Transradial Approach for Neuroendovascular Procedures: A Single-Center Review of Safety and Feasibility

BACKGROUND AND PURPOSE: In recent years, the transradial approach has become more widely adopted for neuroendovascular procedures. The purpose of this study was to evaluate the safety and feasibility of a transradial approach and distal transradial access for neuroendovascular procedures in a single center. MATERIALS AND METHODS: Retrospective analysis was performed for all patients who underwent transradial approach or distal transradial access neuroendovascular procedures from January 2016 to August 2019 at a single center. Exclusion criteria included a Barbeau D waveform, a radial artery of , 2 mm on sonographic evaluation, and known radial artery occlusion. Procedures were evaluated for technical success (de ﬁ ned as successful radial artery access and completion of the intended procedure without crossover to an auxiliary access site), complications, and adverse events during follow-up at 30 days. RESULTS: The transradial approach or distal transradial access was attempted in 279 consecutive patients (58.1% women; median age, 57.7 years) who underwent 328 standard or distal transradial approach procedures. Two-hundred seventy-nine transradial approach and 49 distal transradial approach procedures were performed (cerebral angiography [ n ¼ 213], intracranial intervention [ n ¼ 64], head and neck intervention [ n ¼ 30], and stroke intervention [ n ¼ 21]). Technical success was 92.1%. Immediate adverse events (2.1%) included radial access site hematoma ( n ¼ 5), radial artery occlusion ( n ¼ 1), and acute severe radial artery spasm ( n ¼ 1). Thirty-day adverse events (0.3%) included a radial artery pseudoaneurysm ( n ¼ 1). Twenty-six cases (7.9%) required crossover to transfemoral access. CONCLUSIONS: The transradial approach for neuroendovascular procedures is safe and feasible across a wide range of neuroendovascular interventions.

T he transradial approach (TRA) is well-established within interventional cardiology and is recommended as first-choice access by the American Heart Association. [1][2][3][4][5][6] Large, randomized, multicenter trials from the cardiology literature show that TRA is associated with lower rates of access site morbidity and mortality compared with the transfemoral approach (TFA). [7][8][9][10][11] Additional evidence supports the use of TRA in peripheral vascular interventions. 12 Given the associated benefits to patient safety, satisfaction, and decreased recovery time, a first-line transradial approach for endovascular surgery has gained traction in both peripheral and neurointerventional radiology. [13][14][15] Matsumoto et al 16,17 published the earliest study of TRA for cerebral angiography in 2000, demonstrating the technical feasibility and safety of this approach. Since then, further studies have added to the neuroendovascular literature on TRA. [18][19][20][21] Nevertheless, the common femoral artery remains the primary access site for cerebral angiography and neurointerventional procedures.
The distal transradial approach (dTRA) is a further modification by which the radial artery is accessed distal to the flexor retinaculum as the radial branch courses superficial to the scaphoid bone within the anatomic snuffbox. This access site allows preservation of the more proximal radial artery for future interventions or access requirements.
Here we present our initial institutional experience performing TRA and dTRA for diagnostic neuroangiography and neurointerventions.

Study Design and Exclusion Criteria
This single-center retrospective study was approved by the local institutional review board, and all study activities followed Health Insurance Portability and Accountability Act regulations. A retrospective analysis was performed of 279 patients evaluated for 328 consecutive TRA or dTRA neuroendovascular procedures from January 2016 to December 2019.
All neurointerventional procedures, both diagnostic and interventional, were included and separated into 4 groups: cerebral angiography, head and neck interventions, intracranial interventions, and acute stroke intervention/mechanical thrombectomy.
All patients were given the option of TRA and consented appropriately. Testing for ulnar-palmar arch patency was performed using a technique previously described by Barbeau et al. 22 This practice was abandoned in 2019 following updated guidelines from the Society of Neuro-Interventional Surgery Standards and Guidelines Committee highlighting the limitations of the test in predicting ischemic complications. 23 Before these updated guidelines exclusion criteria for TRA included a Barbeau D waveform (when Barbeau test was performed), a radial artery diameter ,2 mm on initial sonographic evaluation, and known radial artery occlusion. Radial artery diameter , 2 mm is treated as a relative contraindication to radial access at our institution due to the reported association between a smaller radial artery diameter and an increased risk of radial artery occlusion following TRA procedures. 24 Distal TRA was specifically considered in patients with limited supination of the arm, though the decision to procced with dTRA was ultimately made by the operator.
Additional patient demographic data were collected retrospectively using the electronic medical record system, Epic Systems (Verona, Wisconsin) and the Mount Sinai Data Warehouse (Table 1).

TRA Access
Standard radial artery access technique, previously described by Posham et al 12 and Snelling et al, 21 was used for all procedures. Under sonographic guidance, the radial artery was localized with compression on B-mode and color Doppler flowmetry, confirming patency. Local anesthesia (1% lidocaine) was infiltrated in the subdermal layer. A single-wall puncture of the radial artery was performed at the level of the radial epiphysis on the volar radial surface under real-time sonographic guidance. Using the modified Seldinger technique, the proceduralist inserted an access sheath into the radial artery.

dTRA Access
The patient was positioned with the access arm partially pronated at his or her side in a thumbs-up position. Under sonographic guidance, the metacarpals of the thumb and index finger were identified first, followed by the trapezium and the scaphoid bones below the distal radial artery. A single-wall puncture of the artery was performed through the anatomic snuffbox under real-time sonographic visualization. Using the modified Seldinger technique, the proceduralist inserted an access sheath into the radial artery.

Procedural Technique
A 4F or 5F Glidesheath Slender (Terumo) sheath was used for diagnostic studies, and a 5F or 6F Glidesheath Slender sheath was used for most head and neck and intracranial interventions. For stroke interventions and select aneurysm treatments using flow diversion, a 0.088-inch-long guide sheath was used without a short sheath. Maintenance of the radial access sheath on continuous heparinized flush was left to the proceduralist's discretion. Following sheath insertion, a radial artery access cocktail consisting of 3000 U of heparin, 200 mg of nitroglycerin, and 2.5 mg of verapamil was mixed with 10 mL of the patient's blood and slowly infused through the sheath for 5-10 minutes to induce vasodilation. Administration of the access cocktail was omitted during the stroke intervention to expedite the procedure and avoid inducing hypotension. Equipment selection was dependent on the procedure type. For diagnostic angiography, a 5F Simmons 2 Glidecath catheter (Terumo) was advanced over a Bentson guidewire and formed within the aortic arch before selectively catheterizing the vessels of interest. For most nonstroke interventions, a 0.071-inch guide catheter was advanced over a 125-cm Select catheter (Penumbra) and guidewire into the vessel of interest. For stroke interventions and selected flow-diverter cases, a 0.088-inch guide catheter was used without a sheath and advanced over a selecting catheter and stiff guidewire for added support. Once the guide catheter was positioned, the procedure followed a similar vessel-selection technique used during femoral access.
Following the procedure, a radial artery compression device was placed over the arteriotomy site for a minimum of 60 minutes. The band was removed after arterial hemostasis and pulse were reconfirmed. Similar closure techniques were followed for both TRA and dTRA procedures. Repeat evaluation of the access site and radial pulse was performed for all patients before discharge and during the 30 day follow-up outpatient visit.

Technical Success and Complications
Technical success was defined as successful radial artery access and completion of the intended procedure without crossover to an auxiliary access site.
Major complications included the need for prolonged hospitalization, an unplanned increase in the level of care, permanent adverse sequelae, and death. Minor complications included the need for additional nominal therapy, overnight admission for observation, loss of a radial pulse without evidence of distal ischemia, and hematoma or blood loss not requiring transfusion or open surgical repair. In addition to the periprocedural data, the access site was evaluated for bleeding and neurologic events before and after the procedure with additional evaluations up to 30 days postprocedure.

Statistical Analysis
Data are presented as mean and range for continuous variables and frequency for categoric variables. Univariate and multivariate analyses of complications and crossover were performed using logistic regression. A P value of # .05 was considered statistically significant. SPSS Statistics for Windows, Version 23.0 (IBM) was used for statistical analyses.

Procedure
A total of 328 radial procedures were performed during the study period, including cerebral angiography (n ¼ 213), intracranial interventions (n ¼ 64), head and neck interventions (n ¼ 30), and stroke interventions (n ¼ 21) (Table 1). Of the 328 procedures, 279 procedures used the TRA and 49 procedures used the dTRA. A total of 51 procedures were performed in patients who had undergone prior TRA, and 8 cases were transitioned to TRA after failed TFA. Crossover to the TRA was most often necessary due to tortuous vascular anatomy preventing supra-aortic vessel selection from the femoral artery. Radial artery micropuncture and vessel access were successful in all cases. Completion of the intended procedure via TRA was achieved in 255 of 279 cases (91.3%), while it was via the dTRA in 47 of 49 cases (95.9%). Composite success in completing the intended procedure via TRA or dTRA was achieved in 302 of 328 cases (92.1%).

Crossover to TFA
Crossover to TFA was required to successfully complete 26 of 328 (7.9%) cases. Of all TRA crossovers, 19 (73.1%) were attributed to vascular anomalies or severe tortuosity of the subclavian artery (n ¼ 7), aortic arch (n ¼ 7), or carotid artery (n ¼ 5), which undermined stability and/or prevented navigation of the catheter system. An additional 6 failures (23.1%) were related to radial artery vasospasm (n ¼ 3), severe radial and brachial artery tortuosity secondary to hypertensive vasculopathy (n ¼ 1), and an aberrant right subclavian artery morphology that prevented target artery access (n ¼ 2). In 1 patient, crossover to TFA was performed due to patient discomfort.
Analysis of TFA crossover is presented in the Online Table. There were no significant predictors of crossover from TRA to TFA.
The Barbeau test was performed for assessment of collateral circulation to the hand in 306 cases (93.3%) and was stopped in 2019 following guidelines that highlight the inability of the test to predict ischemic complications. 25 All tested patients had favorable Barbeau test outcomes (grade A and B), and no correlation was demonstrated between the test result and access site complication or crossover to TFA.

Complications
A total of 8 adverse events (2.4%) were recorded among all radial artery cases: 7 in the TRA group (2.5%, 7/279) and 1 in the dTRA group (2.0%, 1/49). Seven of these events were immediate, and 1 was delayed (noted at 30-day follow-up). Immediate adverse events included hematoma (n ¼ 5), radial artery occlusion (n ¼ 1), and severe radial artery spasm (n ¼ 1). The delayed adverse event was a radial artery pseudoaneurysm measuring , 2 mm that was managed conservatively with observation. An analysis of the TRA adverse events is presented in Table 2. Larger sheath size ($6 F) was a significant predictor of adverse events (P , .001).

DISCUSSION
The success and widespread adoption of the TRA in cardiology has highlighted the benefits of radial access. Unlike traditional femoral access, TRA is associated with a decreased risk of bleeding and arterial damage, both of which are critical for patients on anticoagulant or antiplatelet therapy. 12 TRA has been associated with lower access site infection rates, expedited postprocedural ambulation, reduced postprocedure nursing care, and shorter hospital stay. 21 Patient preference for TRA over TFA has been demonstrated previously. 26 The present study describes data from 328 TRA neuroendovascular procedures performed in 279 patients. Both diagnostic and interventional procedures were included from a practice of 8 neurointerventionalists, each having varying levels of TRA case experience. Two interventionalists performed most procedures (40% and 30%, respectively). Six interventionalists performed between 11 and 28 procedures each (median, 13.5).
Our overall success rate, defined as successful completion of the intended procedure from the initial radial access, was 92.2% (302/328 cases). The success rate of TRA neuroendovascular procedures in the literature is reportedly 92.7%-99%. 18 A dedicated analysis of TRA to TFA crossover in the current study showed that the primary factors contributing to crossover included the following: vascular anomalies and tortuosity of the subclavian artery (n ¼ 7), aortic arch (n ¼ 7), and carotid artery (n ¼ 5); radial artery spasm (n ¼ 3); severe radial artery tortuosity (n ¼ 1); variant arch anatomy (n ¼ 2); and excessive arm pain (n ¼ 1). A combination of factors led to the high crossover rate (7.9%) in the current study, including operator experience, position on the TRA learning curve, availability of access catheters, and case selection.
Various studies have examined the relationship between operator volume and procedural outcomes, suggesting 30-50 cases as the general learning curve. 27,28 We separately analyzed the 100 most recent cases and demonstrated a technical success rate of 95% with 5 crossover cases (5%). While the operator's learning curve was not directly studied, this is a likely contributor to the overall crossover rate of 7.9% for the entire cohort. To that end, this study was performed at an academic teaching hospital with neurointerventional trainees of varying experience in performing transradial neurointerventional procedures.
Although our study did not elucidate significant predictors for conversion from TRA to TFA, other, non-neurointervention-focused studies have identified such predictors. Posham et al 12 noted procedural type, female sex, and height of ,1.7 m as significant predictors of radial-to-femoral crossover using univariate analysis. However, these were not significant by multivariate analysis. Additionally, Carvalho et al 29 reported short sheaths, female sex, multivessel disease, body surface area, and age older than 66 years as independent predictors of conversion from radial to femoral access. To our knowledge, no prior studies on TRA neuroendovascular procedures have examined the predictors of conversion from radial to femoral access.
The radial artery is easily compressible against the volar radial surface, permitting easy hemostasis. Titano et al 30 showed that TRA was associated with a low incidence of bleeding complications in patients with an elevated international normalized ratio (.1.5) undergoing below-neck interventions. Our study reaffirms this finding in a neurointerventional population with no significant difference in adverse events between patients with and without coagulopathy.
The overall adverse event rate in our study was 2.4%, including 7 immediate adverse events (access site hematoma [n ¼ 5], radial artery occlusion [n ¼ 1], severe radial artery spasm [n ¼ 1], and one 30-day adverse event (radial artery pseudoaneurysm of , 2 mm). No major access site complications occurred. This complication rate of 2.4% is similar to those published for TRA in visceral interventions and TRA in neurointerventions. 12,31 All immediate adverse events were managed conservatively. An analysis of the total adverse events is presented in Table 3. Significantly higher adverse events rates were observed for procedures that used 6F sheaths (P , .001). All access site hematomas were small and resolved following conservative management. One of the 5 patients who developed an access site hematoma had been placed on dual-antiplatelet therapy following intracranial stent placement.
Four of the 8 total adverse events occurred when a 6F sheath was used (access site hematoma [n ¼ 3] and severe radial artery spasm [n ¼ 1]). The significant association observed between larger sheath size and access site compilations is in keeping with previous studies examining both transfemoral and transradial procedures. [32][33][34] This association is thought to arise from larger radial arteriotomy size, increased potential for intimal damage, and longer procedure times associated with interventional cases in which 6F sheaths are common. dTRA was used in 49 cases with a procedural completion rate without crossover to TFA of 95.6% (47/49). This outcome measure and the reported adverse events are comparable with those in the TRA cohort and previously reported dTRA data. dTRA allows preservation of the more proximal radial artery, which may be required for other interventions/therapies, enhanced operator and patient comfort, and easy hemostasis.
Direct comparison of complication rates of TRA and TFA cannot be made from the current study. However, comparison of complication rates for TRA versus TFA have been described previously. 35, 36 Stone et al 35 performed a prospective comparison of TFA to TRA for diagnostic cerebral angiography and demonstrated no significant difference in complication rates between the 2 groups.
Our study did not demonstrate a relationship between repeat catheterization and adverse events. Repeat TRA was performed in 51 cases (15.5%) and accounted for 2 of the 8 total minor adverse events (25%): 1 hematoma and 1 radial artery occlusion. These findings are consistent with the previously published literature, describing same-site repeat TRA success rates of .94% with up to 6 prior TRA procedures. 25 The primary limitation of the present study is the retrospective design and lack of a comparator control arm, making it difficult to appreciate subtle differences in procedural success and complication rates. Additionally, patient selection for TRA was based on both operator and patient preference in addition to the specific procedure performed.

CONCLUSIONS
Our institution's initial experience using TRA for diagnostic angiography and neurointerventional procedures was comparable with that of previously reported TRA studies. No major complications were noted, and the rate of femoral crossovers will likely continue to improve with technical familiarity and case selection. Furthermore, favorable outcomes were observed in patients on anticoagulation, mirroring findings from the coronary/cardiac literature. The transition of practice to a radial first approach is a safe and worthwhile endeavor that may substantially improve the morbidity and mortality of select patients.