Abstract

Background: Linear accelerator (LINAC)-based stereotactic radiosurgery (SRS) treatment of trigeminal neuralgia (TN) may have similar efficacy to Gamma Knife SRS (GK-SRS), but the preponderance of data comes from patients treated with GK-SRS. Our objective was to analyze the outcomes for LINAC-based treatment of TN in patients at our institution.

Methods: We retrospectively analyzed data for patients who underwent LINAC-based SRS for TN from 2006 to 2018. Data were collected from the patients’ medical records. Nonparametric statistics were used for the analysis.

Results: Of the 41 patients treated with LINAC-based SRS (typically 90 Gy dosed using a 4 mm collimator for one fraction) during that time, follow-up data of >3 weeks post-SRS were available for 32 patients. The median pretreatment Barrow Neurological Institute (BNI) pain score was 5 (range 4–5). The follow-up period ranged from 0.9 to 113.2 months (median 5 months). There was significant improvement in postradiation BNI pain score (P 3). Survival analysis showed a median time to pain recurrence of 30 months. There was no relationship between prior microvascular decompression (MVD) surgery and change in BNI pain score pre- to posttreatment.

Conclusion: The results demonstrate that LINAC-based SRS is an effective means to treat TN. Prior MVD surgery did not affect efficacy of SRS in lowering the BNI score from pre- to posttreatment in this patient cohort.

Keywords: BNI pain score, Linear accelerator, Outcomes, Pain, Stereotactic radiosurgery, Trigeminal neuralgia

INTRODUCTION

Trigeminal neuralgia (TN) is classically characterized by episodic facial pain (Type 1). Other subtypes have also been recognized that involve continuous facial pain (Type 2) or pain related to demyelinating disorders.[ 2 , 8 ] Although treatment with microvascular decompression (MVD) has become more common over the past decade,[ 15 , 25 ] noninvasive modalities of treatment such as stereotactic radiosurgery (SRS) may be a favorable option for those patients who refuse invasive surgery or are not able to tolerate open surgical treatment.

The two most common methods of SRS are Gamma Knife (GK) and linear accelerator (LINAC). A retrospective systematic review of the techniques suggested comparable efficacies for the treatment of TN,[ 24 ] but LINAC-based treatment may be less costly for hospitals with small patient pools and does not require maintaining cobalt sources as is the case in GKS.[ 9 , 14 ] LINAC machines can also be used for extracranial radiation and non-SRS cranial treatments and are arguably more versatile for small community hospitals to purchase and maintain.

In this study, we retrospectively analyzed outcomes for LINAC-based treatment of TN in a group of patients at our institution. Importantly, we consistently used a radiation dose of 80–90 Gy delivered to the root entry zone (REZ) to treat our patients, which has been reported as an efficacious dose and target for the treatment of TN.[ 20 , 23 , 24 ]

MATERIALS AND METHODS

Patients

We retrospectively analyzed data for consecutive patients treated at our institution who underwent LINAC-based radiation treatment for TN from 2006 to 2018. For this Institutional Review Board (IRB)-approved case series, data were collected from the electronic medical record. Explicit patient consent was waived for this retrospective data review by the IRB. All data were deidentified for analysis. Data regarding age, sex, time of pain onset, and time of SRS procedure relative to other interventions (including MVD, percutaneous balloon rhizotomy, radiofrequency ablation, and glycerol rhizotomy) were recorded. A medical history of multiple sclerosis was also noted. Time to pain relief was determined based on clinical follow-up visit notes. Brainstem dosimetry data were also collected when available for patients.

TN type was delineated using the Burchiel classification system.[ 2 ] A Burchiel Type 1 classification refers to the classic description of TN facial pain that is episodic and unilateral for >50% of the time, separated by pain-free intervals. A Burchiel Type 2 classification refers to having continuous facial pain for >50% of the time with possible additional episodic pain. Other categories include “symptomatic trigeminal neuralgia” in patients with multiple sclerosis, atypical facial pain associated with a somatoform disorder, and trigeminal neuropathic pain associated with unintentional traumatic injury.[ 2 ] Patients with follow-up times <3 weeks were excluded from the analysis because we deemed that to be too short to reliably detect symptom changes related to radiation. Patients with SRS targets other than the REZ were also excluded from the study.

Treatment protocol

All patients underwent frame-based stereotaxis using a stereotactic Leksell head frame (Elekta Instruments AB Stockholm, Sweden). Patients then underwent a CT scan acquired with a 1–1.25 mm slice thickness for dose calculation and treatment planning purposes. Constructive interference in steady-state or 3D-fast imaging employing steady-state acquisition magnetic resonance imaging had been performed previously, and these scans were imported into the iPlan treatment planning system (Brainlab AG, Munich, Germany). These were used in combination with stereotactic T1-weighted images to identify the trigeminal nerve target and place the isocenter. After fusion with the CT scan, a treatment plan with seven arcs using a 4 mm collimator was devised. The target for all patients was the REZ of the TN in the prepontine cistern such that the brainstem contour was at the ~50% isodose line (IDL).[ 5 , 10 ] For most patients, the prescribed dose was 90 Gy to isocenter (100% IDL) in a single fraction. We ensured that the 20%, 30%, and 50% IDL had a mean brainstem volume of 0.14, 0.047, and 0.008 ml, respectively, per Goss et al.[ 10 ] Treatment was delivered on the BrainLAB Novalis LINAC device or, later, on a Varian TrueBeam (Varian, Palo Alto, CA) device. After treatment, the stereotactic frame was removed, and patients were discharged home. All patients were treated by the same team of providers.

Pain scale

The Barrow Neurological Institute (BNI) scoring system was used to quantify pain before and after the procedure.[ 21 ] A score of 1 constitutes complete pain relief without medication, a score of 2 constitutes occasional pain not requiring medication, a score of 3 constitutes some pain that is controlled on medication (subscore 3a constitutes no pain on pain medications and subscore 3b constitutes some pain that is controlled with medications), a score of 4 constitutes some pain that is inadequately controlled on medication, and a score of 5 is pain that is severe without any relief from medications. This score was determined based on patient progress notes from clinic visits.

Statistical analysis

Data analysis was done using Matlab (version 2019b, Natick, MA). The nonparametric Wilcoxon rank-sum and signed-rank tests were used for the analysis. The Kaplan–Meier estimator was used to estimate the probability of pain control (defined as a BNI <4) from time of SRS treatment in a cohort of patients who had initial pain control.[ 3 ] Censored data represented those patients who did not reach the endpoint of pain recurrence (BNI 4–5) at the time of last follow-up. Patients who were lost to follow-up were not included in the survival probability estimate.

This case series has been reported in line with the PROCESS guidelines.[ 1 ]

RESULTS

Demographics

Forty-one patients with a mean age of 67.4 (range 30–89) years were identified for inclusion. Of these, 25 (61%) had Type 1 TN, 3 (7%) had Burchiel Type 2 TN, 1 (2%) had TN related to trauma (neuropathic TN), 8 (20%) had multiple sclerosis-related TN (symptomatic TN), and 4 (10%) had atypical TN [ Table 1 ]. Thirteen (32%) had received MVD previously and 32 (78%) had another surgical procedure such as glycerol rhizotomy before SRS. Most patients received 90 Gy of radiation prescribed to the isocenter using a 4 mm collimator for one fraction; two patients received a dose of 80 Gy. The median BNI score was 5 (range 4–5) before SRS treatment.

Table 1:

Demographics of 41 patients treated with LINAC for TN.

Outcomes after SRS

Of the 41 patients, 32 patients had sufficient follow-up data. The follow-up period varied from 0.9 to 113.2 months (median 5 months). Of these 32 patients, 23 (72%) reached a BNI pain score of 1–3. There was significant improvement in postradiation BNI score (z = 4.38, P < 0.001, Wilcoxon signed-rank test, [ Figure 1 ]). Across improved patients, 13 patients had a BNI score of 3b and five patients had a BNI score of 3a after treatment. Of the 21 patients with Type 1 TN, 15 (71%) improved to a BNI score <4. Similarly, 6/7 (86%) patients with multiple sclerosis and 2/3 (67%) patients with atypical TN improved to a BNI score <4. One patient had hypoesthesia after radiation. Interestingly, there was no relationship between prior MVD surgery and change in BNI score pre- to posttreatment (z = 1.58, P = 0.11, Wilcoxon rank-sum test). Furthermore, prior history of any type of treatment did not relate to change in BNI score pre- to post-LINAC-SRS treatment (z = −1.8, P = 0.07, Wilcoxon rank-sum test).

Figure 1:

Histograms showing number of patients with a particular (a) pre-LINAC treatment BNI score and (b) post-LINAC treatment score. There was significant improvement in postradiation BNI score across patients (z = 4.38, P < 0.001, Wilcoxon signed-rank test).

Brainstem dosimetry was retrieved for 33 patients. The median maximal dose delivered was 4500 cGy (range 1530–7020 cGy). Volumetric analysis showed that median volume of tissue that received 4500 cGy was 0.001 ml, the median volume that received 2700 cGy was 0.009 ml, and the median volume that received 1800 cGy radiation was 0.029 ml. The maximum dose to the brainstem did not relate to pain persistence (i.e., BNI score >3; Kruskal–Wallis test, Chi-squared (1) = 1.17, P = 0.28). Because only one patient had a documented adverse effect, dosimetric analysis on toxicity could not be performed; however, the maximum dose delivered to that patient was 4050 cGy.

A total of 12 of 32 (38%) patients had recurrence of symptoms after SRS, and 11 underwent an additional intervention such as additional SRS, rhizotomy, or peripheral electrical stimulation. Kaplan–Meier analysis of the probability of pain control from time of SRS treatment was done using a subset of patients who had initial benefit from SRS (BNI score <4). The median time to recurrence was 30.2 months (range 18.3– 94 months; [ Figure 2 ]). The “hazard rate” of pain recurrence was 8%/person-month.

Figure 2:

Kaplan–Meier survival curve showing probability of pain control (BNI < 4) decreases with time. This analysis includes only patients who had initial pain control with LINAC-based SRS treatment. Censored data are those patients who did not reach the endpoint of interest, which was documented pain recurrence at last follow-up (denoted with a “|”). The curve shows decreasing probability of continued pain control from time of intervention. A 95% confidence interval shown in dashed line.

DISCUSSION

TN affects 4.3 in 100,000 people/year,[ 13 ] and the treatment options vary based on the medical history and prior treatments. Radiosurgery offers a noninvasive means of treating TN, with the tradeoff that the treatment effects may take months to happen, although some effects are reported to occur as early as a few weeks,[ 7 ] and that there may be delayed complications such as hypoesthesia that can be bothersome. Several systems exist for delivering radiation to the REZ (or the Gasserian ganglion itself, although this is a less common target), namely, GKS, CyberKnife radiosurgery (CKR), and LINAC-based systems. In a recent systematic analysis, the mean rate of pain freedom in Type 1 patients with or without medication was comparable among the modalities (85% for LINAC, 87% for GKS, and 79% for CKR, meta-analysis covering studies from 1951 to 2015, [ Table 2 ]).[ 4 , 6 , 11 , 12 , 17 , 18 , 22 , 24 , 26 , 27 ] These findings are comparable with our results of 72% of patients having pain control with or without medications (a BNI score of 1–3) and 71% of Burchiel Type 1 patients improving. We reviewed studies published from 2016 to 2022 that reported the rate of pain treatment in patients with Type 1 TN treated with either GK or LINAC [ Table 2 ] and found roughly 70–80% pain treatment effect (BNI 1–3) for GK and an 80–90% rate of pain control with LINAC (which is slightly better than our report). Formal meta-analysis methods are beyond the scope of this paper but suggest comparable efficacy between the modalities as shown in Tuleasca et al.[ 24 ]

Table 2:

Comparison of studies published in 2016–2022 reporting Gamma Knife or linear accelerator-based SRS treatment for Type 1 TN pain, reporting pain relief rate, recurrence rate, and time to recurrence.

As for other types of TN, in Pokhrel et al.,[ 19 ] no patients with atypical TN had improvement whereas we found 2/3 patients with atypical TN improved with SRS. We also found that 86% of patients with sensory TN (related to multiple sclerosis for our patients) achieved pain control with treatment, which is comparable with the previous findings in the literature.[ 16 , 24 ]

Tuleasca et al.[ 24 ] found the range of time to pain relief to be from 8.5 to 60 days, although estimates were biased by patient recall and were the product of follow-up appointment times not being standardized across studies. They reported that the median rate of recurrence was 29% and that time to pain recurrence was 7.5–20.4 months (meta-analysis covering studies from 1951 to 2015). Within that meta-analysis, Smith et al.[ 23 ] found 23% recurrence at 12.6 months. Selected LINAC studies from 2016 to 2022 [ Table 2 ] also indicate a roughly 25–35% pain recurrence rate for Type 1 TN and perhaps 57% for all TN types.[ 19 ] We observed a probability of recurrent pain of 58% (25–84% at the 95% CI) at 30 months across TN types in our patients, and an overall recurrence rate was 38%. Nine patients were lost to follow-up after SRS treatment, which may or may not indicate that the patient had pain relief and limits the interpretation of our results. The recurrence rate was statistically lower for GK compared with LINAC treatment in Tuleasca et al.[ 24 ] The rate of delayed bothersome hypoesthesia in our cohort was 1/32 (3%), which is lower than one other reported study result of 13% in a larger cohort of patients, though that was not explicitly labeled as “bothersome.”[ 23 ]

Interestingly, prior MVD surgery and prior surgical treatment history did not affect efficacy of SRS in lowering the BNI score from pre- to posttreatment in our cohort of patients. Tuleasca et al.[ 24 ] performed a pooled analysis across SRS modalities and showed that prior MVD did not affect efficacy of SRS treatment. Similarly, multivariate analysis has not shown prior surgery to relate to outcome with LINAC treatment.[ 20 ] In contrast, other studies have shown that prior treatments overall decrease the efficacy of SRS.[ 19 , 23 , 24 ] Thus, the results in this realm are mixed.

Limitations

This study has a range of limitations including being retrospective and having a relatively small sample size, although only several other studies have larger cohorts of patients treated with LINAC-based SRS for TN.[ 7 , 19 , 20 , 24 ] In addition, we do not have follow-up on nine patients, which leaves their treatment effect unknown. Studying a larger cohort and collecting longer follow-up data with planned follow-up clinic visits would help support our results. However, we believe that our institutional experience adds to the growing body of evidence showing the efficacy and tolerability of LINAC-SRS across a variety of institutions.

CONCLUSION

LINAC-based SRS is an effective means to treat TN for those with Type 1 as well as atypical TN and TN in the setting of multiple sclerosis. Prior MVD surgery did not affect efficacy of SRS in lowering the BNI score from pre- to posttreatment in this cohort of patients. A larger cohort of patients with prior MVD treatment followed by SRS treatment would help explore this result in future studies.

Declaration of patient consent

Institutional Review Board (IRB) permission obtained for the study.

Financial support and sponsorship

Neurosurgery Research and Education Fund (NREF).

Conflicts of interest

JDR has received consulting fees from Medtronic and Corlieve Therapeutics. RLJ has received consulting fees from Medtronic.

References

1. Agha RA, Sohrabi C, Mathew G, Franchi T, Kerwan A, O’Neill N. The PROCESS 2020 guideline: updating consensus preferred reporting of CasE series in surgery (PROCESS) guidelines. Int J Surg. 2020. 84: 231-5

2. Burchiel KJ. A new classification for facial pain. Neurosurgery. 2003. 53: 1164-6

3. Cardillo G.editors. KMPLOT: Kaplan-Meier Estimation of the Survival Function. San Francisco, California: GitHub; 2018. p.

4. Chang CS, Huang CW, Chou HH, Lin LY, Huang CF. Outcome of Gamma Knife radiosurgery for trigeminal neuralgia associated with neurovascular compression. J Clin Neurosci. 2018. 47: 174-7

5. Chen JC, Girvigian M, Greathouse H, Miller M, Rahimian J. Treatment of trigeminal neuralgia with linear accelerator radiosurgery: Initial results. J Neurosurg. 2004. 101: 346-50

6. Dai ZF, Huang QL, Liu HP, Zhang W. Efficacy of stereotactic gamma knife surgery and microvascular decompression in the treatment of primary trigeminal neuralgia: A retrospective study of 220 cases from a single center. J Pain Res. 2016. 9: 535-42

7. Debono B, Lotterie JA, Sol JC, Bousquet P, Duthil P, Monfraix S. Dedicated linear accelerator radiosurgery for classic trigeminal neuralgia: A single-center experience with long-term follow-up. World Neurosurg. 2019. 121: e775-85

8. Eller J, Raslan A, Burchiel K. Trigeminal neuralgia: Definition and classification. Neurosurg Focus. 2005. 18: E3

9. Friedman W. Linear accelerator radiosurgery. Clin Neurosurg. 1992. 38: 445-71

10. Goss BW, Frighetto L, DeSalles AA, Smith Z, Solberg T, Selch M. Linear accelerator radiosurgery using 90 gray for essential trigeminal neuralgia: Results and dose volume histogram analysis. Neurosurgery. 2003. 53: 823-30

11. Inoue T, Hirai H, Shima A, Suzuki F, Yamaji M, Fukushima T. Long-term outcomes of microvascular decompression and Gamma Knife surgery for trigeminal neuralgia: A retrospective comparison study. Acta Neurochir (Wien). 2017. 159: 2127-35

12. Jung HH, Park CK, Jung NY, Kim M, Chang WS, Chang JW. Gamma knife radiosurgery for idiopathic trigeminal neuralgia: Does the status of offending vessels influence pain control or side effects?. World Neurosurg. 2017. 104: 687-93

13. Katusic S, Beard CM, Bergstralh E, Kurland LT. Incidence and clinical features of trigeminal neuralgia, Rochester, Minnesota, 1945-1984. Ann Neurol. 1990. 27: 89-95

14. Königsmaier H, De Pauli-Ferch B, Hackl A, Pendl G. The costs of radiosurgical treatment: Comparison between gamma knife and linear accelerator. Acta Neurochir (Wien). 1998. 140: 1101-11

15. Kundu B, Rolston JD. Nationwide shift from percutaneous rhizotomy to microvascular decompression for treatment of trigeminal and other cranial nerve neuralgias. Headache. 2018. 58: 1675-9

16. Lee AT, Raygor KP, Elefant F, Ward MM, Wang DD, Barbaro NM. Comparison of stereotactic radiosurgery and radiofrequency ablation for trigeminal neuralgia in multiple sclerosis patients. Stereotact Funct Neurosurg. 2020. 98: 378-85

17. Moreno NE, Gutiérrez-Sárraga J, Rey-Portolés G, JiménezHuete A, Álvarez RM. Long-term outcomes in the treatment of classical trigeminal neuralgia by gamma knife radiosurgery: A retrospective study in patients with minimum 2-year follow-up. Neurosurgery. 2016. 79: 879-87

18. Mureb M, Golub D, Benjamin C, Gurewitz J, Strickland BA, Zada G. Earlier radiosurgery leads to better pain relief and less medication usage for trigeminal neuralgia patients: An international multicenter study. J Neurosurg. 2021. 135: 237-44

19. Pokhrel D, Sood S, Mcclinton C, Saleh H, Badkul R, Jiang H. Linac-based stereotactic radiosurgery (SRS) in the treatment of refractory trigeminal neuralgia: Detailed description of SRS procedure and reported clinical outcomes. J Appl Clin Med Phys. 2017. 18: 136-43

20. Rashid A, Pintea B, Kinfe TM, Surber G, Hamm K, Boström JP. LINAC stereotactic radiosurgery for trigeminal neuralgia-retrospective two-institutional examination of treatment outcomes. Radiat Oncol. 2018. 13: 1-11

21. Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han P, Speiser B. Gamma knife radiosurgery for trigeminal neuralgia: Experience at the barrow neurological institute. Int J Radiat Oncol Biol Phys. 2000. 47: 1013-9

22. Shrivastava A, Mohammed N, Hung YC, Xu Z, Schlesinger D, Heinrichs T. Impact of integral dose on the maintenance of pain relief in patients with idiopathic trigeminal neuralgia treated with upfront gamma knife radiosurgery. World Neurosurg. 2019. 129: e375-80

23. Smith ZA, Gorgulho AA, Bezrukiy N, McArthur D, Agazaryan N, Selch MT. Dedicated linear accelerator radiosurgery for trigeminal neuralgia: A single-center experience in 179 patients with varied dose prescriptions and treatment plans. Int J Radiat Oncol Biol Phys. 2011. 81: 225-31

24. Tuleasca C, Régis J, Sahgal A, De Salles A, Hayashi M, Ma L. Stereotactic radiosurgery for trigeminal neuralgia: A systematic review. J Neurosurg. 2018. 130: 733-57

25. Wang DD, Ouyang D, Englot DJ, Rolston JD, Molinaro AM, Ward M. Trends in surgical treatment for trigeminal neuralgia in the United States of America from 1988 to 2008. J Clin Neurosci. 2013. 20: 1538-45

26. Wang DD, Raygor KP, Cage TA, Ward MM, Westcott S, Barbaro NM. Prospective comparison of long-term pain relief rates after first-time microvascular decompression and stereotactic radiosurgery for trigeminal neuralgia. J Neurosurg. 2018. 128: 68-77

27. Zeng YJ, Zhang H, Yu S, Zhang W, Sun XC. Efficacy and safety of microvascular decompression and gamma knife surgery treatments for patients with primary trigeminal neuralgia: A prospective study. World Neurosurg. 2018. 116: e113-7