Hemodynamic Angioarchitecture as a Determinant of AVM Radiosurgery Outcome: A Systematic Review

Hemodynamic Angioarchitecture as a Determinant of AVM Radiosurgery Outcome: A Systematic Review | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Systematic Review Hemodynamic Angioarchitecture as a Determinant of AVM Radiosurgery Outcome: A Systematic Review RAHUL MODI, Puneet Pareek This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9386788/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Stereotactic radiosurgery (SRS) is an established treatment for intracranial arteriovenous malformations (AVMs). Current grading systems—principally the modified Radiosurgery-Based AVM Score (mRBAS)—assign all non-eloquent lobar locations to a single risk tier and incorporate only binary venous drainage depth as an angioarchitectural variable. 1 Whether granular angioarchitecture—encompassing transit time, draining vein number and depth, feeder artery calibre, and perinidal angiogenesis—independently modifies radiosurgical outcome beyond currently graded anatomical location has not been systematically synthesized. Methods A PRISMA 2020-compliant systematic review of PubMed/MEDLINE, Cochrane Central, Embase, and Scopus was performed (1990–2026). Studies reporting SRS outcomes stratified by anatomical location or angioarchitecture variables were included. Primary outcome was complete obliteration confirmed on imaging. Secondary outcomes included symptomatic radiation-induced changes (RICs), post-SRS hemorrhage, and functional status (modified Rankin Scale). Pooled analysis employed DerSimonian–Laird random-effects meta-analysis; heterogeneity was assessed with I 2 and Cochran's Q. This review is registered with PROSPERO (CRD420261356041). Results Forty-six studies encompassing approximately 38,000 patients met inclusion criteria. Overall pooled obliteration was 69.4% (95% CI 65.2%–73.6%; I 2 = 87.4%). Within mRBAS Tier-0 locations, pooled obliteration ranged from 82%–88% (frontal) to 70%–78% (occipital; Kruskal–Wallis p < 0.001). Angioarchitecture variables demonstrated substantially larger effect sizes: transit time (OR 4.3; 95% CI 3.2–5.8; p < 0.001), 2,3 draining vein number (OR 3.1; 95% CI 2.5–3.8; p < 0.001), and venous drainage depth (OR 2.4; p = 0.008) each individually exceeded the location-tier effect. Perinidal angiogenesis (OR 0.26) 4 and feeder artery enlargement (OR 0.30) were the most adverse individual predictors. Conclusions Hemodynamic angioarchitecture—as assessed on pre-treatment digital subtraction angiography (DSA)—is the primary determinant of AVM radiosurgery outcome, overriding the prognostic advantage of favorable anatomical location. Angioarchitecture variables individually and collectively outperform the mRBAS location tier as predictors of obliteration. Prospective studies should mandate standardized pre-treatment DSA characterization of transit time, draining vein number, venous drainage depth, and perinidal angiogenesis as pre-specified primary outcome predictors. Oncology Arteriovenous malformation Stereotactic radiosurgery Angioarchitecture Obliteration rate Digital subtraction angiography Figures Figure 1 Figure 2 Figure 3 INTRODUCTION Brain arteriovenous malformations (AVMs) are congenital vascular lesions carrying an annual hemorrhage risk of approximately 2%–3% in the unruptured state and substantially higher rates following an index bleed. 5 Hemorrhage risk is not uniform; deep lesion location and deep venous drainage each independently increase rupture probability. 6 For surgically accessible, low-grade lesions, microsurgical resection guided by the Spetzler–Martin grading system has represented the historical standard, stratifying operative risk on nidus size, venous drainage pattern, and eloquence of adjacent cortex. 7 Stereotactic radiosurgery emerged as a definitive treatment alternative for surgically inaccessible or high-risk AVMs. The first large Gamma Knife series demonstrated obliteration in 72%–88% of patients with acceptable complication rates. 8 Subsequent dose–response analyses confirmed that obliteration probability depended on the margin dose, nidus volume, and anatomical location relative to eloquent structures. 9 The location–outcome relationship was codified into the SPIE score, assigning site-specific risk coefficients from 0 (frontal) to 10 (pons/midbrain). 10,11 The Radiosurgery-Based AVM Score (RBAS) incorporated nidus volume, patient age, and location, 12 and the modified RBAS (mRBAS) simplified location to a binary variable: non-eloquent sites scored 0 and eloquent deep structures scored 1. 1 Independent validation confirmed the mRBAS as the most widely applied radiosurgical grading instrument. 13,14 The ARUBA trial heightened the requirement for accurate pre-treatment outcome stratification by reporting that medical management was superior to interventional therapy for unruptured AVMs at a mean follow-up of 33 months, although these findings attracted substantial methodological criticism. 15 A growing but insufficiently synthesized body of evidence suggests that angioarchitecture—the internal hemodynamic vascular anatomy of the AVM—may exert a substantially larger influence on radiosurgical outcomes than location tier alone. One landmark study demonstrated obliteration rates ranging from 80% in low-flow lesions to only 38% in high-flow AVMs with multiple adverse angioarchitectural features. 4 Mean arterial transit time on DSA was subsequently confirmed as an independent predictor of both obliteration probability and time-to-obliteration. 3,2 Location-specific series have further refined outcome expectations within the mRBAS Tier-0 grouping. For cerebellar AVMs, infratentorial location did not significantly reduce obliteration in matched comparisons, 16 while multicenter data confirmed acceptable functional preservation after cerebellar SRS. 18 For temporal lobe AVMs, temporal location was independently associated with pre-existing epilepsy, with radiosurgery achieving meaningful seizure reduction—an endpoint absent from obliteration-based grading instruments. 17 At the unfavorable end of the spectrum, basal ganglia and thalamic AVMs yielded obliteration rates of 60%–68% with symptomatic complication rates reaching 18%–28%, 19 while brainstem series reported obliteration probabilities of 48%–65%. 20 Recent multicenter series reinforced the prognostic importance of draining vein number 21 and deep venous drainage across 8,673 patients. 22 Emerging quantitative DSA techniques further underscore the importance of angioarchitecture, with machine learning models trained on DSA-derived radiomics features predicting obliteration with accuracy exceeding conventional anatomical grading. 23 Meta-analyses of pre-SRS embolization consistently demonstrate no significant improvement in obliteration rates with combined treatment, 24,25,26 reinforcing the biological primacy of intrinsic AVM flow characteristics. Validated composite instruments such as the Virginia Radiosurgery AVM Scale 28 and linear accelerator outcome analyses 29 confirm that nidus volume and dose remain important predictors, but neither incorporates the full angioarchitecture predictor set. In the unruptured AVM population, long-term SRS outcomes support continued radiosurgery for carefully selected patients, 27 where accurate pre-treatment obliteration prediction is particularly important. This systematic review and meta-analysis aims to: (1) characterize the effect sizes of individual angioarchitecture variables as predictors of SRS obliteration, RICs, and post-SRS hemorrhage; (2) determine whether angioarchitecture variables demonstrate larger effect sizes than anatomical location tier in pooled analyses; and (3) quantify the interaction between lobar location and angioarchitecture profile on SRS outcomes. METHODS Study Registration This systematic review was prospectively registered with PROSPERO (registration number CRD420261356041) prior to data extraction. The review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Search Strategy Electronic searches were performed in PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, Embase, and Scopus from January 1, 1990 through March 30, 2026. The search combined MeSH and free-text terms: ("arteriovenous malformation" OR "AVM") AND ("stereotactic radiosurgery" OR "Gamma Knife" OR "LINAC radiosurgery" OR "CyberKnife") AND ("obliteration" OR "outcome") AND ("angioarchitecture" OR "draining vein" OR "transit time" OR "feeder artery" OR "perinidal angiogenesis" OR "location"). No language restriction was applied. Reference lists of included articles and relevant systematic reviews were hand-searched. Forward citation searching was performed for all included studies. Eligibility Criteria Studies were included if they: (1) enrolled patients of any age with intracranial AVMs treated with SRS; (2) reported obliteration rate as a primary or secondary endpoint; (3) provided data stratified by anatomical location or reported angioarchitecture variables as predictors; and (4) included ≥20 patients with minimum radiological follow-up of 24 months. Both ruptured and unruptured AVMs were eligible. Studies evaluating prior embolization or multimodality therapy were included if radiosurgery-specific outcomes could be extracted. Excluded were: case reports and series with < 20 patients, studies of dural arteriovenous fistulae or cavernous malformations, purely surgical or embolization-only series, conference abstracts without full outcome data, and studies in which SRS-specific outcomes could not be separated from combined multimodality results. Data Extraction and Risk of Bias Two reviewers independently extracted: study design, patient number, patient age, AVM volume, prescribed margin dose, obliteration rates by location subgroup, symptomatic RIC rates, post-SRS hemorrhage rates, angioarchitecture variables, and modified Rankin Scale (mRS) functional outcome. Discrepancies were resolved by consensus. Risk of bias was assessed using the Newcastle–Ottawa Scale for observational studies. Certainty of evidence was assessed using the GRADE approach. Statistical Analysis Pooled obliteration rates were calculated using the DerSimonian–Laird random-effects model. Heterogeneity was quantified using I 2 and Cochran's Q. Location-stratified obliteration rates were compared using Kruskal–Wallis one-way analysis with Tukey-equivalent post-hoc pairwise testing and Bonferroni correction. Independent predictors of obliteration identified in multivariate analyses were synthesized narratively and by pooled OR where data permitted. Publication bias was assessed using funnel plots and Egger's regression test. A two-sided significance threshold of p < 0.05 was applied throughout. Outcome Definitions Complete obliteration was defined as absence of AVM nidus on DSA or, where DSA was not performed, absence of flow voids and enhancement on 3T MRI with magnetic resonance angiography. Symptomatic RIC was defined as new neurological symptoms attributable to radiation-induced parenchymal change on imaging. Post-SRS hemorrhage was defined as any intracranial hemorrhage during the latency period from radiosurgery to confirmed obliteration. Favorable functional outcome was defined as mRS score 0–2 at last follow-up. RESULTS Study Characteristics Forty-six studies published between 1990 and 2026 met inclusion criteria, encompassing approximately 38,000 patients (range 20–8,673 per study). Single-center retrospective cohort studies comprised the majority (n = 38); 5 were multicenter series and 3 were meta-analyses or systematic reviews used to supplement pooled outcome data. Gamma Knife SRS was the treatment platform in 34 studies; LINAC-based systems in 8; and mixed or CyberKnife platforms in 4. Mean prescribed margin dose ranged from 16 to 25 Gy. Mean follow-up was 36–84 months. Egger's regression test did not demonstrate significant publication bias ( p = 0.14). Table 1 presents selected key series from the included literature. TABLE 1. Summary of selected included studies (n = 15 shown) Study (Year) N Platform Obliteration, % RIC, % Follow-up, mo Notes Lunsford et al., 1991 227 GK 72-88 9 24-96 Location-stratified Flickinger et al., 1996 197 GK 75 8 24-60 Dose-response Flickinger et al., 1999 332 GK/multi — 11 36+ Location; SPIE Pollock & Flickinger, 2002 356 GK 60-89 10 36-96 RBAS derivation Pollock & Flickinger, 2008 293 GK 64-82 9 36-84 mRBAS derivation Wegner et al., 2011 293 GK 60-80 10 36+ mRBAS validation Taeshineetanakul et al., 2012 139 GK 38-80 11 24-72 Angioarchitecture; flow Ding et al., 2014 134 GK 72-80 10 36-84 Cerebellar; draining vein Ding et al., 2015 444 GK 74 12 36-96 Temporal; seizure outcome Cohen-Inbar et al., 2017 343 GK/multi 72-80 11 48+ Cerebellar; multicenter Panni et al., 2020 191 GK Flow-dep. — 37-173 Flow; angioarchitecture Alzate et al., 2022 71 GK 38-80 — 28-47 Transit time Erickson et al., 2022 210 LINAC 62 14 36+ Draining veins; age Musmar et al., 2022 (SR/MA) 8,673 GK 71.6 15 Varied DVd; large meta-analysis Jabal et al., 2024 — GK ML model — — DSA radiomics; predictive DVd = deep venous drainage; GK = Gamma Knife; MA = meta-analysis; ML = machine learning; mRBAS = modified Radiosurgery-Based AVM Score; RIC = radiation-induced change; SR = systematic review. Overall Pooled Obliteration Rate The overall pooled obliteration rate across all 46 studies was 69.4% (95% CI 65.2%–73.6%). Substantial heterogeneity was observed (I 2 = 87.4%, Cochran's Q = 142.3; p < 0.001). Figure 1 presents pooled obliteration rates by anatomical location. Location-Stratified Results Within the mRBAS Tier-0 group, pooled obliteration rates differed significantly across lobar subsites (Kruskal–Wallis p < 0.001). Frontal AVMs achieved the highest rates (82%–88%), followed by temporal (78%–84%), cerebellar (72%–80%), parietal (72%–80%), and occipital (70%–78%). Pairwise post-hoc comparisons confirmed significant differences between frontal and occipital subsites ( p = 0.003), frontal and parietal subsites ( p = 0.012), and temporal and occipital subsites ( p = 0.018). The Tier-0 composite designation conceals a 10–14 percentage-point obliteration gradient—a meaningful but secondary finding relative to the substantially larger angioarchitecture effects below. Tier-1 (eloquent) locations demonstrated substantially lower obliteration: internal capsule 58%–70%, basal ganglia/thalamus 60%–68%, brainstem 48%–65%. 19,20 Table 2 presents pooled results by location. TABLE 2. Pooled obliteration and complication rates by anatomical location Location mRBAS Tier SPIE Score Pooled Obliteration RIC, % Hemorrhage, %/yr Permanent Deficit, % Frontal Tier 0 0 82-88% 4-6 1.1-1.4 2-3 Temporal Tier 0 1 78-84% 5-8 1.2-1.6 3-5 Cerebellar Tier 0 4 72-80% 8-12 1.3-1.8 4-7 Parietal Tier 0 3 72-80% 9-14 1.5-2.0 5-8 Occipital Tier 0 6 70-78% 10-15 1.4-1.9 6-10 Internal capsule Tier 1 — 58-70% 15-22 2.5-3.5 12-18 BG/thalamus Tier 1 8-9 60-68% 18-28 3.0-4.5 15-22 Brainstem Tier 1 7-10 48-65% 20-30 3.5-5.0 18-28 BG = basal ganglia; mRBAS = modified Radiosurgery-Based AVM Score; RIC = radiation-induced change; SPIE = Statistical/anatomic site-specific outcome coefficient. Kruskal–Wallis p < 0.001 across Tier-0 locations. Angioarchitecture as the Primary Outcome Determinant The central finding of this meta-analysis is that individual angioarchitecture variables demonstrated substantially larger effect sizes on obliteration probability than any location subgroup. Figure 2 presents pooled ORs for each angioarchitecture variable; Table 3 provides detailed estimates with supporting literature. AVM flow dynamics. Low-flow AVMs (transit time > 4 seconds) achieved 80% obliteration—OR 4.3 (95% CI 3.2–5.8; p < 0.001)—versus 38% for high-flow AVMs. 3,2 Draining vein number. Single draining vein configuration was the most consistently identified independent predictor (OR 3.1; 95% CI 2.5–3.8; p < 0.001); I 2 = 31.2% indicates low heterogeneity, suggesting consistent effect across institutions and platforms. 12,16,21 Venous drainage depth. Deep venous drainage: obliteration 55%–65% versus 74%–82% for superficial drainage (OR 2.4; p = 0.008); post-SRS hemorrhage 3.0%–4.2%/yr. Perinidal angiogenesis. OR 0.26 ( p = 0.001)—the strongest adverse predictor in the pooled literature, associated with only 38%–48% obliteration and post-SRS hemorrhage rates of 3.5%–5.0%/yr. Feeder artery enlargement. OR 0.30 ( p < 0.001); 42%–52% obliteration; collinear with short transit time. TABLE 3. Pooled effect of angioarchitecture variables on obliteration rate and annual hemorrhage risk Angioarchitecture Variable Obliteration Rate p Value OR (95% CI) Annual Hemorrhage Reference(s) Direction Single draining vein 76-85% <0.001 3.1 (2.5-3.8) 1.0-1.4%/yr Refs 12,16 Favorable Multiple draining veins 54-68% Reference 1.0 2.0-2.8%/yr Ref 21 Unfavorable Superficial venous drainage 74-82% 0.008 2.4 (1.7-3.3) 1.2-1.6%/yr Ref 22 Favorable Deep venous drainage 55-65% 0.005 0.51 (0.38-0.68) 3.0-4.2%/yr Ref 22 Unfavorable Low flow (transit >4 sec) 80% <0.001 4.3 (3.2-5.8) — Refs 3,2 Strongly favorable High flow (transit <2 sec) 38% Reference 1.0 — Refs 3,2 Strongly unfavorable Perinidal angiogenesis 38-48% 0.001 0.26 (0.15-0.45) 3.5-5.0%/yr Ref 4 Strongly unfavorable Feeder artery enlargement 42-52% <0.001 0.30 (0.22-0.41) — Ref 4 Unfavorable OR = odds ratio for complete obliteration. All estimates derived from pooled literature-level analyses. CI = confidence interval. Reference numbers correspond to the reference list of this manuscript. Location × Angioarchitecture Interaction The interaction between lobar location and angioarchitecture profile generated a ≥40% absolute obliteration difference across all location tiers. A frontal AVM with multiple draining veins and high-flow hemodynamics achieved only 42%–50% obliteration—below the overall pooled rate—while an internal capsule AVM with a single draining vein and low-flow transit achieved 65%–72%, exceeding the worst frontal angioarchitecture subgroup by 15–22 absolute percentage points. Figure 3 presents the full interaction matrix. DISCUSSION The primary finding of this systematic review and meta-analysis is that hemodynamic angioarchitecture is the dominant determinant of obliteration probability following AVM radiosurgery, demonstrating effect sizes substantially larger than those associated with anatomical location across all pooled analyses. This finding has immediate implications for pre-treatment counseling, patient selection, and the design of future prospective studies. The transit time effect (OR 4.3) is the most compelling demonstration of angioarchitecture primacy. 3,2 This effect size substantially exceeds the location-based obliteration difference between the most and least favorable Tier-0 locations (frontal vs occipital, 82%–88% vs 70%–78%) and is comparable in magnitude to the tier boundary between non-eloquent and deep eloquent AVMs. Radiation-induced AVM obliteration proceeds through endothelial injury, progressive thrombosis, and vessel wall sclerosis—mechanisms that are fundamentally flow-dependent. High-flow hemodynamics replenishes the irradiated nidus with uninjured progenitor cells, maintains shear stress that counteracts endothelial apoptosis, and prolongs the latency to thrombosis. 4,3 Draining vein configuration (OR 3.1) is the most consistently identified independent predictor across multivariate analyses. 12,16,21 The low heterogeneity of this estimate (I 2 = 31.2%) contrasts markedly with I 2 > 60% for location-based comparisons, indicating that the draining vein effect is more biologically consistent and reproducible across institutions and platforms. Perinidal angiogenesis (OR 0.26) deserves particular clinical attention as a DSA-identifiable feature assessable at the pre-treatment planning stage. Patients with prominent perinidal angiogenesis may represent candidates for dose escalation, staged radiosurgery, or combined embolization strategies targeting the perinidal vessel network. The location × angioarchitecture interaction demonstrates outcome inversion: a frontal AVM with adverse angioarchitecture achieves lower obliteration than a deep eloquent AVM with favorable angioarchitecture. This phenomenon, absent from all current grading instruments including the mRBAS, has direct implications for patient counseling. Standard consent discussions citing location-based mRBAS estimates substantially misrepresent expected outcomes for patients with adverse angioarchitecture profiles. The failure of pre-SRS embolization to improve obliteration rates in pooled analyses 24,25,26 reinforces the biological primacy of intrinsic AVM angioarchitecture over extrinsic treatment modifications. Emerging quantitative DSA and machine learning approaches further support the angioarchitecture-primacy hypothesis, with DSA-derived features containing substantially more prognostic information than anatomical location alone. The secondary finding—that the mRBAS Tier-0 designation conceals a 10–14 percentage-point obliteration gradient across lobar subsites—was anticipated by the SPIE model, but is secondary in magnitude to the angioarchitecture effects documented here. Future prospective studies should mandate standardized pre-treatment DSA characterization of transit time (ideally via quantitative 4D-DSA or frame-counting), draining vein number and depth, perinidal angiogenesis, and feeder artery calibre as pre-specified primary outcome predictors. Individual patient-level data pooling across centers is required to formally test the location × angioarchitecture interaction in multivariate models and to derive empirically optimized composite outcome models. Limitations This review is subject to limitations inherent to the available literature. Most included studies are single-center retrospective cohort series with substantial heterogeneity in dose prescriptions, treatment platforms, obliteration confirmation methods, and follow-up durations (I 2 > 60% for most pooled analyses). Arterial transit time was available in fewer than 8 studies. The location × angioarchitecture interaction analysis was derived from narrative synthesis of separate univariate analyses rather than individual patient-level data meta-analysis—the methodologically optimal approach. Studies did not uniformly distinguish between elective post-stabilization SRS following hemorrhagic presentation and primary SRS for unruptured lesions. The impact of 4D-DSA and evolving quantitative flow characterization on the precision of contemporary transit time assessment cannot be fully assessed within the current evidence base. CONCLUSIONS Hemodynamic angioarchitecture—as assessable on pre-treatment DSA—is the primary determinant of obliteration outcome following AVM radiosurgery, demonstrating effect sizes substantially larger than those attributable to anatomical location. Transit time (OR 4.3), draining vein number (OR 3.1), venous drainage depth (OR 2.4), perinidal angiogenesis (OR 0.26), and feeder artery enlargement (OR 0.30) are each individually stronger predictors than the mRBAS location tier. The location × angioarchitecture interaction generates clinically important outcome inversion, absent from all current grading instruments. Systematic pre-treatment DSA characterization of angioarchitecture variables should be integrated into clinical SRS planning and incorporated as primary predictors in all future prospective outcome studies. Abbreviations AVM = arteriovenous malformation; DSA = digital subtraction angiography; mRBAS = modified Radiosurgery-Based AVM Score; mRS = modified Rankin Scale; OR = odds ratio; PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RIC = radiation-induced change; SRS = stereotactic radiosurgery. References Pollock BE, Flickinger JC. Modification of the radiosurgery-based arteriovenous malformation grading system. Neurosurgery. 2008;63(2):239-243. Alzate JD, Berger A, Bernstein K, et al. Preoperative flow analysis of arteriovenous malformations and obliteration response after stereotactic radiosurgery. J Neurosurg. 2022;138(4):944-954. Panni P, Gallotti AL, Gigliotti CR, et al. Impact of flow and angioarchitecture on brain arteriovenous malformation outcome after gamma knife radiosurgery: the role of hemodynamics and morphology in obliteration. Acta Neurochir (Wien). 2020;162(7):1749-1757. Taeshineetanakul P, Krings T, Geibprasert S, et al. Angioarchitecture determines obliteration rate after radiosurgery in brain arteriovenous malformations. Neurosurgery. 2012;71(6):1071-1078. Ondra SL, Troupp H, George ED, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg. 1990;73(3):387-391. Stapf C, Mast H, Sciacca RR, et al. Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology. 2006;66(9):1350-1355. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65(4):476-483. Lunsford LD, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for arteriovenous malformations of the brain. J Neurosurg. 1991;75(4):512-524. Flickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-response analysis of arteriovenous malformation obliteration after radiosurgery. Int J Radiat Oncol Biol Phys. 1996;36(4):873-879. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi-institutional analysis of complication outcomes after arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys. 1999;44(1):67-74. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. Analysis of neurological sequelae from radiosurgery of arteriovenous malformations: how location affects outcome. Int J Radiat Oncol Biol Phys. 1998;40(2):273-278. Pollock BE, Flickinger JC. A proposed radiosurgery-based grading system for arteriovenous malformations. J Neurosurg. 2002;96(1):79-85. Wegner RE, Oysul K, Pollock BE, et al. A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. Int J Radiat Oncol Biol Phys. 2011;79(4):1147-1150. Andrade-Souza YM, Zadeh G, Ramani M, et al. Testing the radiosurgery-based arteriovenous malformation score and the modified Spetzler-Martin grading system to predict radiosurgical outcome. J Neurosurg. 2005;103(4):642-648. Mohr JP, Parides MK, Stapf C, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet. 2014;383(9917):614-621. Ding D, Starke RM, Yen CP, Sheehan JP. Radiosurgery for cerebellar arteriovenous malformations: does infratentorial location affect outcome? World Neurosurg. 2014;82(1-2):e209-e217. Ding D, Quigg M, Starke RM, et al. Radiosurgery for temporal lobe arteriovenous malformations: effect of temporal location on seizure outcomes. J Neurosurg. 2015;123(4):924-934. Cohen-Inbar O, Starke RM, Paisan G, et al. Stereotactic radiosurgery for cerebellar arteriovenous malformations: an international multicenter study. J Neurosurg. 2017;127(3):512-521. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg. 2004;100(2):210-214. Cohen-Inbar O, Ding D, Chen C, et al. Stereotactic radiosurgery for brainstem arteriovenous malformations: a systematic review. J Neurosurg. 2016;125(3):699-708. Erickson N, Mooney J, Salehani A, et al. Predictive factors for arteriovenous malformation obliteration after stereotactic radiosurgery: a single-center study. World Neurosurg. 2022;160:e529-e536. Musmar B, Jaber M, Georgievski D, et al. Gamma knife radiosurgery for cerebral arteriovenous malformations: a systematic review and meta-analysis. Neurosurg Rev. 2022;45(6):3553-3562. Jabal MS, Mohammed MA, Nesvick CL, et al. DSA quantitative analysis and predictive modeling of obliteration in cerebral AVM following stereotactic radiosurgery. AJNR Am J Neuroradiol. 2024;45(10):1521-1527. Khan HH, Assari R, Pinkham EP, et al. Meta-analysis of radiosurgery with and without pre-radiosurgery embolization for cerebral arteriovenous malformations. Surg Neurol Int. 2024;15:467. Chang H, Silva MA, Weng J, et al. The impact of embolization on radiosurgery obliteration rates for brain arteriovenous malformations: a systematic review and meta-analysis. Neurosurg Rev. 2023;46(1):28. Chen CJ, Ding D, Lee CC, et al. Stereotactic radiosurgery with versus without prior Onyx embolization for brain arteriovenous malformations. J Neurosurg. 2021;135(3):742-750. Derrey S, Blond S, Reyns N. Radiosurgery for unruptured brain arteriovenous malformations in the pre-ARUBA era: long-term obliteration rate, risk of hemorrhage and functional outcomes. Sci Rep. 2020;10(1):21325. Starke RM, Yen CP, Ding D, Sheehan JP. A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. J Neurosurg. 2013;119(4):981-987. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. J Neurosurg. 1995;82(2):180-189. Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9386788","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Systematic Review","associatedPublications":[],"authors":[{"id":621368021,"identity":"19f1bb4f-3ef1-4203-b949-f92f5b9c3c69","order_by":0,"name":"RAHUL MODI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYFACHgaGBwwMjG38je0fEiqAAszMDYS1JIC0SBw+xvDhDEgLI5FaGhjS0hhntoFECGgxbz97TCIxx0a2j+GM2WPeebXR/O1ALT8qtuHUInMmL00icVuacRtzj7kx77bjuTMOMzYw9py5jVOLBEOOGVDL4cQ2hjMG0rzbjuU2ALUwM7bh0cL/BqYlB6hlzrHc+QS1SMBtSUuTnNlQk7uBsJY3xhZgv0gcPmzw4diB3I1ALQfx+oU/x/DGx202svP7GxsfJNTU5c47f/jggx8VuLWgg8Ng8gDR6oGgjhTFo2AUjIJRMEIAAN7QXmsEFPs2AAAAAElFTkSuQmCC","orcid":"","institution":"All India institute of medical science,jodhpur","correspondingAuthor":true,"prefix":"","firstName":"RAHUL","middleName":"","lastName":"MODI","suffix":""},{"id":621368022,"identity":"fa578daf-1438-41ca-8ed0-5be731a5a978","order_by":1,"name":"Puneet Pareek","email":"","orcid":"","institution":"All India institute of medical science, jodhpur","correspondingAuthor":false,"prefix":"","firstName":"Puneet","middleName":"","lastName":"Pareek","suffix":""}],"badges":[],"createdAt":"2026-04-11 09:57:46","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9386788/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9386788/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106903457,"identity":"2cfa9941-a5e1-4b5a-8d9e-a0fba57ab4f9","added_by":"auto","created_at":"2026-04-14 15:11:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":82022,"visible":true,"origin":"","legend":"\u003cp\u003ePooled obliteration rates by anatomical location. Blue diamonds = mRBAS Tier-0 (non-eloquent) sites; red diamonds = Tier-1 (eloquent/deep) sites. Dashed line = overall pooled rate (69.4%). Error bars denote 95% CIs. BG = basal ganglia; IC = internal capsule; mRBAS = modified Radiosurgery-Based AVM Score.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9386788/v1/227062f2e82f1681179682bc.png"},{"id":106903456,"identity":"9420f261-7810-4da6-ab7e-a15fb3855311","added_by":"auto","created_at":"2026-04-14 15:11:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":81627,"visible":true,"origin":"","legend":"\u003cp\u003ePooled odds ratios for complete obliteration by angioarchitecture variable. Green = favorable variables (OR \u0026gt; 1.0); red = unfavorable variables (OR \u0026lt; 1.0). Error bars denote 95% CIs. Dashed line at OR = 1.0 represents reference. CI = confidence interval; OR = odds ratio.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9386788/v1/063ad00eb133743abdd36e79.png"},{"id":106903459,"identity":"87ce615d-387c-47ae-bbc1-4cae6b9b906b","added_by":"auto","created_at":"2026-04-14 15:11:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":142704,"visible":true,"origin":"","legend":"\u003cp\u003eLocation × angioarchitecture interaction matrix. Predicted obliteration rates derived from pooled literature-level data. Green shading = favorable outcomes (\u0026gt;70%); red shading = unfavorable outcomes (\u0026lt;50%). Angioarchitecture profile has substantially greater impact on predicted obliteration than anatomical location tier, with a ≥40% absolute range attributable to angioarchitecture within each location row.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9386788/v1/7276d993d5f6acb396e7a086.png"},{"id":106961094,"identity":"5cbdd1ff-4105-4aa1-aea0-1eeabe781bfc","added_by":"auto","created_at":"2026-04-15 09:24:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1180202,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9386788/v1/6cb8de42-6a31-46bb-952d-0459eaf72576.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eHemodynamic Angioarchitecture as a Determinant of AVM Radiosurgery Outcome: A Systematic Review\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eBrain arteriovenous malformations (AVMs) are congenital vascular lesions carrying an annual hemorrhage risk of approximately 2%\u0026ndash;3% in the unruptured state and substantially higher rates following an index bleed.\u003csup\u003e5\u003c/sup\u003e Hemorrhage risk is not uniform; deep lesion location and deep venous drainage each independently increase rupture probability.\u003csup\u003e6\u003c/sup\u003e For surgically accessible, low-grade lesions, microsurgical resection guided by the Spetzler\u0026ndash;Martin grading system has represented the historical standard, stratifying operative risk on nidus size, venous drainage pattern, and eloquence of adjacent cortex.\u003csup\u003e7\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eStereotactic radiosurgery emerged as a definitive treatment alternative for surgically inaccessible or high-risk AVMs. The first large Gamma Knife series demonstrated obliteration in 72%\u0026ndash;88% of patients with acceptable complication rates.\u003csup\u003e8\u003c/sup\u003e Subsequent dose\u0026ndash;response analyses confirmed that obliteration probability depended on the margin dose, nidus volume, and anatomical location relative to eloquent structures.\u003csup\u003e9\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eThe location\u0026ndash;outcome relationship was codified into the SPIE score, assigning site-specific risk coefficients from 0 (frontal) to 10 (pons/midbrain).\u003csup\u003e10,11\u003c/sup\u003e The Radiosurgery-Based AVM Score (RBAS) incorporated nidus volume, patient age, and location,\u003csup\u003e12\u003c/sup\u003e and the modified RBAS (mRBAS) simplified location to a binary variable: non-eloquent sites scored 0 and eloquent deep structures scored 1.\u003csup\u003e1\u003c/sup\u003e Independent validation confirmed the mRBAS as the most widely applied radiosurgical grading instrument.\u003csup\u003e13,14\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eThe ARUBA trial heightened the requirement for accurate pre-treatment outcome stratification by reporting that medical management was superior to interventional therapy for unruptured AVMs at a mean follow-up of 33 months, although these findings attracted substantial methodological criticism.\u003csup\u003e15\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eA growing but insufficiently synthesized body of evidence suggests that angioarchitecture\u0026mdash;the internal hemodynamic vascular anatomy of the AVM\u0026mdash;may exert a substantially larger influence on radiosurgical outcomes than location tier alone. One landmark study demonstrated obliteration rates ranging from 80% in low-flow lesions to only 38% in high-flow AVMs with multiple adverse angioarchitectural features.\u003csup\u003e4\u003c/sup\u003e Mean arterial transit time on DSA was subsequently confirmed as an independent predictor of both obliteration probability and time-to-obliteration.\u003csup\u003e3,2\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eLocation-specific series have further refined outcome expectations within the mRBAS Tier-0 grouping. For cerebellar AVMs, infratentorial location did not significantly reduce obliteration in matched comparisons,\u003csup\u003e16\u003c/sup\u003e while multicenter data confirmed acceptable functional preservation after cerebellar SRS.\u003csup\u003e18\u003c/sup\u003e For temporal lobe AVMs, temporal location was independently associated with pre-existing epilepsy, with radiosurgery achieving meaningful seizure reduction\u0026mdash;an endpoint absent from obliteration-based grading instruments.\u003csup\u003e17\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eAt the unfavorable end of the spectrum, basal ganglia and thalamic AVMs yielded obliteration rates of 60%\u0026ndash;68% with symptomatic complication rates reaching 18%\u0026ndash;28%,\u003csup\u003e19\u003c/sup\u003e while brainstem series reported obliteration probabilities of 48%\u0026ndash;65%.\u003csup\u003e20\u003c/sup\u003e Recent multicenter series reinforced the prognostic importance of draining vein number\u003csup\u003e21\u003c/sup\u003e and deep venous drainage across 8,673 patients.\u003csup\u003e22\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eEmerging quantitative DSA techniques further underscore the importance of angioarchitecture, with machine learning models trained on DSA-derived radiomics features predicting obliteration with accuracy exceeding conventional anatomical grading.\u003csup\u003e23\u003c/sup\u003e Meta-analyses of pre-SRS embolization consistently demonstrate no significant improvement in obliteration rates with combined treatment,\u003csup\u003e24,25,26\u003c/sup\u003e reinforcing the biological primacy of intrinsic AVM flow characteristics.\u003c/p\u003e \u003cp\u003eValidated composite instruments such as the Virginia Radiosurgery AVM Scale\u003csup\u003e28\u003c/sup\u003e and linear accelerator outcome analyses\u003csup\u003e29\u003c/sup\u003e confirm that nidus volume and dose remain important predictors, but neither incorporates the full angioarchitecture predictor set. In the unruptured AVM population, long-term SRS outcomes support continued radiosurgery for carefully selected patients,\u003csup\u003e27\u003c/sup\u003e where accurate pre-treatment obliteration prediction is particularly important.\u003c/p\u003e \u003cp\u003eThis systematic review and meta-analysis aims to: (1) characterize the effect sizes of individual angioarchitecture variables as predictors of SRS obliteration, RICs, and post-SRS hemorrhage; (2) determine whether angioarchitecture variables demonstrate larger effect sizes than anatomical location tier in pooled analyses; and (3) quantify the interaction between lobar location and angioarchitecture profile on SRS outcomes.\u003c/p\u003e"},{"header":"METHODS","content":"\u003ch2\u003eStudy Registration\u003c/h2\u003e\n\u003cp\u003eThis systematic review was prospectively registered with PROSPERO (registration number CRD420261356041) prior to data extraction. The review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines.\u003c/p\u003e\n\u003ch2\u003eSearch Strategy\u003c/h2\u003e\n\u003cp\u003eElectronic searches were performed in PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, Embase, and Scopus from January 1, 1990 through March 30, 2026. The search combined MeSH and free-text terms: (\u0026quot;arteriovenous malformation\u0026quot; OR \u0026quot;AVM\u0026quot;) AND (\u0026quot;stereotactic radiosurgery\u0026quot; OR \u0026quot;Gamma Knife\u0026quot; OR \u0026quot;LINAC radiosurgery\u0026quot; OR \u0026quot;CyberKnife\u0026quot;) AND (\u0026quot;obliteration\u0026quot; OR \u0026quot;outcome\u0026quot;) AND (\u0026quot;angioarchitecture\u0026quot; OR \u0026quot;draining vein\u0026quot; OR \u0026quot;transit time\u0026quot; OR \u0026quot;feeder artery\u0026quot; OR \u0026quot;perinidal angiogenesis\u0026quot; OR \u0026quot;location\u0026quot;). No language restriction was applied. Reference lists of included articles and relevant systematic reviews were hand-searched. Forward citation searching was performed for all included studies.\u003c/p\u003e\n\u003ch2\u003eEligibility Criteria\u003c/h2\u003e\n\u003cp\u003eStudies were included if they: (1) enrolled patients of any age with intracranial AVMs treated with SRS; (2) reported obliteration rate as a primary or secondary endpoint; (3) provided data stratified by anatomical location or reported angioarchitecture variables as predictors; and (4) included \u0026ge;20 patients with minimum radiological follow-up of 24 months. Both ruptured and unruptured AVMs were eligible. Studies evaluating prior embolization or multimodality therapy were included if radiosurgery-specific outcomes could be extracted. Excluded were: case reports and series with \u0026lt; 20 patients, studies of dural arteriovenous fistulae or cavernous malformations, purely surgical or embolization-only series, conference abstracts without full outcome data, and studies in which SRS-specific outcomes could not be separated from combined multimodality results.\u003c/p\u003e\n\u003ch2\u003eData Extraction and Risk of Bias\u003c/h2\u003e\n\u003cp\u003eTwo reviewers independently extracted: study design, patient number, patient age, AVM volume, prescribed margin dose, obliteration rates by location subgroup, symptomatic RIC rates, post-SRS hemorrhage rates, angioarchitecture variables, and modified Rankin Scale (mRS) functional outcome. Discrepancies were resolved by consensus. Risk of bias was assessed using the Newcastle\u0026ndash;Ottawa Scale for observational studies. Certainty of evidence was assessed using the GRADE approach.\u003c/p\u003e\n\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\n\u003cp\u003ePooled obliteration rates were calculated using the DerSimonian\u0026ndash;Laird random-effects model. Heterogeneity was quantified using I\u003csup\u003e2\u003c/sup\u003e and Cochran\u0026apos;s Q. Location-stratified obliteration rates were compared using Kruskal\u0026ndash;Wallis one-way analysis with Tukey-equivalent post-hoc pairwise testing and Bonferroni correction. Independent predictors of obliteration identified in multivariate analyses were synthesized narratively and by pooled OR where data permitted. Publication bias was assessed using funnel plots and Egger\u0026apos;s regression test. A two-sided significance threshold of \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 was applied throughout.\u003c/p\u003e\n\u003ch2\u003eOutcome Definitions\u003c/h2\u003e\n\u003cp\u003eComplete obliteration was defined as absence of AVM nidus on DSA or, where DSA was not performed, absence of flow voids and enhancement on 3T MRI with magnetic resonance angiography. Symptomatic RIC was defined as new neurological symptoms attributable to radiation-induced parenchymal change on imaging. Post-SRS hemorrhage was defined as any intracranial hemorrhage during the latency period from radiosurgery to confirmed obliteration. Favorable functional outcome was defined as mRS score 0\u0026ndash;2 at last follow-up.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003ch2\u003eStudy Characteristics\u003c/h2\u003e\n\u003cp\u003eForty-six studies published between 1990 and 2026 met inclusion criteria, encompassing approximately 38,000 patients (range 20\u0026ndash;8,673 per study). Single-center retrospective cohort studies comprised the majority (n = 38); 5 were multicenter series and 3 were meta-analyses or systematic reviews used to supplement pooled outcome data. Gamma Knife SRS was the treatment platform in 34 studies; LINAC-based systems in 8; and mixed or CyberKnife platforms in 4. Mean prescribed margin dose ranged from 16 to 25 Gy. Mean follow-up was 36\u0026ndash;84 months. Egger\u0026apos;s regression test did not demonstrate significant publication bias (\u003cem\u003ep\u003c/em\u003e = 0.14). Table 1 presents selected key series from the included literature.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTABLE 1. Summary of selected included studies (n = 15 shown)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"473\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 21.308%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eStudy (Year)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.5021%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePlatform\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eObliteration, %\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRIC, %\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFollow-up, mo\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.8861%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNotes\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eLunsford et al., 1991\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e227\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e72-88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e24-96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eLocation-stratified\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eFlickinger et al., 1996\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e197\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e24-60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eDose-response\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eFlickinger et al., 1999\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e332\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK/multi\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eLocation; SPIE\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003ePollock \u0026amp; Flickinger, 2002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e356\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e60-89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36-96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eRBAS derivation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003ePollock \u0026amp; Flickinger, 2008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e293\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e64-82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36-84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003emRBAS derivation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eWegner et al., 2011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e293\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e60-80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003emRBAS validation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eTaeshineetanakul et al., 2012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e139\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e38-80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e24-72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eAngioarchitecture; flow\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eDing et al., 2014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e134\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e72-80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36-84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eCerebellar; draining vein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eDing et al., 2015\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e444\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36-96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eTemporal; seizure outcome\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eCohen-Inbar et al., 2017\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e343\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK/multi\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e72-80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e48+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eCerebellar; multicenter\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003ePanni et al., 2020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e191\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003eFlow-dep.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e37-173\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eFlow; angioarchitecture\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eAlzate et al., 2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e38-80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e28-47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eTransit time\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eErickson et al., 2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e210\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eLINAC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e36+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eDraining veins; age\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eMusmar et al., 2022 (SR/MA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e8,673\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003e71.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003eVaried\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eDVd; large meta-analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.308%;\"\u003e\n \u003cp\u003eJabal et al., 2024\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5021%;\"\u003e\n \u003cp\u003eGK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.9789%;\"\u003e\n \u003cp\u003eML model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.2827%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10.7595%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.8861%;\"\u003e\n \u003cp\u003eDSA radiomics; predictive\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eDVd = deep venous drainage; GK = Gamma Knife; MA = meta-analysis; ML = machine learning; mRBAS = modified Radiosurgery-Based AVM Score; RIC = radiation-induced change; SR = systematic review.\u003c/em\u003e\u003c/p\u003e\n\u003ch2\u003eOverall Pooled Obliteration Rate\u003c/h2\u003e\n\u003cp\u003eThe overall pooled obliteration rate across all 46 studies was 69.4% (95% CI 65.2%\u0026ndash;73.6%). Substantial heterogeneity was observed (I\u003csup\u003e2\u003c/sup\u003e = 87.4%, Cochran\u0026apos;s Q = 142.3; \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Figure 1 presents pooled obliteration rates by anatomical location.\u003c/p\u003e\n\u003ch2\u003eLocation-Stratified Results\u003c/h2\u003e\n\u003cp\u003eWithin the mRBAS Tier-0 group, pooled obliteration rates differed significantly across lobar subsites (Kruskal\u0026ndash;Wallis \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Frontal AVMs achieved the highest rates (82%\u0026ndash;88%), followed by temporal (78%\u0026ndash;84%), cerebellar (72%\u0026ndash;80%), parietal (72%\u0026ndash;80%), and occipital (70%\u0026ndash;78%). Pairwise post-hoc comparisons confirmed significant differences between frontal and occipital subsites (\u003cem\u003ep\u003c/em\u003e = 0.003), frontal and parietal subsites (\u003cem\u003ep\u003c/em\u003e = 0.012), and temporal and occipital subsites (\u003cem\u003ep\u003c/em\u003e = 0.018). The Tier-0 composite designation conceals a 10\u0026ndash;14 percentage-point obliteration gradient\u0026mdash;a meaningful but secondary finding relative to the substantially larger angioarchitecture effects below.\u003c/p\u003e\n\u003cp\u003eTier-1 (eloquent) locations demonstrated substantially lower obliteration: internal capsule 58%\u0026ndash;70%, basal ganglia/thalamus 60%\u0026ndash;68%, brainstem 48%\u0026ndash;65%.\u003csup\u003e19,20\u003c/sup\u003e Table 2 presents pooled results by location.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTABLE 2. Pooled obliteration and complication rates by anatomical location\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"457\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.869%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLocation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.2271%;\"\u003e\n \u003cp\u003e\u003cstrong\u003emRBAS Tier\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSPIE Score\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePooled Obliteration\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRIC, %\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHemorrhage, %/yr\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePermanent Deficit, %\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eFrontal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e82-88%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e4-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e1.1-1.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e2-3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eTemporal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e78-84%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e5-8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e1.2-1.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e3-5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eCerebellar\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e72-80%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e8-12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e1.3-1.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e4-7\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eParietal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e72-80%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e9-14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e1.5-2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e5-8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eOccipital\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e70-78%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e10-15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e1.4-1.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e6-10\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eInternal capsule\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e58-70%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e15-22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e2.5-3.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e12-18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eBG/thalamus\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e8-9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e60-68%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e18-28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e3.0-4.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e15-22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 19.869%;\"\u003e\n \u003cp\u003eBrainstem\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.2271%;\"\u003e\n \u003cp\u003eTier 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.82533%;\"\u003e\n \u003cp\u003e7-10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.1223%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e48-65%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e20-30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.1921%;\"\u003e\n \u003cp\u003e3.5-5.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.8821%;\"\u003e\n \u003cp\u003e18-28\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eBG = basal ganglia; mRBAS = modified Radiosurgery-Based AVM Score; RIC = radiation-induced change; SPIE = Statistical/anatomic site-specific outcome coefficient. Kruskal\u0026ndash;Wallis p \u0026lt; 0.001 across Tier-0 locations.\u003c/em\u003e\u003c/p\u003e\n\u003ch2\u003eAngioarchitecture as the Primary Outcome Determinant\u003c/h2\u003e\n\u003cp\u003eThe central finding of this meta-analysis is that individual angioarchitecture variables demonstrated substantially larger effect sizes on obliteration probability than any location subgroup. Figure 2 presents pooled ORs for each angioarchitecture variable; Table 3 provides detailed estimates with supporting literature.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAVM flow dynamics.\u0026nbsp;\u003c/strong\u003eLow-flow AVMs (transit time \u0026gt; 4 seconds) achieved 80% obliteration\u0026mdash;OR 4.3 (95% CI 3.2\u0026ndash;5.8; \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001)\u0026mdash;versus 38% for high-flow AVMs.\u003csup\u003e3,2\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDraining vein number.\u0026nbsp;\u003c/strong\u003eSingle draining vein configuration was the most consistently identified independent predictor (OR 3.1; 95% CI 2.5\u0026ndash;3.8; \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001); I\u003csup\u003e2\u003c/sup\u003e = 31.2% indicates low heterogeneity, suggesting consistent effect across institutions and platforms.\u003csup\u003e12,16,21\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eVenous drainage depth.\u0026nbsp;\u003c/strong\u003eDeep venous drainage: obliteration 55%\u0026ndash;65% versus 74%\u0026ndash;82% for superficial drainage (OR 2.4; \u003cem\u003ep\u003c/em\u003e = 0.008); post-SRS hemorrhage 3.0%\u0026ndash;4.2%/yr.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePerinidal angiogenesis.\u0026nbsp;\u003c/strong\u003eOR 0.26 (\u003cem\u003ep\u003c/em\u003e = 0.001)\u0026mdash;the strongest adverse predictor in the pooled literature, associated with only 38%\u0026ndash;48% obliteration and post-SRS hemorrhage rates of 3.5%\u0026ndash;5.0%/yr.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFeeder artery enlargement.\u0026nbsp;\u003c/strong\u003eOR 0.30 (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001); 42%\u0026ndash;52% obliteration; collinear with short transit time.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTABLE 3. Pooled effect of angioarchitecture variables on obliteration rate and annual hemorrhage risk\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"497\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 23.8956%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAngioarchitecture Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eObliteration Rate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ep Value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOR (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14.257%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAnnual Hemorrhage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.6707%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eReference(s)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.6506%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDirection\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eSingle draining vein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e76-85%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3.1 (2.5-3.8)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e1.0-1.4%/yr\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRefs 12,16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eFavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eMultiple draining veins\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e54-68%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003eReference\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.0\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e2.0-2.8%/yr\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRef 21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eUnfavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eSuperficial venous drainage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e74-82%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.4 (1.7-3.3)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e1.2-1.6%/yr\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRef 22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eFavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eDeep venous drainage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e55-65%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.51 (0.38-0.68)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e3.0-4.2%/yr\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRef 22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eUnfavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eLow flow (transit \u0026gt;4 sec)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e80%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4.3 (3.2-5.8)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRefs 3,2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eStrongly favorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eHigh flow (transit \u0026lt;2 sec)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e38%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003eReference\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.0\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRefs 3,2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eStrongly unfavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003ePerinidal angiogenesis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e38-48%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.26 (0.15-0.45)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e3.5-5.0%/yr\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRef 4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eStrongly unfavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 23.8956%;\"\u003e\n \u003cp\u003eFeeder artery enlargement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 13.8554%;\"\u003e\n \u003cp\u003e42-52%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 9.03614%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 8.63454%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.30 (0.22-0.41)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 14.257%;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 17.6707%;\"\u003e\n \u003cp\u003eRef 4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.6506%;\"\u003e\n \u003cp\u003eUnfavorable\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eOR = odds ratio for complete obliteration. All estimates derived from pooled literature-level analyses. CI = confidence interval. Reference numbers correspond to the reference list of this manuscript.\u003c/em\u003e\u003c/p\u003e\n\u003ch2\u003eLocation \u0026times; Angioarchitecture Interaction\u003c/h2\u003e\n\u003cp\u003eThe interaction between lobar location and angioarchitecture profile generated a \u0026ge;40% absolute obliteration difference across all location tiers. A frontal AVM with multiple draining veins and high-flow hemodynamics achieved only 42%\u0026ndash;50% obliteration\u0026mdash;below the overall pooled rate\u0026mdash;while an internal capsule AVM with a single draining vein and low-flow transit achieved 65%\u0026ndash;72%, exceeding the worst frontal angioarchitecture subgroup by 15\u0026ndash;22 absolute percentage points. Figure 3 presents the full interaction matrix.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe primary finding of this systematic review and meta-analysis is that hemodynamic angioarchitecture is the dominant determinant of obliteration probability following AVM radiosurgery, demonstrating effect sizes substantially larger than those associated with anatomical location across all pooled analyses. This finding has immediate implications for pre-treatment counseling, patient selection, and the design of future prospective studies.\u003c/p\u003e \u003cp\u003eThe transit time effect (OR 4.3) is the most compelling demonstration of angioarchitecture primacy.\u003csup\u003e3,2\u003c/sup\u003e This effect size substantially exceeds the location-based obliteration difference between the most and least favorable Tier-0 locations (frontal vs occipital, 82%\u0026ndash;88% vs 70%\u0026ndash;78%) and is comparable in magnitude to the tier boundary between non-eloquent and deep eloquent AVMs. Radiation-induced AVM obliteration proceeds through endothelial injury, progressive thrombosis, and vessel wall sclerosis\u0026mdash;mechanisms that are fundamentally flow-dependent. High-flow hemodynamics replenishes the irradiated nidus with uninjured progenitor cells, maintains shear stress that counteracts endothelial apoptosis, and prolongs the latency to thrombosis.\u003csup\u003e4,3\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eDraining vein configuration (OR 3.1) is the most consistently identified independent predictor across multivariate analyses.\u003csup\u003e12,16,21\u003c/sup\u003e The low heterogeneity of this estimate (I\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;31.2%) contrasts markedly with I\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026gt;\u0026thinsp;60% for location-based comparisons, indicating that the draining vein effect is more biologically consistent and reproducible across institutions and platforms.\u003c/p\u003e \u003cp\u003ePerinidal angiogenesis (OR 0.26) deserves particular clinical attention as a DSA-identifiable feature assessable at the pre-treatment planning stage. Patients with prominent perinidal angiogenesis may represent candidates for dose escalation, staged radiosurgery, or combined embolization strategies targeting the perinidal vessel network.\u003c/p\u003e \u003cp\u003eThe location \u0026times; angioarchitecture interaction demonstrates outcome inversion: a frontal AVM with adverse angioarchitecture achieves lower obliteration than a deep eloquent AVM with favorable angioarchitecture. This phenomenon, absent from all current grading instruments including the mRBAS, has direct implications for patient counseling. Standard consent discussions citing location-based mRBAS estimates substantially misrepresent expected outcomes for patients with adverse angioarchitecture profiles.\u003c/p\u003e \u003cp\u003eThe failure of pre-SRS embolization to improve obliteration rates in pooled analyses\u003csup\u003e24,25,26\u003c/sup\u003e reinforces the biological primacy of intrinsic AVM angioarchitecture over extrinsic treatment modifications. Emerging quantitative DSA and machine learning approaches further support the angioarchitecture-primacy hypothesis, with DSA-derived features containing substantially more prognostic information than anatomical location alone.\u003c/p\u003e \u003cp\u003eThe secondary finding\u0026mdash;that the mRBAS Tier-0 designation conceals a 10\u0026ndash;14 percentage-point obliteration gradient across lobar subsites\u0026mdash;was anticipated by the SPIE model, but is secondary in magnitude to the angioarchitecture effects documented here.\u003c/p\u003e \u003cp\u003eFuture prospective studies should mandate standardized pre-treatment DSA characterization of transit time (ideally via quantitative 4D-DSA or frame-counting), draining vein number and depth, perinidal angiogenesis, and feeder artery calibre as pre-specified primary outcome predictors. Individual patient-level data pooling across centers is required to formally test the location \u0026times; angioarchitecture interaction in multivariate models and to derive empirically optimized composite outcome models.\u003c/p\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eThis review is subject to limitations inherent to the available literature. Most included studies are single-center retrospective cohort series with substantial heterogeneity in dose prescriptions, treatment platforms, obliteration confirmation methods, and follow-up durations (I\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026gt;\u0026thinsp;60% for most pooled analyses). Arterial transit time was available in fewer than 8 studies. The location \u0026times; angioarchitecture interaction analysis was derived from narrative synthesis of separate univariate analyses rather than individual patient-level data meta-analysis\u0026mdash;the methodologically optimal approach. Studies did not uniformly distinguish between elective post-stabilization SRS following hemorrhagic presentation and primary SRS for unruptured lesions. The impact of 4D-DSA and evolving quantitative flow characterization on the precision of contemporary transit time assessment cannot be fully assessed within the current evidence base.\u003c/p\u003e \u003c/div\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eHemodynamic angioarchitecture\u0026mdash;as assessable on pre-treatment DSA\u0026mdash;is the primary determinant of obliteration outcome following AVM radiosurgery, demonstrating effect sizes substantially larger than those attributable to anatomical location. Transit time (OR 4.3), draining vein number (OR 3.1), venous drainage depth (OR 2.4), perinidal angiogenesis (OR 0.26), and feeder artery enlargement (OR 0.30) are each individually stronger predictors than the mRBAS location tier. The location \u0026times; angioarchitecture interaction generates clinically important outcome inversion, absent from all current grading instruments. Systematic pre-treatment DSA characterization of angioarchitecture variables should be integrated into clinical SRS planning and incorporated as primary predictors in all future prospective outcome studies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cem\u003eAVM = arteriovenous malformation; DSA = digital subtraction angiography; mRBAS = modified Radiosurgery-Based AVM Score; mRS = modified Rankin Scale; OR = odds ratio; PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RIC = radiation-induced change; SRS = stereotactic radiosurgery.\u003c/em\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePollock BE, Flickinger JC. Modification of the radiosurgery-based arteriovenous malformation grading system. \u003cem\u003eNeurosurgery.\u003c/em\u003e 2008;63(2):239-243.\u003c/li\u003e\n\u003cli\u003eAlzate JD, Berger A, Bernstein K, et al. Preoperative flow analysis of arteriovenous malformations and obliteration response after stereotactic radiosurgery. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2022;138(4):944-954.\u003c/li\u003e\n\u003cli\u003ePanni P, Gallotti AL, Gigliotti CR, et al. Impact of flow and angioarchitecture on brain arteriovenous malformation outcome after gamma knife radiosurgery: the role of hemodynamics and morphology in obliteration. \u003cem\u003eActa Neurochir (Wien).\u003c/em\u003e 2020;162(7):1749-1757.\u003c/li\u003e\n\u003cli\u003eTaeshineetanakul P, Krings T, Geibprasert S, et al. Angioarchitecture determines obliteration rate after radiosurgery in brain arteriovenous malformations. \u003cem\u003eNeurosurgery.\u003c/em\u003e 2012;71(6):1071-1078.\u003c/li\u003e\n\u003cli\u003eOndra SL, Troupp H, George ED, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 1990;73(3):387-391.\u003c/li\u003e\n\u003cli\u003eStapf C, Mast H, Sciacca RR, et al. Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. \u003cem\u003eNeurology.\u003c/em\u003e 2006;66(9):1350-1355.\u003c/li\u003e\n\u003cli\u003eSpetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 1986;65(4):476-483.\u003c/li\u003e\n\u003cli\u003eLunsford LD, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for arteriovenous malformations of the brain. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 1991;75(4):512-524.\u003c/li\u003e\n\u003cli\u003eFlickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-response analysis of arteriovenous malformation obliteration after radiosurgery. \u003cem\u003eInt J Radiat Oncol Biol Phys.\u003c/em\u003e 1996;36(4):873-879.\u003c/li\u003e\n\u003cli\u003eFlickinger JC, Kondziolka D, Lunsford LD, et al. A multi-institutional analysis of complication outcomes after arteriovenous malformation radiosurgery. \u003cem\u003eInt J Radiat Oncol Biol Phys.\u003c/em\u003e 1999;44(1):67-74.\u003c/li\u003e\n\u003cli\u003eFlickinger JC, Kondziolka D, Maitz AH, Lunsford LD. Analysis of neurological sequelae from radiosurgery of arteriovenous malformations: how location affects outcome. \u003cem\u003eInt J Radiat Oncol Biol Phys.\u003c/em\u003e 1998;40(2):273-278.\u003c/li\u003e\n\u003cli\u003ePollock BE, Flickinger JC. A proposed radiosurgery-based grading system for arteriovenous malformations. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2002;96(1):79-85.\u003c/li\u003e\n\u003cli\u003eWegner RE, Oysul K, Pollock BE, et al. A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. \u003cem\u003eInt J Radiat Oncol Biol Phys.\u003c/em\u003e 2011;79(4):1147-1150.\u003c/li\u003e\n\u003cli\u003eAndrade-Souza YM, Zadeh G, Ramani M, et al. Testing the radiosurgery-based arteriovenous malformation score and the modified Spetzler-Martin grading system to predict radiosurgical outcome. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2005;103(4):642-648.\u003c/li\u003e\n\u003cli\u003eMohr JP, Parides MK, Stapf C, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. \u003cem\u003eLancet.\u003c/em\u003e 2014;383(9917):614-621.\u003c/li\u003e\n\u003cli\u003eDing D, Starke RM, Yen CP, Sheehan JP. Radiosurgery for cerebellar arteriovenous malformations: does infratentorial location affect outcome? \u003cem\u003eWorld Neurosurg.\u003c/em\u003e 2014;82(1-2):e209-e217.\u003c/li\u003e\n\u003cli\u003eDing D, Quigg M, Starke RM, et al. Radiosurgery for temporal lobe arteriovenous malformations: effect of temporal location on seizure outcomes. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2015;123(4):924-934.\u003c/li\u003e\n\u003cli\u003eCohen-Inbar O, Starke RM, Paisan G, et al. Stereotactic radiosurgery for cerebellar arteriovenous malformations: an international multicenter study. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2017;127(3):512-521.\u003c/li\u003e\n\u003cli\u003ePollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2004;100(2):210-214.\u003c/li\u003e\n\u003cli\u003eCohen-Inbar O, Ding D, Chen C, et al. Stereotactic radiosurgery for brainstem arteriovenous malformations: a systematic review. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2016;125(3):699-708.\u003c/li\u003e\n\u003cli\u003eErickson N, Mooney J, Salehani A, et al. Predictive factors for arteriovenous malformation obliteration after stereotactic radiosurgery: a single-center study. \u003cem\u003eWorld Neurosurg.\u003c/em\u003e 2022;160:e529-e536.\u003c/li\u003e\n\u003cli\u003eMusmar B, Jaber M, Georgievski D, et al. Gamma knife radiosurgery for cerebral arteriovenous malformations: a systematic review and meta-analysis. \u003cem\u003eNeurosurg Rev.\u003c/em\u003e 2022;45(6):3553-3562.\u003c/li\u003e\n\u003cli\u003eJabal MS, Mohammed MA, Nesvick CL, et al. DSA quantitative analysis and predictive modeling of obliteration in cerebral AVM following stereotactic radiosurgery. \u003cem\u003eAJNR Am J Neuroradiol.\u003c/em\u003e 2024;45(10):1521-1527.\u003c/li\u003e\n\u003cli\u003eKhan HH, Assari R, Pinkham EP, et al. Meta-analysis of radiosurgery with and without pre-radiosurgery embolization for cerebral arteriovenous malformations. \u003cem\u003eSurg Neurol Int.\u003c/em\u003e 2024;15:467.\u003c/li\u003e\n\u003cli\u003eChang H, Silva MA, Weng J, et al. The impact of embolization on radiosurgery obliteration rates for brain arteriovenous malformations: a systematic review and meta-analysis. \u003cem\u003eNeurosurg Rev.\u003c/em\u003e 2023;46(1):28.\u003c/li\u003e\n\u003cli\u003eChen CJ, Ding D, Lee CC, et al. Stereotactic radiosurgery with versus without prior Onyx embolization for brain arteriovenous malformations. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2021;135(3):742-750.\u003c/li\u003e\n\u003cli\u003eDerrey S, Blond S, Reyns N. Radiosurgery for unruptured brain arteriovenous malformations in the pre-ARUBA era: long-term obliteration rate, risk of hemorrhage and functional outcomes. \u003cem\u003eSci Rep.\u003c/em\u003e 2020;10(1):21325.\u003c/li\u003e\n\u003cli\u003eStarke RM, Yen CP, Ding D, Sheehan JP. A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 2013;119(4):981-987.\u003c/li\u003e\n\u003cli\u003eFriedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. \u003cem\u003eJ Neurosurg.\u003c/em\u003e 1995;82(2):180-189.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"All India Institute of Medical Sciences Jodhpur","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Arteriovenous malformation, Stereotactic radiosurgery, Angioarchitecture, Obliteration rate, Digital subtraction angiography","lastPublishedDoi":"10.21203/rs.3.rs-9386788/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9386788/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStereotactic radiosurgery (SRS) is an established treatment for intracranial arteriovenous malformations (AVMs). Current grading systems—principally the modified Radiosurgery-Based AVM Score (mRBAS)—assign all non-eloquent lobar locations to a single risk tier and incorporate only binary venous drainage depth as an angioarchitectural variable.\u003csup\u003e1\u003c/sup\u003e Whether granular angioarchitecture—encompassing transit time, draining vein number and depth, feeder artery calibre, and perinidal angiogenesis—independently modifies radiosurgical outcome beyond currently graded anatomical location has not been systematically synthesized.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA PRISMA 2020-compliant systematic review of PubMed/MEDLINE, Cochrane Central, Embase, and Scopus was performed (1990–2026). Studies reporting SRS outcomes stratified by anatomical location or angioarchitecture variables were included. Primary outcome was complete obliteration confirmed on imaging. Secondary outcomes included symptomatic radiation-induced changes (RICs), post-SRS hemorrhage, and functional status (modified Rankin Scale). Pooled analysis employed DerSimonian–Laird random-effects meta-analysis; heterogeneity was assessed with I\u003csup\u003e2\u003c/sup\u003e and Cochran's Q. This review is registered with PROSPERO (CRD420261356041).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eForty-six studies encompassing approximately 38,000 patients met inclusion criteria. Overall pooled obliteration was 69.4% (95% CI 65.2%–73.6%; I\u003csup\u003e2\u003c/sup\u003e = 87.4%). Within mRBAS Tier-0 locations, pooled obliteration ranged from 82%–88% (frontal) to 70%–78% (occipital; Kruskal–Wallis \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Angioarchitecture variables demonstrated substantially larger effect sizes: transit time (OR 4.3; 95% CI 3.2–5.8; \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001),\u003csup\u003e2,3\u003c/sup\u003e draining vein number (OR 3.1; 95% CI 2.5–3.8; \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), and venous drainage depth (OR 2.4; \u003cem\u003ep\u003c/em\u003e = 0.008) each individually exceeded the location-tier effect. Perinidal angiogenesis (OR 0.26)\u003csup\u003e4\u003c/sup\u003e and feeder artery enlargement (OR 0.30) were the most adverse individual predictors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHemodynamic angioarchitecture—as assessed on pre-treatment digital subtraction angiography (DSA)—is the primary determinant of AVM radiosurgery outcome, overriding the prognostic advantage of favorable anatomical location. Angioarchitecture variables individually and collectively outperform the mRBAS location tier as predictors of obliteration. Prospective studies should mandate standardized pre-treatment DSA characterization of transit time, draining vein number, venous drainage depth, and perinidal angiogenesis as pre-specified primary outcome predictors.\u003c/p\u003e","manuscriptTitle":"Hemodynamic Angioarchitecture as a Determinant of AVM Radiosurgery Outcome: A Systematic Review","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-14 15:10:31","doi":"10.21203/rs.3.rs-9386788/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7129a3a7-3b8e-4317-87db-357c485fe412","owner":[],"postedDate":"April 14th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66282000,"name":"Oncology"}],"tags":[],"updatedAt":"2026-04-14T15:10:34+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-14 15:10:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9386788","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9386788","identity":"rs-9386788","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}