or poorly controlled using the four category system. Treatment-related facial numbness was defined as BNI class III or IV. For both pain failure and numbness, patients
M AN U
without the respective endpoint were censored at time of last follow-up. Logrank (Mantel-Cox) tests were performed to test differences in time to pain failure or development of facial numbness for patients treated with different doses. A univariate analysis was performed to determine the effect of age, gender, typical vs. atypical symptoms, MS, history of prior surgical procedures,
TE D
and time between initial onset of TN symptoms and SRS treatment date on treatment outcomes. Statistically significant factors were then analyzed in a multivariable model using the backwards method. Given overlapping variables (prior procedure and number of procedures), two
EP
multivariable models were constructed. Data analysis was performed using SAS version 9.4
AC C
(SAS Institute, Cary, NC). The threshold of statistical significance was set at p<0.05.
Results:
The patient demographics and baseline characteristics are presented in Table 1. A total of 870 patients were treated across the two institutions. The only significant demographic variable differing between the institutions was age at first pain onset (median 60 vs. 66 years, see Supplemental Table 1). The majority of the patients had typical TN (827 patients, 95%); 78
6
ACCEPTED MANUSCRIPT
patients (9%) had a diagnosis of MS. The median age at first onset of symptoms was 63 years (Range [R]: 11-96 years) and the median age at the time of the SRS was 71 years (R: 25-98 years). The median time from first pain episode to SRS treatment was 48.7 months (R: 0-610
RI PT
months). SRS was the primary treatment for 637 patients (73%) while 233 patients (27%) had undergone ≥1 prior procedure.
SC
The 870 patients were divided into three dose groups for comparative analysis: 352 patients (40%) were treated to ≤ 82 Gy, 85 patients (10%) to 83-86 Gy, and 433 patients (50%) to ≥ 90
M AN U
Gy. The median dose prescribed was 86 Gy to the 100% isodose line and the median doses prescribed in the three dose cohorts were 82 Gy (R: 70-82 Gy), 86 Gy (R: 83-86 Gy), and 90 Gy (R: 90-95 Gy). The median follow-up was 36.5 months (R: 0-207.9 months). The median followup time for patients treated to 83-86 Gy was shorter (6.6 months) than the median follow-up for
TE D
patients treated to ≤ 82 Gy (22.4 months) or ≥90 Gy (48.7 months).
In total, 69 patients (8%) did not have a pain response recorded at follow-up and these patients
EP
were omitted from the analysis for pain response, though other variables and outcomes may have been collected. An improvement in pain after SRS was noted in 747 patients (86% of the total
AC C
group, 93% of patients with follow-up data regarding pain response). The 4-year rate of excellent to good pain relief across all patients was 86.7% (95% CI: 83.6-89.8%). There was a stepwise improvement in the 4-year rate of excellent or good pain response with increasing dose: 79.0% (95% CI: 73.1-85.0%), 81.6% (95% CI: 63.4-99.9%), and 92.0% (95% CI: 88.7%-95.4%) in patients treated to ≤82 Gy, 83-86 Gy, and ≥90 Gy, respectively (p=0.0019, Figure 1), and patients treated to ≥90 Gy had longer times to pain failure compared to patients treated to ≤82
7
ACCEPTED MANUSCRIPT
Gy (p=0.0019). For patients with MS or atypical trigeminal neuralgia, the 4-year rate of excellent or good pain response was 71.6% (95% CI: 56.4-86.8%) and 85.3% (95% CI: 72.0-98.7%) for those treated to ≤82 Gy and ≥90 Gy, respectively (p=0.21, Supplemental Figure 1). For patients
RI PT
who had previously undergone at least one procedure, the 4-year rate of excellent or good pain response was 75.6% (95% CI: 65.4-85.8%) and 82.5% (95% CI: 71.4-93.6%) for those treated to
SC
≤82 Gy and ≥90 Gy, respectively (p=0.24).
Univariate analysis revealed that dose, age at time of SRS, and a history of prior procedures were
M AN U
predictors of pain failure following SRS. Specifically, patients treated to lower doses (≤82 Gy) were at higher risk of a having a recurrence of their pain (HR: 2.0, 95% CI: 1.3-3.0) compared to patients treated to ≥90 Gy (p=0.0007). Although patients treated to doses from 83-86 Gy also appeared to experience a higher rate of pain failure (HR: 2.2, 95% CI: 0.8-6.1), this result was
TE D
not statistically significant (p=0.15). Additionally, patients who had a prior surgical procedure were at higher risk for pain failure after SRS (HR: 1.70, 95% CI: 1.12-2.52). When the number of prior procedures (R: 1-6) were analyzed as a continuous variable, they were also associated
EP
with an increased risk of pain failure (HR: 1.28, 95% CI: 1.1-1.5). Multivariate analysis confirmed that prescribed dose influenced pain control outcomes (p=0.008) and in a second
AC C
multivariable model examining the effect of the number of procedures prior to SRS, an increasing number of procedures was also significantly associated with an increased risk of pain failure following SRS (p=0.033) (Table 2). Time between the first onset of symptoms to SRS did not appear to influence pain control.
8
ACCEPTED MANUSCRIPT
Follow-up records for the assessment of facial numbness were available for 802 patients (92%). The worst facial numbness score was class I in 379 patients (43%), class II in 170 patients (20%), class III in 175 patients (20%), and class IV in 78 patients (9%). The 5-year rate of
RI PT
freedom from class III or class IV facial numbness was 58.4% (95% CI: 53.9-62.8%). The 4-year rates of freedom from class III/IV numbness were similar among patients treated to 83-86 Gy (50.7%, 95% CI: 20.7-80.8%) and ≥90 Gy (59.7%, 95% CI: 54.3-65.2%) and were significantly
SC
lower than the 4-year rate for patients treated to the ≤82 Gy (74.9%, 95% CI: 68.8-80.9%)
M AN U
(p=0.0056, Figure 2).
On univariate analysis, patients treated to doses ≥90 Gy were at higher risk of developing a class III/IV facial numbness (HR: 1.53, 95% CI: 1.16-2.04) compared to patients treated to the lowest doses (≤ 82 Gy, p=0.0028). Gender, age at time of GK, MS, or history of a previous surgical
TE D
procedure did not relate to risk of treatment-related facial numbness (see Table 3).
At last follow-up, 539 patients (62%) reported excellent and 162 patients (19%) reported good
EP
pain control. Similarly, at last follow-up, 511 patients (59%), 130 patients (15%), 112 patients (13%), and 45 patients (5%) experienced class I-IV facial numbness, respectively. Nine patients
AC C
(1%) were diagnosed with anesthesia dolorosa, all of whom were treated to doses ≤86 Gy.
Discussion:
In our analysis of over 800 patients with a median follow-up of approximately three years, we concluded that pain control outcomes are best when SRS is the first procedure in the management of TN and doses >82 Gy are prescribed to the DREZ. Moreover, patients can
9
ACCEPTED MANUSCRIPT
expect favorable outcomes with treatment regardless of the length of symptoms prior to radiosurgery. They must be cautioned, however, that efficacy is reduced when SRS follows other procedures. Patients should also be counseled about the risks of treatment-related facial
RI PT
numbness as a result of therapy.
The dose for SRS was first established by a multi-institutional study in which 50 patients were
SC
treated at five centers to doses ranging from 60-90 Gy prescribed to the 100% isodose line using a single 4-mm isocenter targeted at the nerve root entry zone [2]. A dose of ≥70 Gy was
M AN U
associated with a significantly greater chance of pain relief (72% vs. 9%, p=0.0003). Limitations of applicability of this data include lack of control for other factors influencing treatment outcome, such as age or a history of a prior procedure. Moreover, the prescribed doses have gradually increased over time, with some studies suggesting a benefit to doses as high as 90 Gy
TE D
(although caution should be used for doses exceeding 90 Gy) [9,10,11]. In general, there is a significant variation in prescription doses across institutions, as is best illustrated when reviewing the literature for large experiences with comparably long-term treatment outcomes
EP
(Table 4) [12-16]. In general, significant heterogeneity in patient selection, previous neurosurgical procedures, type of trigeminal neuralgia, prescription doses, and target placement
AC C
limit the extent of cross-comparisons. Besides these variations, the types of outcomes data reported, including the scales used, and the management style of post-treatment recurrences and complications are institutionally dependent and this adds considerable variation to the data pool. However, consideration should be given to a larger multi-institutional analysis to determine the effect of dose on treatment outcome using standardized and validated scales, ideally in a prospective manner.
10
ACCEPTED MANUSCRIPT
In addition to a higher chance of pain relief, there are other advantages to dose escalation. Kim et al. reported a more rapid response to treatment in 44 patients treated to a dose of 85 Gy when retrospectively compared to 66 patients treated to a dose of 80 Gy; however, they did not observe
RI PT
a significant difference in pain control rate between the two groups [17]. Moreover, MorbidniGaffney et al. observed a longer duration of pain response in patients treated to doses exceeding 85 Gy [18]. Our results suggest a significant long-term benefit to dose escalation, with step-wise
SC
improvements in rates of pain response at 4-years with increasing dose: from 79% for those patients treated to doses ≤82 Gy compared to 92% in those treated to doses ≥90 Gy. At eight
M AN U
years post SRS, with approximately 50 patients in both the lowest and highest dose cohorts, higher rates of pain relief were still noted in patients treated to ≥90 Gy (78%, 95% CI: 71.484.4%) vs. (67%, 95% CI: 59.0-75.6%) for patients treated to ≤82 Gy. Collectively, these retrospective reports and the results of our study suggest a benefit with regards to dose escalation
TE D
in terms of likelihood of pain response and duration of symptom relief.
There is an ultimate trade-off to dose escalation and multiple series have cautioned about
EP
increased risk for treatment-related morbidity. For example, Marshall and colleagues reported on a series of 448 patients treated to doses ranging from 80 to 97 Gy, where the mean maximum
AC C
prescribed dose was 88 Gy and the majority of patients received 90 Gy [19]. The mean dose to the dorsal root entry zone was 57.6 Gy in those who developed facial numbness compared to 47.3 Gy in those who did not develop numbness (p=0.02). In the present series, patients treated to doses >82 Gy were at increased risk for treatment-related numbness. Interestingly, this risk was not significantly higher in patients treated to doses ≥90 Gy vs. patients treated to 83-86 Gy. Since there was a benefit with dose escalation beyond 83 Gy but no significant increase in facial
11
ACCEPTED MANUSCRIPT
numbness, doses up to 90 Gy to the DREZ may be safe for clinical practice if mindful of the isocenter placement and brainstem tolerance and as long as patients are counseled about the risks of facial numbness. Longer follow-up in patients treated to doses ranging from 83 to 90 Gy may
RI PT
help determine the optimal dose which provides a favorable balance between pain control and treatment-related facial numbness.
SC
Dose escalation to 90 Gy to the DREZ may be considered when treating patients with apriori risk factors for pain failure following radiosurgery. In our series, the total number of prior procedures
M AN U
before SRS was significant on multivariable analysis as a predictor for early pain failure. When using a similar four-category pain response scale, Flickinger and colleagues also observed a similar relationship with rates of complete pain relief of 75%, 70%, and 63% at a median followup of 26 months in patients who had previously undergone 0, 1, and ≥2 prior neurosurgical
TE D
procedures, respectively [8]. On multivariable analysis, we did not observe a statistically significant improvement in freedom from pain failure with dose escalation in patients who had received ≥1 procedure; however, dose escalation may be considered in patients who have had
EP
multiple prior procedures and are at the highest risk for early pain failure. Determining this threshold will be key to improving outcomes for these patients and should be studied in a larger
AC C
multi-institutional analysis with adequate power.
We acknowledge several limitations to our study. Foremost, when performing multi-institutional studies, there always exists a risk for differences in data collection and outcomes assessment between institutions. We tried to limit this by using a four category classification system for pain response, albeit with a tradeoff of reduced sensitivity. The BNI scale is most commonly used in
12
ACCEPTED MANUSCRIPT
evaluation of trigeminal neuralgia patients, however, it has yet to be empirically validated. It contains as many as 12 permutations of outcomes (four pain categories and three medication-use categories) and was thought to be too complicated for a retrospective multi-institutional analysis.
RI PT
Therefore, we used a simplified classification system, in which pain response and medication use could be divided into four categories. Simplified scales such as this have been used in similar multi-institutional and large single institution studies, though we recognize that the best data
SC
would be collected prospectively using a sensitive patient-reported test obtained at baseline and at standard follow-up appointments [8]. Also, use of a categorical system in this series (similar to
M AN U
the BNI) as opposed to a quantifiable pain response scale prevents any quantitative estimations of benefit and limits analysis to dichotomous outcomes (pain relief vs. pain failure). To help improve data coding reliability, we limited coding to a few individuals. Given the lack of a true threshold for initiation of salvage procedures and the heterogeneity in re-treatment approaches,
TE D
we only used pain failure or facial numbness to evaluate SRS treatment efficacy and were not able to evaluate potential benefits of salvage therapies. Furthermore, as this was a retrospective study, a standardized follow-up schedule was not mandated. Considering the clinical complexity
EP
regarding patient symptoms from TN, the primary endpoints are often based on subjective outcomes which make data analysis difficult. Therefore, clinical questions regarding treatment
AC C
delivery techniques or comparisons of treatment parameters would be best addressed by prospectively collected data with quality-of-life endpoints, medication diaries, and patient satisfaction outcomes [20].
Conclusion:
13
ACCEPTED MANUSCRIPT
Our study demonstrates that SRS is an effective treatment option for patients with TN that provides durable pain relief. Patients should be counseled about reduced efficacy if they have undergone multiple prior procedures. In addition, pain control outcomes are best when patients
RI PT
are treated to >82 Gy to the DREZ, however, they must be counseled about the risks of
treatment-related facial numbness. We hope that future studies will analyze additional variables
AC C
EP
TE D
M AN U
SC
to help improve the outcomes in patients with this complex neurological disorder.
14
ACCEPTED MANUSCRIPT
Figure Legends:
dose group: ≤ 82 Gy, 83-86 Gy, and ≥ 90 Gy.
RI PT
Figure 1. Kaplan-Meier curve of freedom from pain failure (fair or poor pain control) based on
Figure 2. Kaplan-Meier curve of freedom from Barrow Neurologic Institute class III or IV
AC C
EP
TE D
M AN U
SC
treatment-related facial numbness based on dose group: ≤ 82 Gy, 83-86 Gy, and ≥ 90 Gy.
15
ACCEPTED MANUSCRIPT
References 1. Regis J, Metellus P, Hayashi M, Roussel P, Donnet A, Bille-Turc F. Prospective controlled trial of gamma knife surgery for essential trigeminal neuralgia. J Neurosurg
RI PT
2006;104:913-924.
2. Kondziolka D, Lunsford LD, Flickinger JC, Young RF, Vermeulen S, Duma CM, Jacques DB, Rand RW, Regis J, Peragut JC, Manera L, Epstein MH, Lindquist C. Stereotactic
SC
radiosurgery for trigeminal neuralgia: A multiinstitutional study using the gamma unit. J
3. XXXXX
M AN U
Neurosurg 1996;84:940-945.
4. Tsai JS, Rivard MJ, Engler MJ, Mignano JE, Wazer DE, Shucart WA. Determination of the 4 mm gamma knife helmet relative output factor using a variety of detectors. Medical physics 2003;30:986-992.
TE D
5. Burchiel KJ. A new classification for facial pain. Neurosurgery 2003;53:1164-1166; discussion 1166-1167.
6. Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han PP, Speiser BL. Gamma knife
EP
radiosurgery for trigeminal neuralgia: The initial experience of the barrow neurological institute. International journal of radiation oncology, biology, physics 2000;47:1013-
AC C
1019.
7. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta chirurgica Scandinavica 1971;137:311-314.
8. Flickinger JC, Pollock BE, Kondziolka D, Phuong LK, Foote RL, Stafford SL, Lunsford LD. Does increased nerve length within the treatment volume improve trigeminal neuralgia
16
ACCEPTED MANUSCRIPT
radiosurgery? A prospective double-blind, randomized study. International journal of radiation oncology, biology, physics 2001;51:449-454. 9. Pollock BE, Phuong LK, Gorman DA, Foote RL, Stafford SL. Stereotactic radiosurgery for
RI PT
idiopathic trigeminal neuralgia. J Neurosurg 2002;97:347-353.
10. Kondziolka D, Lacomis D, Niranjan A, Mori Y, Maesawa S, Fellows W, Lunsford LD. Histological effects of trigeminal nerve radiosurgery in a primate model: Implications for
SC
trigeminal neuralgia radiosurgery. Neurosurgery 2000;46:971-976; discussion 976-977. 11. Longhi M, Rizzo P, Nicolato A, Foroni R, Reggio M, Gerosa M. Gamma knife radiosurgery
M AN U
for trigeminal neuralgia: Results and potentially predictive parameters--part i: Idiopathic trigeminal neuralgia. Neurosurgery 2007;61:1254-1260; discussion 1260-1251. 12. Verheul JB, Hanssens PE, Lie ST, Leenstra S, Piersma H, Beute GN. Gamma knife surgery for trigeminal neuralgia: A review of 450 consecutive cases. J Neurosurg 2010;113
TE D
Suppl:160-167.
13. Kondziolka D, Zorro O, Lobato-Polo J, Kano H, Flannery TJ, Flickinger JC, Lunsford LD. Gamma knife stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg
EP
2010;112:758-765.
14. Young B, Shivazad A, Kryscio RJ, St Clair W, Bush HM. Long-term outcome of high-dose
AC C
gamma knife surgery in treatment of trigeminal neuralgia. J Neurosurg 2013;119:11661175.
15. Lucas JT, Jr., Nida AM, Isom S, Marshall K, Bourland JD, Laxton AW, Tatter SB, Chan MD. Predictive nomogram for the durability of pain relief from gamma knife radiation surgery in the treatment of trigeminal neuralgia. International journal of radiation oncology, biology, physics 2014;89:120-126.
17
ACCEPTED MANUSCRIPT
16. Regis J, Tuleasca C, Resseguier N, Carron R, Donnet A, Gaudart J, Levivier M. Long-term safety and efficacy of gamma knife surgery in classical trigeminal neuralgia: A 497patient historical cohort study. J Neurosurg 2015:1-9.
RI PT
17. Kim YH, Kim DG, Kim JW, Kim YH, Han JH, Chung HT, Paek SH. Is it effective to raise the irradiation dose from 80 to 85 gy in gamma knife radiosurgery for trigeminal neuralgia? Stereotactic and functional neurosurgery 2010;88:169-176.
SC
18. Morbidini-Gaffney S, Chung CT, Alpert TE, Newman N, Hahn SS, Shah H, Mitchell L, Bassano D, Darbar A, Bajwa SA, Hodge C. Doses greater than 85 gy and two isocenters
M AN U
in gamma knife surgery for trigeminal neuralgia: Updated results. J Neurosurg 2006;105 Suppl:107-111.
19. Marshall K, Chan MD, McCoy TP, Aubuchon AC, Bourland JD, McMullen KP, deGuzman AF, Munley MT, Shaw EG, Tatter SB, Ellis TL. Predictive variables for the successful
TE D
treatment of trigeminal neuralgia with gamma knife radiosurgery. Neurosurgery 2012;70:566-572; discussion 572-563.
20. Zakrzewska JM, Lopez BC. Quality of reporting in evaluations of surgical treatment of
EP
trigeminal neuralgia: Recommendations for future reports. Neurosurgery 2003;53:110-
AC C
120; discussion 120-112.
18
ACCEPTED MANUSCRIPT
≤ 82 Gy N (%) 352 (40%)
≥ 90 Gy N (%) 433 (50%)
p-value
60 30-90
RI PT
60 11-89
66 18-96
0.2427
69 30-96
71 25-98
SC
71 28-93
0.4596
83 (98%) 2 (2%)
409 (95%) 24 (6%)
M AN U
335 (95%) 17 (5%) 204 (58%) 148 (42%) 0 (0%) 314 (89%) 38 (11%)
48 (56%) 37 (43%) 0 (0%) 76 (89%) 9 (11%)
0.0650
281 (65%) 149 (34%) 3 (1%) 0.1782
402 (93%) 31 (7%)
211 (60%) 140 (40%) 1 (0.3%)
56 (66%) 29 (34%) 0 (0%)
257 (60%) 176 (41%) 0 (0%)
108 (31%) 255 (73%) 212 (60%)
25 (29%) 67 (79%) 47 (55%)
103 (24%) 311 (72%) 239 (55%)
EP
AC C
Abbreviations NA = not available yrs = years
83-86 Gy N (%) 85 (10%)
<0.0001
TE D
Table 1. Patient characteristics All Patients N (%) Number of Patients 870 Age at first pain onset Median (yrs) 63 Range (yrs) 11-96 Age at stereotactic radiosurgery Median (yrs) 71 Range (yrs) 25-98 Typical vs. Atypical Symptoms Typical 827 (95%) Atypical 43 (5%) Gender Female 533 (61%) Male 334 (38%) NA 3 (0.3%) Multiple Sclerosis No 792 (91%) Yes 78 (9%) Right vs. Left Right 524 (60%) Left 345 (40%) NA 1 (0.1%) Trigeminal nerve distribution I 236 (27%) II 633 (73%) III 498 (57%)
0.6008
0.0855 0.4095 0.0340
ACCEPTED MANUSCRIPT
Table 2. Cox proportional hazards regression analysis of predictors of pain failure following stereotactic radiosurgery
1.755 0.824 2.131 0.988 1.478 -
RI PT
0.001 0.888 0.151 0.399 0.068 0.639 0.126 0.043 0.560 0.011 0.005
95% CI Lower Upper P value Multivariable analysis 1.157 0.290 0.749 0.972 0.976 -
SC
2.986 2.620 6.123 1.791 3.752 2.179 1.003 1.000 1.002 2.529 1.523
Hazard ratio
M AN U
1.336 0.329 0.757 0.793 0.953 0.620 0.975 0.969 0.997 1.127 1.078
EP
Abbreviations: CI = confidence interval TN = trigeminal neuralgia SRS = stereotactic radiosurgery
1.997 0.928 2.152 1.192 1.891 1.163 0.989 0.984 0.999 1.689 1.281
AC C
Dose Group ≤ 82 Gy vs. ≥ 90 Gy ≤ 82 Gy vs. 83-86 Gy 83-86 Gy vs. ≥ 90 Gy Gender (F vs. M) Atypical vs. typical TN Multiple sclerosis Age at first pain onset Age at time of SRS Time from first pain onset to SRS SRS after previous procedure No. of previous procedures prior to SRS
95% CI Lower Upper P value Univariate analysis
TE D
Hazard ratio
2.663 2.339 6.067 1.003 2.239 -
0.008 0.716 0.156 0.118 0.0652 -
Hazard ratio
95% CI Lower Upper P value Multivariable analysis
1.777 0.872 2.039 0.988
1.175 0.308 0.715 0.972
2.687 2.464 5.815 1.003
0.006 0.795 0.183 0.120
1.215
1.016
1.452
0.033
ACCEPTED MANUSCRIPT
95% CI
SC
1.157 0.929 0.442 0.754 0.583 0.778 0.999 0.999 0.999 0.607 0.775
Upper
M AN U
1.534 1.821 0.842 0.972 0.982 1.204 1.008 1.009 1.000 0.813 0.910
TE D
AC C
Abbreviations: CI = confidence interval TN = trigeminal neuralgia SRS = stereotactic radiosurgery
Lower
EP
Univariate analysis covariates Dose Group ≥ 90 Gy vs. ≤ 82 Gy 83-86 Gy vs. ≤ 82 Gy ≥ 90 Gy vs. 83-86 Gy Gender (F vs. M) Atypical vs. typical TN Multiple sclerosis Age at first pain onset Age at time of SRS Time from first pain onset to SRS SRS after previous procedure No. of previous procedures prior to SRS
Hazard ratio
RI PT
Table 3. Cox proportional hazards regression analysis of predictors of facial numbness following stereotactic radiosurgery
2.033 3.571 1.603 1.254 1.655 1.866 1.017 1.020 1.002 1.089 1.067
P value
0.003 0.081 0.601 0.828 0.945 0.405 0.101 0.074 0.831 0.165 0.245
ACCEPTED MANUSCRIPT
Table 4. Selected large institutional experiences of stereotactic radiosurgery for trigeminal neuralgia
Young et al. Lucas et al. Régis et al. Present study
Median FUP (months)
Dose Range
Pain Control
Pain Control Scale
Freedom from facial numbness
Facial numbness scale
2010
365
28.0
80 Gy
58.0%1
BNI
93.5%2
BNI
BNI
89.5%
2
BNI
67.1%
2
WBFS
2010 2013 2014 2015 2015
503 315 446 497 870
24.0
60 - 90 Gy
58.8
90 Gy
21.2
80 - 97 Gy
43.8
70 - 90 Gy
36.5
70 - 95 Gy
46.0%
1
71.4%
2
46.9%
1
64.9%
1
84.0%
1
RI PT
Kondziolka et al.
No. of patients
MMS
SC
Vergeul et al.
Year
M AN U
Series Author
BNI BNI
EGFP
-
-
79.6%
1
BNI
58.4%
1
BNI
AC C
EP
TE D
EGFP = Excellent, Good, Fair, Poor categorical scoring system, BNI = Barrow Neurologic Institute, MMS = Modified Marseille Scale, WBFS = Wong-Baker FACES scale 1 = 5-year rates, 2 = crude-rate, 3 = 4-year rates
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT