Safety and Efficacy of Laser Interstitial Thermal Therapy as Upfront Therapy in Primary Glioblastoma and IDH-Mutant Astrocytoma: A Meta-Analysis

Simple Summary Laser interstitial thermal therapy (LITT) is a minimally invasive, MRI-guided procedure that causes localized tissue hyperthermia. This results in the ablation of targeted cells and disruption of the blood–brain barrier. LITT offers an alternative treatment option to standard-of-care resection for glioma in addition to potentially enhancing drug delivery of adjuvant therapy. Here, we present the first meta-analysis since the fifth edition of the WHO Classification of Tumors of the Central Nervous System (WHO CNS5) to assess the current literature on the safety and efficacy of LITT in the upfront treatment of primary brain tumors. This study summarizes the current relevant literature on LITT and supports its viability as a treatment option for glioblastoma and IDH-mutant astrocytoma. Abstract Although primary studies have reported the safety and efficacy of LITT as a primary treatment in glioma, they are limited by sample sizes and institutional variation in stereotactic parameters such as temperature and laser power. The current literature has yet to provide pooled statistics on outcomes solely for primary brain tumors according to the 2021 WHO Classification of Tumors of the Central Nervous System (WHO CNS5). In the present study, we identify recent articles on primary CNS neoplasms treated with LITT without prior intervention, focusing on relationships with molecular profile, PFS, and OS. This meta-analysis includes the extraction of data from primary sources across four databases using the Covidence systematic review manager. The pooled data suggest LITT may be a safe primary management option with tumor ablation rates of 94.8% and 84.6% in IDH-wildtype glioblastoma multiforme (GBM) and IDH-mutant astrocytoma, respectively. For IDH-wildtype GBM, the pooled PFS and OS were 5.0 and 9.0 months, respectively. Similar to rates reported in the prior literature, the neurologic and non-neurologic complication rates for IDH-wildtype GBM were 10.3% and 4.8%, respectively. The neurologic and non-neurologic complication rates were somewhat higher in the IDH-mutant astrocytoma cohort at 33% and 8.3%, likely due to a smaller cohort size.


Introduction
In 2021, the World Health Organization (WHO) reclassified primary central nervous system (CNS) tumors by incorporating molecular markers, as opposed to solely relying on histologic and genetic features as they had in years past [1][2][3][4].This integration of novel diagnostic technologies, including next-generation sequencing and DNA methylation profiling, resulted in the publication of the fifth edition of the WHO Classification of Tumors of the CNS (WHO CNS5) [3,5].The updated nomenclature recategorizes IDHmutant glioblastoma (GBM) without a 1p/19q co-deletion under the umbrella of IDHmutant astrocytoma that may be classified as WHO grade 2-4 depending on further molecular profiling.Instead, GBM became defined by its IDH-wildtype status, as it is commonly recognized today [3,6,7].While advancements in neuroimaging techniques and the implementation of 5-ALA in intraoperative fluorescence mapping have augmented surgeons' ability to obtain gross total resection, instances remain in which surgery is not an option or not preferred by the patient [7][8][9][10][11][12].Furthermore, certain factors are associated with reduced response to resection with adjuvant chemoradiation, such as IDH-wildtype GBM without methylguanine methyltransferase (MGMT) promoter methylation, poor tumor excisability, and resistance to chemotherapy [13,14].In cases of unresectable tumors, resistance to adjuvant therapies, deep lesions, significant comorbidity, and, to an extent, tumors located proximal to eloquent structures, minimally invasive laser-interstitial thermal therapy (LITT) is considered an effective alternative treatment modality for primary CNS tumors [15,16].
LITT (also known as stereotactic laser ablation) involves stereotactic navigation in delivering laser energy using an inserted optical fiber that ablates the tumor and has secondary effects of forming a hypoxic antitumor microenvironment by local hyperthermia, cellular necrosis, and focal coagulation [17][18][19].The earliest case of LITT for brain tumors was reported in 1983, but its efficacy was limited [15,20,21].The later development of magnetic resonance-guided LITT (MR-LITT) introduced high-resolution imaging with real-time magnetic resonance imaging (MRI) thermometry and spurred the widespread application of LITT for the management of CNS tumors, radiation necrosis, and epilepsy [15,22].Given its minimally invasive nature, LITT is often preferable in patients with significant comorbidities, deep-seated lesions, low functional scores, or, to an extent, an inability to tolerate anesthesia [23,24].Relative to standard open surgery for tumor resection, LITT is hypothesized to have equivalent efficacy with improved recovery time, shorter hospital stay, and lower physiologic impact of blood loss [12,25,26].Although LITT may be complicated by hemorrhage, seizure, edema, and radiation to neighboring tissue, the incidence rate of these complications is generally lower than in other invasive treatment modalities [12].
Currently, there are multiple studies investigating the possibility of multimodal treatment in combination with LITT to provide synergistic delivery of other therapeutics to the tumor microenvironment [27][28][29][30][31]. Beyond just directed tumor cell death, LITT is also hypothesized to decrease the stability of tight junctions and the rate of transcytosis across the blood-blood-brain barrier (BBB) [32,33].In doing so, LITT increases the permeability of the BBB and enables the greater delivery of therapeutics to target CNS neoplasms by presumptively increasing chemokine production and antigen presentation [32,34,35].In this way, LITT offers a unique advantage in comparison to non-stereotactic treatment modalities by enhancing the delivery of immunotherapies, as well as other treatments [36][37][38][39][40][41][42][43].
Over the years, several retrospective and prospective studies have reported the efficacy of LITT [44][45][46].Here, we present a comprehensive summary of the most recent advances since WHO CNS5 was released in LITT as an upfront treatment for primary tumors.

Data Collection
Original studies using LITT for primary brain neoplasms were identified per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [47].Boolean combinations of (Primary Brain Tumor OR Central Nervous System Tumor OR Brain Tumor OR Intracranial Tumor OR Intracerebral Neoplasm) AND (LITT OR Laser Interstitial Thermal Therapy) were queried across four databases (PubMed, Embase, Cochrane, and Web of Science).All studies were included based on LITT's first descriptions from 1983 and after with no specified time range [48].Two independent reviewers screened all studies using the Covidence systematic review manager.Articles reporting primary tumors treated with LITT without prior intervention in the adult patient population were included.The term 'primary' included all non-recurrent, non-metastatic tumors that originated in the brain.Non-human, pediatric, and pregnant patients were excluded, as were duplicated articles, non-English manuscripts, case reports, and review articles.

Data Extraction
All reported data from studies was extracted using the Covidence systematic review manager.Demographic data such as age, sex, tumor type, and other pertinent clinical data were included.LITT-specific variables of interest included laser type, laser developer, equipment power, and other equipment specifications.Preoperative variables of interest included preoperative tumor volume, tumor location, tumor laterality, methylation status, mutational profile, and preoperative Karnofsky performance score.Clinical outcomes included postoperative tumor volume, the extent of ablation, and resulting complications.To consolidate heterogeneous descriptions of postoperative complications, adverse events pertaining to neurologic outcomes were defined as "neurologic complications", which comprised "transient aphasia", "seizure", "motor function deficits", and "hemiparesis".Some studies grouped their neurologic complications and reported them only as 'neurologic deficit' without further specification.These have been grouped together as "non-specified neurologic deficits".Other complications were deemed "non-neurologic complications" which comprised: "edema", "hemorrhage", "hydrocephalus", "pulmonary embolism", "DVT", "wound infection", and "meningitis".Overall survival, progression-free survival, and mortality rates were also recorded.

Statistical Analysis
All data were analyzed using the "meta" function in Stata 18 Standard Edition (SE) [49][50][51].During the abstract screening, single patient reports and studies were not included in the analysis due to limited sample size (Figure 1).Due to the limited sample size of studies reporting primary meningiomas and oligodendrogliomas, we excluded them.Overall, we statistically analyzed two cohorts: patients diagnosed with IDH-wildtype GBM, and patients diagnosed with IDHmutant astrocytoma.Descriptive data were reported as means with standard deviations.Due to limited data availability across selected studies, pooled proportional analysis was performed to calculate pooled percentages for all binomial study characteristics.Study heterogeneity was quantified via the Cochrane Q statistic.Forest plots were used to summarize individual study proportions, study heterogeneity, and the pooled proportional effects observed throughout the study.Meta-regressions were used to evaluate the correlations between proportional data and continuous variables of interest while subgroup analysis was performed amongst categorical data evaluating proportional data of interest.All binary data were reported as pooled percentages while continuous data were reported as pooled means with standard deviations.Publication bias was evaluated via contour-enhanced funnel plots at contour significance levels of 1%, 5%, and 10%.Egger's test was utilized to evaluate if funnel plot asymmetry leading to publication bias was statistically significant.Random effects modeling (REML) was utilized to account for inter-study heterogeneity.Sensitivity analysis of outcomes with no significant study heterogeneity was performed if the Cochrane Q statistic was insignificant (p > 0.05) using a common-effects inverse variance model.A p-value ≤ 0.05 was deemed significant for all other statistical testing.
icance levels of 1%, 5%, and 10%.Egger's test was utilized to evaluate if funnel plot asymmetry leading to publication bias was statistically significant.Random effects modeling (REML) was utilized to account for inter-study heterogeneity.Sensitivity analysis of outcomes with no significant study heterogeneity was performed if the Cochrane Q statistic was insignificant (p > 0.05) using a common-effects inverse variance model.A p-value ≤ 0.05 was deemed significant for all other statistical testing.

Search Results
Following the specified search criteria, a total of 875 deduplicated studies were screened for inclusion (Figure 1).Following the abstract screening, 621 studies were excluded if the inclusion criteria was not met, and the search terms were not appropriately

Search Results
Following the specified search criteria, a total of 875 deduplicated studies were screened for inclusion (Figure 1).Following the abstract screening, 621 studies were excluded if the inclusion criteria was not met, and the search terms were not appropriately included.During full text screening, the top reasons why we excluded studies involved the reporting of pooled data in which outcomes for metastatic tumors were not differentiated, were not stratified by tumor type, or stratified for recurrent and primary cohorts.Ultimately, we excluded another 241 studies and 22 were included after the completion of the full-text review.A majority of the studies were screened out due to presenting data that were unstratified by tumor type, metastatic tumors, and recurrent tumors.Among this cohort of 206 patients, 16 studies included LITT for primary IDH-wild type glioblastoma (N = 185, 90%), while 6 studies reported on LITT for IDH-mutant astrocytoma (N = 21, 10%) (Table 1).The patient demographic data and characteristics from the 22 included studies differentiated by tumor type (GBM or astrocytoma).*: Johnson et al. [53] is reported twice for independent astrocytoma grades, the first representing two patients with astrocytoma grade 2 and the second with three patients with astrocytoma grade 3. †: Murayi et al. [56] is reported twice for two independent tumor type cohorts: the first reports astrocytoma and the second reports only GBM patients from the same study.IDH1: Isocitrate dehydrogenase 1; MGMT: O(6)-methylguanine-DNA methyltransferase; KPS: Karnofsky performance status.

IDH-Mutant Astrocytoma
The average of patients presenting with IDH-mutant astrocytoma was 38.6 years (SD: 6.3, Range: 33.5-51.0),roughly 60% of patients were male.About 44% of the IDH-mutant astrocytoma were located in the frontoparietal region, 16% were in the thalamus and temporal lobe, 11% were located in the insula, while the remaining 5.6% were either in the corpus callosum or occipital lobe.Additionally, the average preoperative tumor volume was 13.4 cm 3 (SD: 13.4,Range: 3.1-23.7)with similar lateralization to the right (44.8%) and left (55.2%) hemispheres.Initial KPS was 81.7 on average (SD: 7.6, Range: 75.0-90.0).IDH status was not explicitly reported for 25% of the astrocytoma included in the analysis.
The patient postoperative outcomes from the 22 included studies differentiated by tumor type (GBM or astrocytoma).*: Johnson et al. [53] is reported twice for independent astrocytoma grades, the first representing two patients with astrocytoma grade 2 and the second with three patients with astrocytoma grade 3. †: Murayi et al. [56] is reported twice for two independent tumor type cohorts: the first reports astrocytoma and the second reports only GBM patients from the same study.Note: The postoperative outcomes are reported from the patient's last follow-up at the time of each study.Mean KPS postoperative scores were recorded at

Meta-Regression Analysis
Meta-regression analysis was performed between all preoperative characteristics and postoperative outcomes of interest.While no significant correlation was found between IDH-wildtype GBM and IDH-mutant astrocytoma cohorts for EOA, tumor progression, complication rates, tumor volume, or demographic characteristics, a significant negative correlation was noted in three studies between postoperative KPS and tumor progression (β = −0.07,p < 0.0001).Tumor progression was also negatively correlated with MGMT methylation (β = −33.7,p < 0.0001).

015). (D)
There is a significant negative relationship between postoperative KPS and tumor progression (p < 0.001).

001). (B)
There is a significant negative relationship between the proportion of MGMT methylation and neurologic complications (p = 0.002).(C) There is a significant negative relationship between the proportion of MGMT methylation and non-neurologic complications (p = 0.015).(D) There is a significant negative relationship between postoperative KPS and tumor progression (p < 0.001).

Adjuvant Therapy in Combination with LITT
We reviewed adjuvant therapies (ATs) used post-LITT among the included studies.Temozolomide, dexamethasone, bevacizumab, lomustine, chemotherapy, tumor treating fields, and pembrolizumab with and without radiation therapy were the reported ATs used post-LITT (Supplementary Table S2).The most commonly occurring AT reported in this study is chemoradiation in the form of temozolomide + radiotherapy (46.5%), as seen in the IDH-wildtype GBM cohort.Although close to 50% of the IDH-wildtype GBM cohort reported the use of ATs post-LITT, just 18% of the IDH-mutant astrocytoma cohort reported receiving any.Overall, the included articles report that the usage of ATs resulted in a higher PFS and OS on average [16,60].

Publication Bias
Publication bias was utilized to evaluate the variance in reporting of LITT efficacy through pooled rates of neurologic and non-neurologic outcomes.Egger's test did not reveal statistically significant publication bias for non-neurologic complications in either the IDH-wildtype GBM or IDH-mutant astrocytoma cohorts.However, significant publication bias was found to influence the reporting of neurologic complications in studies that evaluated LITT for IDH-mutant astrocytoma (β = −6.2,p < 0.01) (Supplementary Figure S2).Specifically, sample size was found to influence significance level, with publication bias against smaller studies.

Discussion
To the best of our knowledge, this is the first meta-analysis since the 2021 WHO CNS5 guidelines to evaluate the outcomes of LITT as initial therapy in both primary IDH-wildtype GBM and IDH-mutant astrocytoma.Currently, LITT is often used in cases of deep-seated lesions, when there are contraindications to radiotherapy or resection, proximity to eloquent regions, or in cases where patients are particularly vulnerable to wound complications [70][71][72].In the past, LITT has been shown to reduce postoperative complications while maintaining high ablation rates [69,73].This study summarizes the current relevant literature on LITT and discusses its viability as an upfront treatment option in the management of primary IDH-wildtype GBM and IDH-mutant astrocytoma.

Survival and Function
The pooled PFS and OS rates for the IDH-wildtype GBM subgroup were 5 and 9 months, respectively.Prior studies have reported similar survival data following surgical resection with PFS and OS of 5 months and 10 months, respectively [25,74,75].However, it is difficult to place these survival rates in the context of the current literature as many studies combined IDH-wildtype GBM with IDH-mutant astrocytoma when reporting survival rates.Given this heterogeneity in reporting, survival rates in these cohorts are highly variable.One example of this was a 2016 meta-analysis by Ivan et al. in which grade 3 astrocytoma were included with IDH-wildtype GBMs and thus OS was reported as higher at 14.2 months [14].Also, their conclusions were limited by study power given their analysis only included 3 papers.Our study experienced similar limitations in that LITT is often recommended in instances of unresectable tumors presenting with associated edema, thus potentially biasing the patient selection.While we stratified by tumor type, Cancers 2024, 16, 2131 9 of 16 a secondary limitation was that we were unable to do so by tumor grade given limited patient numbers.Future studies should stratify patients by factors including tumor size, depth, and grade to more accurately reflect LITT's efficacy compared to standard resection.
One factor that was found to improve survival in our analysis was MGMT methylation status.We found that MGMT-methylated tumors were significantly correlated with improved PFS after LITT (p < 0.001), which is consistent with prior studies [16,76].We also found a significant reduction in postoperative complication rates among patients with MGMT methylation status, suggesting that LITT is particularly safe and effective in these tumors.
In patients whose GBM did not progress during the study period, their KPS score significantly improved following LITT, which is consistent with the previous literature [25,77].Even in those with recurrence, the safety of LITT has been demonstrated by patients with KPS scores that remain stable long after surgery [69].However, in this study, the pooled average KPS score decreased somewhat in the immediate postoperative period.The extent and severity of this decline was limited by variations in reporting and lack of long-term follow-up.
Overall, our data suggest that LITT may be a safe primary management option and warrants further investigation through randomized controlled studies to better characterize survival outcomes.

Extent of Ablation
The noted EOA for the IDH-wildtype GBM subgroup (94.8%) is similar to previous studies which report a range between 91 and 99% for newly diagnosed tumors [16,26,62,73].Although preoperative tumor volume was not controlled, the high EOA supports LITT's ability to achieve maximal safe ablation in a variety of tumor sizes and locations.The average EOA in the IDH-mutant astrocytoma subgroup was 84.6%, similar to the existing literature.For instance, the study by Ivan et al. reported a tumor ablation rate of 82.9% [14].However, this study had high heterogeneity with regard to tumor types and molecular profiles included [14].Past studies have indicated that a 78% extent of resection is the threshold for significantly improved overall survival in patients undergoing surgery [33,78,79].The extent of ablation in both cohorts exceeded this threshold on average, though it is unclear if this percentage should be applied given its original context was in regard to the extent of the resection.Maximization of EOA involves optimizing LITT-associated parameters such as optimal trajectory, integration of intraoperative neuromonitoring, and a risk assessment of lesion-to-tract distance to prevent significant disability in tracts thermally susceptible [48,80].More prospective studies are needed to address the role of EOA on both survival and postoperative neurologic deficit to determine ideal parameters for these cases.

Postoperative Complications
In our analysis, LITT for IDH-wildtype GBM demonstrated lower postoperative neurologic (10.3%) and non-neurologic (4.8%) complication rates relative to those reported after other treatment modalities noted by the literature.Advancements in LITT have resulted in lower complication rates in comparison to the Viozzi et al. study that reported a rate of 33.7% in newly diagnosed GBM post-LITT [81].A prospective study by Zetterling et al. examined complications following surgical resection using 5-ALA contrast and found that 41% experienced neurologic deficits postoperatively [82].In another prospective analysis, Gempt et al. found ischemic lesions following surgery for GBM in 31% of cases, 24% were symptomatic [83].Similarly, other national outcome studies report high complication rates between 13 and 19% after craniotomy [84][85][86][87][88].This appears true with regard to patients undergoing hypofractionated radiotherapy as well [78,79].As reported by Floyd et al., late toxicity effects based on the Radiation Therapy Oncology Group neurotoxicity scores, were reported in 50% of individuals with an average radiation dosage of 50 Gy applied during hypofractionated radiotherapy [79].Our findings suggest a comparable extent of ablation/resection with lower complication rates in patients with GBM primarily treated with LITT as opposed to surgery or radiation.Interestingly, we found neurologic complication rates that were three times as high for IDH-mutant astrocytoma when compared to GBM (33% vs. 10%).However, the significance of this is difficult to interpret given the low sample size in the IDH-mutant astrocytoma cohort and the variability in IDH-mutant astrocytoma grades included [89,90].Aside from neurologic complications in the IDH-mutant astrocytoma cohort, all other complication rates for both tumor subtypes were similar to the neurologic rates presented in the literature ranging from 2.17% to 7.14% [80,91,92].Given that LITT is often used for deep-seated tumors located near eloquent structures, these relatively low complication rates are reassuring.While LITT is a minimally-invasive technique, technical complications may arise, such as hyperthermic deposition, catheter placement, and malfunctioning of the cooling system [93].Institutional recognition and reporting of these complications is necessary for improving the software and hardware involved.

Adjuvant Therapies
Currently, adjuvant chemotherapy is recommended along with maximal surgical resection.However, 30% of GBM cases are considered inoperable [94].The ability of LITT to target these tumors in concert with AT to augment the penetration of the peritumoral BBB highlights its utility in multimodal management [32].Leuthardt et al. reported increased peritumoral BBB permeability from weeks 1 to 2 and 4 to 6, highlighting the ability of LITT to enhance drug delivery to allow for the use of certain adjuvant therapies [37].Although we were unable to run a meta-analysis for multimodal therapies in combination with LITT due to lack of enough data points and inconsistent reporting, our study was able to systematically review and characterize post-LITT adjuvant therapies from the few studies that did not include them.The most commonly occurring AT reported in this study is chemoradiation in the form of temozolomide + radiotherapy (46.5%), while the use of the immunotherapeutic pembrolizumab alone was the least utilized AT in the GBM cohort.Based on our findings, adjuvant therapies appear to provide improved survival outcomes after LITT, and this is reflected in the literature.For example, one multimodal study of LITT and chemoradiation for newly diagnosed IDH-wildtype GBM found an OS of 16 months and PFS of 12 months with combined treatment versus an OS of 10 months and PFS of 6 months after LITT alone [16].Currently, the active expedited laser interstitial thermal therapy and chemoradiation clinical trial (NCT02970448) is evaluating the feasibility of LITT with standard concurrent radiotherapy and temozolomide in newly diagnosed highgrade gliomas [95].In the GBM cohort, two individuals (1.1%) received tumor-treating field therapy, which delivers low intermediate frequency electric fields to selectively remove proliferative cells, increase survival outcomes, and reduce recurrence [56].While the adjuvant immunotherapeutic pembrolizumab was administered in only one GBM patient in our cohort (0.5%), it has been shown to increase peripheral immune cell recruitment and immune response in GBM [16,96].Although a current phase I/II study (NCT02311582) is evaluating the in situ vaccination effect of LITT using PD-L1 inhibitor pembrolizumab at varying dosages for recurrent tumors, similar trials need to be conducted for primary CNS tumors [97].Prospective studies are needed to evaluate the potential of adjuvant LITT to improve drug delivery, radiotherapy efficacy, and survival.

Limitations and Future Steps
This meta-analysis had inherent limitations due to nonuniform reporting of demographics and outcomes amongst the included studies, as well as overall diminished sample size, especially in the IDH-mutant astrocytoma cohort.The lack of standardized technical parameters in LITT, such as ablation duration or temperature and operator experience, resulted in variability between studies.Additionally, the methodology of EOA measurement was inconsistently reported by studies, often being limited to MRI utilized post-procedurally.We utilized random-effects modeling in the meta-analyses to account for interstudy heterogeneity.Due to limited sample sizes and study count, tumor types such as meningiomas and oligodendrogliomas were not included.As different tumors are characterized by varying growth patterns and invasion, we performed subgroup analyses by tumor type and excluded metastatic and recurrent tumors.Within the IDH-mutant astrocytoma cohort, we did not differentiate by tumor grade as we were limited by the number of studies and heterogeneity in reporting.In terms of genetic characteristics, IDH1 mutation and MGMT methylation could not be extracted for studies in which these statuses were not reported for the entire study cohort.Consequently, any IDH-wildtype tumors reported in the studies, but unidentified for IDH mutation, could not be reclassified based on the WHO CNS5 classification.This highlights the necessity of future research to report all pertinent genetic characteristics, especially due to potential revisions to the WHO CNS5 stratification for tumor type.Thus, future meta-analyses can reclassify such tumors, offering insights into survival outcomes based on the most current tumor classification criteria.Further, we were not able to control for follow-up time in our study, which resulted in variability of the reported outcomes.Only the IDH-mutant astrocytoma group appeared to have significant publication bias in association with neurologic outcomes, which is likely due to the small sample size of this cohort.Nonetheless, this is the first proportional meta-analysis exploring the utility of LITT in the management of primary CNS tumors following the updated WHO CNS5 stratification by tumor type.In the evaluation of treatment modalities, this study assumes EOR and EOA are comparable in terms of tumor removal, modeling after previous retrospective studies [77,98].Adoption of a more uniform procedure and data reporting may aid in future analysis of the efficacy of adjuvant therapies post-LITT.Future prospective studies are required to validate these results and for direct comparison of LITT to other forms of treatment.Large-scale studies may also aid in the evaluation of differences in outcome, wavelength, cooling modality, and laser power between the three FDA-approved LITT systems Monteris NeuroBlate ® , Medtronic Visualase ™ , and ClearPoint Prism ® in optimizing their parameters [12,48,70,99,100].

Conclusions
We present the first meta-analysis of clinical outcomes following LITT since the publication of the 2021 WHO CNS5 guidelines for two of the most common primary CNS tumor types.We found that LITT for IDH-wildtype GBM provides acceptable EOA with similar to reduced complication rates when compared to surgical resection.Our investigation into IDH-mutant astrocytoma demonstrated slightly lower EOA on average with higher complications, though it is difficult to draw any robust conclusions from this cohort given its extremely limited size of just 21 patients.However, these findings combined with the unique ability of LITT to enhance drug delivery and radiotherapy efficacy suggests its utility as an adjuvant treatment modality bears further investigation.Larger prospective studies are needed to establish clear recommendations for LITT as a primary treatment option for these challenging lesions.

Figure 1 .
Figure 1.PRISMA flowchart for the filtering of 1402 primary scientific articles into 22 studies included in the meta-analysis.

Figure 1 .
Figure 1.PRISMA flowchart for the filtering of 1402 primary scientific articles into 22 studies included in the meta-analysis.

Figure 2 .
Figure 2. MGMT methylation status, tumor progression, complications, and post-op KPS.The metaregression bubble plots, only included for the IDH-wildtype GBM subgroup, were utilized to evaluate the correlation coefficient between two patient parameters [16,57,60,63,66,69]. (A) There is a significant negative correlation between the proportion of MGMT methylation and tumor progression (p < 0.001).(B)There is a significant negative relationship between the proportion of MGMT methylation and neurologic complications (p = 0.002).(C) There is a significant negative relationship between the proportion of MGMT methylation and non-neurologic complications (p = 0.015).(D) There is a significant negative relationship between postoperative KPS and tumor progression (p < 0.001).

Figure 2 .
Figure 2. MGMT methylation status, tumor progression, complications, and post-op KPS.The metaregression bubble plots, only included for the IDH-wildtype GBM subgroup, were utilized to evaluate the correlation coefficient between two patient parameters [16,57,60,63,66,69]. (A) There is a significant negative correlation between the proportion of MGMT methylation and tumor progression (p < 0.001).(B)There is a significant negative relationship between the proportion of MGMT methylation and neurologic complications (p = 0.002).(C) There is a significant negative relationship between the proportion of MGMT methylation and non-neurologic complications (p = 0.015).(D) There is a significant negative relationship between postoperative KPS and tumor progression (p < 0.001).

Table 1 .
Patient Demographics and Characteristics.
1 month follow-up in Jamshidi et al. and Beaumont et al. while at 3 months follow-up for deGroot et al. and Viozzi et al.EOA: Extent of ablation; OS: Overall survival; PFS: Progression-free survival; KPS: Karnofsky performance status.