Stereotactic Radiosurgery versus Neuroablative Techniques for Medically Refractory Trigeminal Neuralgia: A Systematic Review and Meta-Analysis of Outcomes

Trigeminal neuralgia (TN) is defined by sudden, intense, sharp, electric shock-like shooting, or burning recurrent paroxysms of unilateral nerve pain in the distribution of the trigeminal nerve [1]. When the pain is severe enough to induce contraction of the facial muscles, this condition is called tic douloureux. TN is categorized into three types (classical, secondary, idiopathic), based on the aetiology.

Classical TN (type 1) stems from neurovascular compression (NVC) to the trigeminal nerve, requiring demonstration of NVC intraoperatively or within imaging. Patients with type 2 classical neuralgia also have persistent background facial pain in addition to the sudden pain characteristic of the disease. Secondary TN follows from an underlying disease, usually multiple sclerosis, tumours, or space-occupying lesions. Idiopathic TN type 1 includes patients with no abnormalities on electrophysiological tests or imaging, while patients with type 2 idiopathic neuralgia also have continuous background pain. Globally, trigeminal affects women more than men, and the incidence increases with age [1, 2].

A variety of studies have highlighted the significant detriment TN can have on health-related quality of life metrics such as the SF-36 and the Hospital Anxiety and Depression Scale [3, 4]. In the same vein, patients describe significant quality of life benefit from successful treatment, both medically and surgically [5, 6]. However, there are barriers to receiving adequate treatment. Misdiagnosis is common, and a large proportion of cases are refractory to first-line medications such as carbamazepine or oxcarbazepine [7]. If symptoms persist, clinicians increase the dose of these sodium-channel blocking agents but also risk increasing side effects such as dizziness, central nervous system depression, and hyponatremia. Alternatively, adjunct medications such as gabapentin, pregabalin, lamotrigine, or baclofen can be prescribed [7]. If medications still fail to control symptoms or medication side effects are intolerable, surgical approaches offered include microvascular decompression (MVD), neuroablative techniques such as percutaneous balloon compression (PBC), glycerol rhizolysis (GR), and radiofrequency thermocoagulation (RFT), or stereotactic radiosurgery (SRS) [7, 8]. Surgery offers the only potentially curative therapy for TN, but due to risk of complications, is usually only offered to patients with medically refractory TN [9‒11], generally defined as patients who fail to respond to at least 2 different medical treatments for TN [10, 11].

While MVD, which addresses trigeminal nerve irritation due to neurovascular compression, is considered the favoured surgical treatment [7], NVC is not observed in all patients. Studies have cited an overall prevalence of NVC to be 10% for idiopathic patients and have shown that factors including nerve atrophy or displacement also play significant roles in the symptomatic presentation of patients with classic TN [12]. While the current landscape is still developing, varying studies have cited the prevalence of idiopathic TN to range between 20 and 60% across populations with TN [13, 14]. As such, there remains a significant portion of patients with TN for whom MVD may not be indicated. MVD is also much more invasive, consequently having a higher rate of mortality (0.1%) and major complications including hearing loss, cranial nerve palsies, CSF leaks, and haemorrhage/stroke (5%) than other approaches [7], which is particularly pertinent for older patients or those who have significant comorbidities.

When the potential risks of MVD outweigh the benefits, the remaining surgical options are neuroablative percutaneous techniques or SRS. There are no studies that provide a comprehensive comparison of the outcomes of these alternatives to MVD. This study aims to perform a systematic review and meta-analysis of previous studies evaluating outcomes of neuroablative techniques and SRS in treating medically refractory TN.

A structured search was conducted in accordance with the systematic review guidelines outlined by the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) checklist [15]. Articles were searched using PubMed, MEDLINE, and Embase from inception up to June 2024. No restrictions or filters were used for study design or date. Only studies in English were included. The keywords “Stereotactic Radiosurgery OR Stereotactic OR Gamma Knife OR LINAC OR Cyberknife OR Radiofrequency Thermocoagulation OR Balloon Compression OR Glycerol OR Rhizolysis OR Ablative OR Neuroablative AND Trigeminal Neuralgia OR Tic Douloureux” were used in MeSH format to identify all articles referencing SRS or neuroablactive techniques (and their analogous descriptions) in the context of TN and the BNI scale. A further search using the terms “Stereotacti* OR OR Gamma Knife OR LINAC* OR Cyberknife* OR Radiofrequen* OR Balloon Com* OR Glycer* OR Rhizo* OR Ablat* OR Neuroablat* OR Lesion* AND Trigeminal Neur* OR Tic Douloureux*” was done in text-word format to account for any descriptions of the techniques that have not been formally indexed in the MeSH format.

Inclusion was limited to experimental and observational studies in English, consisting of patients with primary TN (classical and idiopathic) of any nerve root (V1, V2, V3), either unilaterally or bilaterally. Studies consisting of patients with secondary TN were excluded. Studies were required to have all participants of ≥16 years of age, and to have a sample size of ≥20 participants. All participants were required to be first-time candidates for the intervention being studied. If the study population consisted of participants who had previously undergone other surgical treatments, demographic details of this population were considered separately. Only studies covering the treatment modalities of SRS, RFT, GR, and PBC were included, including their subtypes. Studies reporting on MVD were excluded but may have been included if they consisted of a treatment arm with one of the four interventions. All articles were required to report pain outcomes using the Barrow Neurological Institute (BNI) pain intensity scale for TN [16]. This scale ranges from one to five, with one representing complete pain relief following treatment without the need for medication, and five representing little to no pain relief at all following treatment (Fig. 1). The BNI scale is one of the most commonly used scales to assess the efficacy of treatment of TN and provides a more detailed breakdown of the severity of disease, based on factors such as pain intensity and response to treatment. Moreover, without controlling for a common metric across studies, we risk assuming equivalence in the comparability of studies and levels of treatment efficacy. As such, we believe using the BNI scale allows the inclusion of a large sample of studies assessing outcomes in a standardized and disease-specific manner, providing key insights into the frequency and extent of pain relief patients can expect across treatment types [16]. All studies were also required to report overall rates of recurrence and complications at final follow-up. Follow-up duration was restricted to a minimum of 12 months following intervention. In cases where a study contained a partial population that fit the inclusion criteria with sufficient details regarding their demographic details and follow-up outcomes, only subjects/treatments that were relevant to our inclusion criteria were included.

Studies from the database search were input into the systematic review software Covidence. Two reviewers (Y.A., J.S.) independently completed initial abstract and title screening, followed by full-text screening in accordance with the inclusion and exclusion criteria. Any conflicts in decisions were discussed until a consensus was reached with the assistance of senior authors (C.H., L.T.). Data extraction was then completed based on the following set of outcomes: author, study name, study design, country, sample size, male:female ratio, population age (mean, median, range), aetiology of TN, intervention (SRS vs. neuroablative procedures, including any subtypes), pain relief (in BNI), recurrence and complications, follow-up time, dropout rate, reason for dropout.

Quality assessment was conducted independently by both reviewers, with any conflicts discussed with a senior author until consensus was reached. The Newcastle-Ottawa Scale (NOS) was used for observational studies. Cochrane Risk of Bias 2.0 tool was used for interventional studies.

Following pooling of all participant demographic and treatment-data, the Shapiro-Wilk test, two-tailed T test, and Spearman’s-R were used to test for the degree of significance between participant demographic factors and BNI outcomes, with studies weighted according to a fixed-effects model. The Shapiro-Wilk test, Mann-Whitney U test, and ANCOVA were used to test for statistical significance of the differing treatment modalities and patient outcomes. Pain outcomes were stratified by time, and included all studies reporting outcomes at the specified time-interval, regardless of whether further follow-up data were present. Data were compiled and analysed using GraphPad Prism Version 10.1.1 for Mac (GraphPad, San Diego, CA, USA). Parametric variables are presented as mean (±standard deviation). Nonparametric data sets are presented as median (interquartile range). p < 0.05 was considered statistically significant.

This review has been registered with the international prospective register of systematic reviews (PROSPERO). Registration ID: CRD42024557229.

The structured search resulted in 6,620 studies with 3,640 duplicates, leaving 2,980 studies. Initial title and abstract screening were performed, leaving 314 studies eligible for full-text screening. Following screening, 277 studies were excluded, leaving 37 studies to be included within this meta-analysis (Fig. 2). Of the 37 studies, 26 studies covered participants who underwent SRS, followed by 5 studies covering GR, 5 studies covering PBC, and 3 studies covering RFT. Two studies of the 37 covered 2 treatment modalities, with participants within each treatment group considered separately. Study and treatment characteristics are described in the online supplementary Table (for all online suppl. material, see https://doi.org/10.1159/000543859).

Of the 26 SRS studies, all studies included participants with primary idiopathic TN. One study included a group of patients with bilateral idiopathic TN, 3 studies included patients with atypical idiopathic TN, and 2 studies included patients with classical TN. Within the 13 neuroablative studies, 11 covered patients with primary idiopathic TN, 1 study covered patients with idiopathic bilateral TN, 2 studies covered patients with atypical idiopathic TN, and 1 study had patients with classical TN.

Across all studies, the pooled sample size consisted of 3,288 participants. Of these, 2,537 participants had undergone SRS (including LINAC and GK), while 751 underwent neuroablative procedures. Participant demographic characteristics are described in Table 1.

Patients with bilateral idiopathic TN were significantly more likely to undergo neuroablative procedures (p < 0.05), while those with atypical idiopathic TN were significantly more likely to undergo SRS (p < 0.01). The SRS cohort had a significantly higher proportion of patients with a V1 pathology as compared to the neuroablative cohort (p < 0.001), while the neuroablative cohort had significantly higher proportions of patients with V2 and V3 pathologies (p < 0.001). The SRS cohort also reported a significantly higher mean follow-up (p < 0.01), and a larger proportion of patients who had previously undergone another surgical procedure for TN (p < 0.05). Within the SRS cohort, mean GyMax delivered was 83.27 Gy, with a range of 70 Gy–95 Gy.

BNI outcomes (Fig. 1) were initially assessed in relation to demographic factors. Average BNI outcomes were significantly correlated (Fig. 3a) with participant sex and the male:female ratio (r = −0.59, 95% CI [−0.80, −0.16], p < 0.001). No significant findings were found with respect to the relationship of age, study location, nerve root distribution, and TN aetiology with BNI pain outcomes (Fig. 3b). Specifically for the patients in the SRS cohort, there was a non-significant relationship between the dosage of radiation delivered (GyMax) and average BNI outcomes (r = 0.16, 95% CI [−0.25, 0.52], p = 0.4352). There was no significant difference observed in outcomes between patients undergoing GK and LINAC SRS (Fig. 3c).

Across all follow-up periods, neuroablative techniques led to significantly improved BNI outcomes compared to SRS (Fig. 4a), with 901 (35.0%), 428 (16.6%), 582 (22.6%), 388 (15.1%), and 238 (9.2%) patients reporting BNI I, II, III, IV, and V, respectively, in the SRS cohort (n = 2,537), compared to 478 (63.6%), 78 (10.4%), 83 (11.1%), 55 (7.3%), and 57 (7.6%) of patients reporting BNI I, II, III, IV, V, respectively, in the neuroablative cohort (n = 751). Median BNI was also significantly lower in the neuroablative cohort compared to the SRS cohort (1.0 vs. 2.0, p < 0.0001). Within the neuroablative group, PBC reported a significantly higher frequency of BNI I and II (80.4%, p < 0.001) compared to GR (67.4%) and RFT (74.0%), while also reporting a significantly lower frequency of BNI IV and V (6.2%, compared to 20.8% and 21.2% in GR and RFT, respectively, p < 0.001). No significant difference was observed in patients with idiopathic and primary TN, or between patients with a first-time surgical intervention and those who previously underwent another surgical procedure for TN.

Across all studies reporting outcomes for cohorts at 12 months following treatment (Fig. 4b), neuroablative techniques demonstrated significantly lower median reported BNI (2.0 vs. 3.0, p = 0.0023) compared to SRS, with 564 (32.0%), 86 (4.9%), 447 (25.4%), 303 (17.2%), and 361 (20.5%) patients in the SRS cohort (n = 1,761) reporting BNI I–V, respectively, compared to 142 (32.9%), 58 (13.5%), 144 (33.4%), 48 (11.1%), 39 (9.0%) in the neuroablative cohort (n = 431).

Similarly, within the follow-up period of 13 months–36 months (Fig. 4c), neuroablative techniques demonstrated significantly lower median reported BNI compared to SRS (1.0 vs. 2.0, p < 0.0001). In the neuroablative cohort (n = 690), 478 (69.3%), 57 (8.3%), 70 (10.1%), 79 (11.4%), and 6 (0.9%), patients reported BNI I–V, respectively, in comparison to 493 (25.5%), 529 (27.4%), 450 (23.3%), 284 (14.7%), 177 (9.2%) patients in the SRS cohort (n = 1,933).

Finally, for the cohort within the follow-up period of >36 months (Fig. 4d), neuroablative techniques continued to demonstrate significantly lower median reported BNI compared to SRS (1.0 vs. 2.0, p < 0.0001). In the neuroablative cohort (n = 547), 352 (64.4%), 72 (13.1%), 48 (8.8%), 31 (5.7%), and 44 (8.0%), patients reported BNI I–V, respectively, in comparison to 758 (41.6%), 268 (14.7%), 352 (19.3%), 259 (14.2%), 186 (10.2%) patients in the SRS cohort (n = 1,823).

A significant difference was also observed in the average BNI outcome over time for both treatment modalities (Fig. 4e), resulting from an overall lower BNI outcome reported in patients undergoing neuroablative techniques (1.398 vs. 2.118, p < 0.0001) across all time points. A significant difference was not found in the rates of change of average BNI outcome between the treatment modalities.

To examine recurrence and side-effects at a comparable scale, we analysed BNI outcomes at all time points across all studies between 12 months and 5 years following intervention, with measurements taken beyond this period excluded. Overall recurrence was significantly lower in the neuroablative cohort in comparison to the SRS cohort (p < 0.0001), with 1,070 (42.2%) patients experiencing some degree of recurrence of TN in the SRS cohort, in comparison to 169 (22.5%) patients in the neuroablative cohort (Fig. 5a). Within the three neuroablative interventions, patients undergoing PBC reported significantly lower rates of recurrence (12.9%, p < 0.0001) compared to GR (30.0%) and RFT (28.8%).

The most commonly reported adverse effect across all treatment modalities was hypoesthesia/facial numbness. Patients undergoing SRS were significantly less likely to develop hypoesthesia following treatment compared to the neuroablative cohort (p < 0.0001), with 478 (18.8%) SRS patients reporting some degree of hypoesthesia at last follow-up, compared to 379 (50.5%) patients undergoing neuroablative procedures (Fig. 5b). Within the neuroablative treatments, patients undergoing GR were significantly more likely to report hypoesthesia (63.4%, p < 0.001), followed by PBC (45.0%) and RFT (37.0%).

Other adverse effects were grouped into minor and major adverse effects. Minor adverse effects included masseter muscle weakness, corneal anaesthesia, herpetic reactivation, minimal haematoma formation, vomiting, and headache. Major adverse effects included diplopia, hearing disturbance, swallowing and speaking difficulties, corneal blindness, large haematoma formation, anaesthesia dolorosa, and meningitis. Neuroablative techniques were associated with significantly higher levels of reported minor adverse effects (p < 0.0001), occurring in 147 (19.6%) patients compared to only 86 (3.4%) patients in the SRS cohort. Neuroablative techniques were also associated with significantly higher levels of major adverse effects compared to SRS (p < 0.001), occurring in 17 patients (2.3%) compared to 34 patients (1.3%) in the SRS cohort (Fig. 5c).

This study found that neuroablative techniques yielded significantly improved median pain outcomes across all follow-up periods, with significantly lower overall recurrence rates than SRS. Specifically, the largest improvement in BNI outcomes was observed for follow-up periods >12 months, where patients undergoing neuroablative techniques reported substantially higher proportions of BNI I and II compared to their SRS counterparts [16]. Even for follow-up periods >36 months, neuroablative techniques continued to display significantly improved pain outcomes.

Additionally, it was found that female sex was positively and significantly correlated with increased BNI pain scores, which is in line with the consensus regarding the more frequent and intense presentation of TN in women [2]. Patient age, nerve root distribution, TN aetiology, and SRS dose were not shown to have any significant relationships with BNI pain outcomes. On the other hand, neuroablative techniques were shown to have a significantly poorer side effect profile than SRS. Approximately 1 in 5 patients who underwent neuroablation reported minor adverse events compared to only 3.4% of SRS patients. While rare overall across both treatment modalities, neuroablative techniques were associated with a statistically significant increased rate of major adverse effects compared to SRS (2.3% vs. 1.3%).

This systematic review of 37 featured populations of comparable age across all treatment modalities. This study included 26 SRS (including LINAC and gamma knife) studies (n = 2,537), and 13 studies outlining the efficacy of neuroablative techniques including radiofrequency thermocoagulation (n = 146), PBC (n = 318), and GR (n = 287). With respect to SRS, we believe our study provides novel findings across larger samples in comparison to the recent European Academy of Neurology guidelines [9], which reviewed 8 gamma knife radiosurgery single intervention trials (n = 1,168). On the other hand, our study was unable to reach a comparable sample size to the guidelines (n = 5,577) within the neuroablative cohort, particularly due to its reliance on the BNI scale. This was particularly relevant within the RFT cohort and may have potentially resulted in the selection of studies that highlight a trend that would otherwise be absent in larger samples/without the restriction to the BNI scale. On the other hand, despite a smaller sample size, the BNI scale adds a substantial level of granularity to each study’s findings, providing novel findings with regard to the character of pain relief each treatment modality can result in. Moreover, restricting outcomes to the BNI scale allows us to directly compare outcomes across both treatment cohorts using a standard scale of efficacy. As such, our study allows clinicians to better guide patients on the extent of pain relief they can receive and whether medication may be required, which to our knowledge, has not been described in the literature so far. This study explores outcomes across a longitudinal scale, with an overall mean follow-up of 45.0 months for SRS and 34.1 months for neuroablative techniques. As such, it presents novel findings with respect to the longitudinal outcomes of both treatments, particularly within the follow-up duration of >12 months, where the difference between the two treatment modalities tends to become more pronounced. At 12 months, however, no significant difference was observed in the overall reporting of BNI I between the two treatment groups. This is an interesting and somewhat contradictory finding since, unlike SRS, the pain relief following neuroablative techniques is often immediate, and hence would be expected to be larger at 12 months based on our overall results. We speculate this may be due to a level of bias within the studies reporting outcomes at 12 months. While studies with longer follow-up measurements that measured BNI at 12 months were included, there were many studies that solely evaluated efficacy at 12 months in both SRS and neuroablative cohorts. As such, it is possible that these studies may have inadvertently selected patients who were expected to have successful outcomes with both SRS and neuroablative techniques according to their baseline status, resulting in similar outcomes. Both cohorts also had significantly lower sample sizes at 12 months compared to other follow-up intervals, as only studies measuring BNI values at exactly 1 year following intervention (rather than a larger interval) were included. These factors, coupled with the fact that an interval of 12 months following treatment often falls within the window of maximal efficacy of SRS [17], may have resulted in the similarity observed. Further work within this timeframe may be needed to better characterize pain profiles at 1 year following intervention.

Current guidelines for TN state that MVD is favoured over SRS based on superior pain outcomes at 1–2 years (BNI: 1.0, 68–88% vs. 24–71%) and up to 4–5 years (BNI: 1.0, 61–88% vs. 33–56%) of postoperative follow-up [7, 9, 18]. The findings of this study suggest that neuroablative techniques produce superior pain relief along with decreased risk of recurrence when compared to SRS. In cases where MVD is contraindicated for patients, the results suggest that neuroablative techniques should be the next-line treatment for patients seeking pain relief. This is especially relevant to patients with pathologies in the V2 and V3 distribution of the trigeminal nerve, as these nerve roots may be more accessible through ablative procedures (Table 1), and hence may result in a higher degree of success. On the other hand, SRS has consistently been shown to have the lowest risk of complications amongst surgical approaches to TN, likely resulting from the lack of physically invasive intervention. Moreover, our findings suggest a significantly higher uptake of SRS in patients with V1 distributions or atypical pathologies compared to neuroablative techniques (Table 1). This may suggest clinicians may feel more comfortable with SRS in situations where access via neuroablation is complicated, or where the aetiology of disease is atypical. Finally, in cases where patients may be contraindicated to undergo neuroablative procedures (such as due to anticoagulation), SRS may be the only viable surgical intervention. Thus, while neuroablative techniques may offer higher levels of efficacy, patients and clinicians may nevertheless choose SRS based on the risks of morbidity from adverse effects, degree of invasiveness, and clinical factors such as nerve root distribution and contraindications. As ever, nuance is required in terms of the choice of treatment, with clinicians ensuring adequate consultation with patients based on the weightage of potential benefits against the risk of adverse effects in light of available alternative treatments to guide their decision.

This review presents novel findings with regard to the longitudinal outcomes of neuroablative techniques and SRS for patients with medically refractory TN. The pooled estimates, alongside relative study weighing, provide robust results based on which clinical decisions for these patients can be better informed, while also reducing variance. The findings are also highly replicable and reflect overall longitudinal trends observed across the literature.

The study is somewhat limited by its emphasis on longitudinal outcomes, warranting the inclusion of observational and cohort studies that inherently have lesser degrees of control over extraneous factors. Similarly, the inclusion criteria solely comprising articles that report BNI pain outcomes eliminates a variety of suitable studies due to an inability to directly translate results presented in other scales. Moreover, due to the restriction of articles to the BNI scale, certain treatment cohorts (such as RFT and GR) had relatively lower sample sizes in comparison to the overall cohort. This may potentially result in selection bias, as it is possible that the studies reporting outcomes in BNI may demonstrate results that are incongruent with the larger literature. Thus, our findings may require further evidence to draw definitive conclusions, especially when comparing treatment modalities within the neuroablative cohort, due to its smaller sample size. Another key limitation of the study lies in its inclusion of largely single-interventional studies (35/37), introducing a degree of selection bias. This may result in the skewed inclusion or representation of specific patient samples that could bias overall pain outcomes. While the inclusion criteria set out minimum age and sample size requirements, no restrictions were placed on population and clinical factors. Although there is a large scarcity of studies in the current literature and guidelines with respect to multi-intervention studies exploring surgical treatments of TN [9], their lack in our study could potentially affect the validity of our pooled results. Our study also had a significantly smaller sample size in both cohorts at the 12-month interval, potentially resulting in differences in results that are otherwise absent in the larger trend. A small portion of our sample (13.4%) also included patients who had undergone other procedures prior to their primary treatment with the intervention being studied. This patient group could potentially have had more resistant symptoms and were more prevalent in the SRS group, which may have introduced bias into the results. It is also important to note that heterogeneity in the delivery of treatment to patients across studies may also result in reduced validity of the pooled results. While we generally found the use of similar techniques across all treatment modalities (online suppl. Table [19‒55]), some studies had significant differences in treatment characteristics. For example, Kodeeswaran et al. [26] used a substantially larger volume of anhydrous glycerin in comparison to the other GR studies. Similarly, Son et al. [30] made 3 lesions in RFT, compared to single lesions in other RFT studies. While no significant differences were observed in BNI outcomes between these outlying studies and the larger treatment cohort, heterogeneity in treatment can introduce a level of confounding that may have affected the overall results, especially in the neuroablative cohort due to its lower sample size. A key factor to also consider is the distribution of nerve roots, wherein the majority of patients with a pathology in the V1 nerve root underwent SRS, indicating a potential level of selection bias within the studies of the SRS cohort. Finally, patients undergoing SRS were significantly more likely to have atypical idiopathic TN, and/or have undergone other surgical procedures prior to the studied treatment, which may have led to a proportion of the cohort being inherently more difficult to treat and consequently report worse outcomes.

Another important consideration is also the role of potential uncertainties within diagnosis, given that most patients had primary idiopathic TN. We cannot tell from the information available how many subjects would have met the current diagnostic criteria for idiopathic TN, and it is possible that any patients misdiagnosed with TN could have been unevenly split between the two study groups. While these patients cannot be directly identified in this review, the use of clearly defined and standard criteria for a diagnosis of TN in the literature could alleviate this uncertainty. Finally, this study had a significantly smaller sample size for the neuroablative cohort compared to the SRS cohort, primarily as a result of a scarcity of studies covering treatment outcomes of PBC, RFT, or GR using the BNI scale, limiting the overall study power achievable. The literature would likely benefit from further prospective and controlled work regarding the outcomes of SRS and neuroablative treatments, both in the short-term (up to the 12-month interval), and more longitudinally. Moreover, studies with independent assessments of pain phenotypes, reporting outcomes using a standardized metric (such as the BNI scale), would significantly enhance the comparability of different treatments across patient populations.

For patients with neurovascular compromise, MVD remains the preferred surgical treatment for TN assuming the absence of significant comorbidities or frailty. However, since many TN patients do not have clear evidence of neurovascular compression on imaging, clinicians and patients must evaluate risks and benefits of the various interventions to decide on the preferred course of action. This meta-analysis suggests that neuroablative techniques are likely to produce better pain relief than SRS in these patients, especially at follow-up intervals between 1 and 3-years and beyond. However, in cases of V1 pathologies, significant comorbidities, frailty, or high risks of adverse effects, SRS may be the preferred intervention due to a lower rate of potential complications or better targeting of the site of pathology.

Future clinical trials comparing pain outcomes and adverse effects following MVD, neuroablative techniques, and SRS would all be beneficial to determine the superiority of each intervention in patients with TN. In general, outcomes of neuroablation are less well studied than those of MVD or radiosurgery. While the findings obtained in this study are promising, an absence of high-quality clinical trials and studies continues to exist, both for primary and repeat treatments. As such, further prospective work could help inform guidelines for treatment, especially in cases where one or more treatments have been ineffective in treating TN.

A statement of ethics is not applicable because this study is based exclusively on published literature. This study was registered as a systematic review on PROSPERO (CRD42024557229).

The authors have nothing to disclose.

This study was not supported by any sponsor or funder.

Yash Akkara and Ciaran Scott Hill helped with conceptualization and design. Yash Akkara, Jolene Marie Singh, and Ciaran Scott Hill helped with data collection. Yash Akkara and Ciaran Scott Hill helped with data analysis. Yash Akkara, Jolene Marie Singh, Lewis Thorne, and Ciaran Scott Hill helped with article drafting, final manuscript, and editing.

The list of studies included within this analysis can be found in the online supplementary materials. Further enquiries can be directed to the corresponding author.