In Reply: Stereotactic Radiosurgery for Dural Arteriovenous Fistulas: A Systematic Review and Meta-Analysis and International Stereotactic Radiosurgery Society Practice Guidelines.

Abstract

To the Editor: We appreciate the insights regarding our systematic review and meta-analysis and practice guidelines from the International Stereotactic Radiosurgery Society (ISRS) on stereotactic radiosurgery (SRS) for dural arteriovenous fistulas (dAVFs) in a recent Letter to the Editor by Raper et al.1,2 As noted, our analysis found that SRS was estimated to result in complete obliteration (CO) rates of approximately 70% with very favorable symptom improvement and symptom cure rates of roughly 97% and 79%, respectively. CO rates after SRS compare quite favorably with those with Onyx embolization only per the Consortium for Dural Arteriovenous Fistula Outcomes Research (CONDOR) of approximately 55%.3 Of note, multimodality therapy with prior embolization was also noted to be associated with improved symptom cure rates in our analysis, suggesting a potential benefit in the therapeutic ratio for patients with poor prognostic factors for achievement of CO after either SRS or embolization alone. Treatment-related morbidity was also found to be quite limited with incidences of either permanent neurological deficits or hemorrhage after SRS both estimated to be roughly 1%. In our guidelines analysis, we were unable to report on favorable outcomes (as was performed in CONDOR and defined as achievement of CO without hemorrhage) because of lack of patient-level data. As noted, the risk of permanent neurological deficits after SRS compares favorably with embolization with estimates ranging from 4% to 16.1%.3,4 Previous analyses have noted similar rates of hemorrhage after embolization of 1.6%, quite similar to those observed in our analysis after SRS.3 We also noted in our analysis that previous hemorrhage was associated with a higher risk of hemorrhage after SRS similar to findings from CONDOR noting patients with prior hemorrhage had lower rates of favorable outcomes. The favorable risk profile and outcomes of SRS highlighted by our meta-analysis suggest it is an excellent therapeutic option for management of appropriately selected patients with dAVFs. Of note, limitations regarding both our analysis and CONDOR are inclusion of data from experienced and high volume sites that may not be reflective of sites with less experience in management of dAVFs. In the absence of prospective randomized data in part secondary to the rarity of dAVFs, prospective registries are key to further characterize both efficacy and toxicity of SRS for treatment of dAVFs. The prospectively maintained NeuroPoint Alliance SRS registry allows an avenue for analysis of patient-level data and further characterization of the impact of patient, lesion, and treatment characteristics on clinical outcomes after SRS for dAVFs.5 In addition, the International Radiosurgery Research Foundation (IRRF) has a long-standing history of conducting and publishing multi-institutional outcomes on SRS for a variety of indications. On the topic of SRS for dAVFs, the IRRF has published CO rates without hemorrhage ranging from 37% to 80% for 120 patients dependent on the presence of corticovenous reflux (CVR), prior hemorrhage, and noncavernous sinus (NCS) location.6 These findings are similar to our analysis where prior hemorrhage was associated with higher risk of hemorrhage after SRS and NCS dAVF associated with poorer CO and symptom control rates. Another analysis by the IRRF of 131 patients noted inferior CO rates in patients with NCS dAVFs as compared with cavernous sinus (CS) dAVFs with an association with CO and symptom improvement among NCS dAVFs but not among CS dAVFs.7 These findings in combination with those of our analysis suggest a need for improved CO rates, potentially through multimodality approaches for NCS dAVFs. The latency period for dAVFs, similar to arteriovenous malformations treated with SRS, has been characterized through work by the IRRF with similar rates of hemorrhage noted before and after SRS before CO being achieved.8 As noted by Raper et al, counseling on the risk of hemorrhage after SRS during the latency period is key as is patient selection given the association of prior hemorrhage and risk of hemorrhage during the latency period after SRS. Through these successful multi-institutional initiatives through the IRRF, robust clinical evidence has demonstrated the efficacy and safety of SRS for dAVFs. Further studies examining the potential role of multimodality approaches for patients at risk of unfavorable outcomes (particularly those with prior hemorrhages or NCS dAVFs) are warranted.