INTRODUCTION
Intra-operative cerebral angiography is critical during open surgery for vascular pathology such as aneurysms, arteriovenous malformations (AVMs), or dural arteriovenous fistulas (dAVFs), enabling real-time visualization of complex anatomy, confirmation of complete pathology resolution prior to closure, and early detection of complications such as arterial occlusion or residual aneurysm [1,2]. While cerebral angiography was historically performed through transfemoral access (TFA), there has been a recent trend towards transradial access (TRA), inspired by interventional cardiology literature reporting associations between TRA and reduced access site complications [3], reduced major bleeding events [4], shorter hospital length of stay, and lower overall mortality [5-9], though these cardiovascular findings may not directly translate to neurovascular applications.
However, as TRA becomes more widely incorporated into the neurointerventionist’s armamentarium, certain circumstances remain where both TRA and TFA present technical challenges, namely during prone positioning for cerebrovascular surgery. Prone positioning obstructs easy access to both the groin and radial artery for traditional TFA or TRA, respectively, while also compromising sterile field maintenance and creating risk for access site complications from compression against the operating table. Previous workarounds, such as interrupting surgery for patient repositioning or using extended femoral sheaths, carry risks including increased operative time, loss of surgical positioning, and catheter complications. Transulnar access (TUA) may represent a potentially ergonomically advantageous alternative for patients in the prone position, given the superficial and accessible location of the ulnar artery when patients are positioned prone with thumbs pointing down. While the interventional cardiology literature has investigated TUA through randomized controlled trials and prospective studies, reporting comparable outcomes to TRA in cardiovascular procedures [10-12], detailed technical descriptions specifically focused on TUA during prone positioning remain limited in the literature.
Here, we present our experience with prone TUA for intra-operative cerebral angiography and discuss its technical nuances and clinical applications.
MATERIALS AND METHODS
Study Cohort
We retrospectively reviewed 10 consecutive patients from April 2020 to August 2025 who underwent prone surgery for vascular pathology where TUA was used for intraoperative angiogram. Inclusion criteria were: (1) patients undergoing open neurosurgical resection or treatment of vascular pathology (AVMs or dAVFs) in the prone position and (2) intraoperative cerebral angiography deemed necessary by the operating surgeon. Exclusion criteria were: (1) patients positioned supine or lateral, (2) absence of palpable ulnar artery pulse on preoperative examination, and (3) lack of confirmed radial artery patency on ultrasound evaluation, as radial artery patency was required to prevent potential hand ischemia. All patients meeting inclusion criteria during the study period were included, representing a consecutive series. All cases were performed at the Hospital of the University of Pennsylvania, a quaternary neurovascular referral hospital.
Data Collection
We collected demographic data (age, sex), diagnostic categories (AVMs and dAVFs by location), and lesion laterality for all patients. Prior to TUA, a detailed ultrasound evaluation confirmed patency of ulnar, radial, and distal radial arteries, with ulnar artery access only performed when radial artery patency was confirmed to prevent hand ischemia. We measured technical success of obtaining TUA, whether cerebral angiography was entirely performed via TUA, and mean ulnar artery diameter through ultrasound imaging at the time of access. Procedural variables included surgery type (microsurgical resection vs. clipping), catheter specifications (angled taper glide vs. SIM2 glide), catheter diameter, radiation dose, contrast volume, and fluoroscopy time. We documented any access site complications including hematomas, pseudoaneurysms, peripheral ischemic events, and nerve injury. Follow- up was limited to the immediate perioperative period. No systematic long-term clinical follow-up or delayed vascular imaging was obtained to assess for late complications such as arterial occlusion, delayed pseudoaneurysm formation, or chronic hand ischemia.
Transulnar Access Technique
TUA was performed in all patients while in the supine position before turning prone. Cerebral angiography was performed in all patients while maintained in the prone position in Mayfield pins. With the patient supine, a tailored arm board was placed to support the targeted upper extremity corresponding to the side of the dominant vertebral artery. The wrist was supinated and slightly hyperextended. Manual pressure was applied to the ulnar artery proximal and distal to the entry site with care taken to not occlude the vessel proximal to the puncture site. Ultrasound guidance was used to puncture the ulnar artery 1–3 cm proximal to the pisiform bone with a 21-gauge needle. Arterial access was confirmed through visualization of brisk, pulsating blood flow at the needle hub. Standard access wire was then advanced into the ulnar artery through the cannulation needle. After the wire cleared the intercondylar line of the humerus, the cannulation needle was exchanged for a 5-French sheath. The sheath was secured to the skin with a Tegaderm transparent film dressing (3M). An anti-spasmolytic cocktail consisting of heparin 2,000 u, nitroglycerin 200 mcg, and verapamil 5 mg was then infused through the sheath, followed by attachment to a heparinized saline flush (5,000 units/L) for the duration of the surgery. Prior to positioning, pulse oximetry was placed on the ipsilateral hand for continuous monitoring throughout the procedure, and intraoperative neuromonitoring was established using somatosensory evoked potentials (SSEPs) of the median and ulnar nerves bilaterally, which remained unaffected by the ulnar access. The patient was then placed in Mayfield pins and carefully flipped into the prone position, with care taken to preserve the patency of the sheath during positioning. Fig. 1 shows an example of final patient positioning during prone surgery and subsequent intra-operative angiography. At the conclusion of each case, the patient was returned to the supine position to maintain sterile field integrity. The guide catheter was then removed, and a standard TR arterial compression band (Terumo) was inverted and positioned slightly proximal to the ulnar arteriotomy site for enhanced coverage. The compression band was inflated per standard protocol and subsequently deflated by removing 1 cc of air every 2–3 minutes until hemostasis was achieved while maintaining distal perfusion. Nursing staff monitored the access site throughout the perioperative period for signs of bleeding or vascular compromise.
RESULTS
Intra-operative cerebral angiography was successfully performed via TUA in all 10 patients while maintained in the prone position during surgery for vascular lesions. Eight patients underwent surgery for AVMs and 2 patients for dAVFs. Among the AVM patients, 4 were occipital, 1 was parietal, 1 was vermian, 1 was spinal cord (Type IV), and 1 was cerebellar pial. The 2 dAVF patients had occipital dural fistula and cervicomedullary dural fistula, respectively. Seven patients were male, and 3 patients were female. The average age at surgery was 54.7±15.3 years old. Eight patients had ulnar artery access obtained on the left side, while 2 patients had access obtained on the right side. Mean ulnar artery diameter as determined by ultrasound was 2.4±0.3 mm. Seven patients underwent microsurgical resection while 3 underwent microsurgical clipping. Catheter types used were evenly distributed between angled taper glide and SIM2 glide (5 patients each). Mean fluoroscopy time was 7.5±2.6 minutes, mean contrast used was 50.0±21.6 mL, and mean radiation dose was 3,002.3±3,005.5 mGy*cm (milligray-centimeter). There were no immediate perioperative access site complications in any patient. Follow-up was limited to the immediate postoperative period, and systematic long-term clinical or imaging outcomes were not obtained (Table 1). Individual patient characteristics and outcomes are detailed in Table 2.
Illustrative Case: Right Occipital Arteriovenous Malformation Resection
A healthy patient in their 30s presented to the emergency department with sudden-onset severe right-sided headache and visual disturbance in the left eye while exercising at the gym. Pre-operative MRI and cerebral angiogram demonstrated a Spetzler-Martin grade 2 right occipital AVM. Injection of the left vertebral artery showed a 1.5-cm AVM nidus with supply from the distal calcarine branch of the right posterior cerebral artery and superficial venous drainage into a cortical vein (Fig. 2B). The patient underwent elective right occipital craniotomy for resection of AVM. A lumbar drain was placed pre-operatively for brain relaxation. Left ulnar artery access was established under ultrasound guidance using a 5-French sheath (Fig. 2A, C), as outlined above in the supine position. The patient was then placed in Mayfield pins and flipped prone. Following craniotomy and exposure of the AVM a 2-mm clip was placed on the feeding branch of the right posterior cerebral artery, and a 5-mm clip was placed over the arterialized draining vein. The AVM nidus was circumferentially dissected and removed. Intra-operative cerebral angiogram was performed after AVM resection showing slow drainage within a deep vein with no direct arteriovenous shunting in the arterial phase (Fig. 2D). At completion of the procedure, the catheter and sheath were removed, and wrist hemostasis was achieved with TR band. No complications were encountered during or immediately following completion of the procedure. Post-operatively, the patient recovered in the neurological intensive care unit and was discharged home on post-operative day 2. At 1-month follow-up, the patient reported complete resolution of headache with a small left scotoma.
DISCUSSION
While recent studies have reported using TUA for intraoperative angiography [13], including in prone patients [14-16], we present a detailed technical description and consecutive case series specifically focused on the application of TUA for intraoperative cerebral angiography during prone neurovascular surgery.
Ulnar access for angiography was first described by Terashima et al. [17] in 2001 with a case series of 9 patients that demonstrated the safety and feasibility of ulnar access for coronary angiography, with the primary goal of enabling immediate ambulatory mobilization and preserving the radial artery as a potential coronary bypass graft for surgical myocardial revascularization. Subsequent studies have demonstrated non-inferiority of ulnar access when compared to TRA an TFA using well-described comparative approaches [10-12]. Additionally, Andrade et al. [18] reported that TUA’s high success and low incidence of access-site complications, demonstrating the approach as an alternative to the TRA in selected cases when operated by radial-trained surgeons. Based on cardiovascular literature, ulnar access has been reported to have lower incidence of severe vasospasm, potentially due to decreased adrenergic receptor density in the ulnar artery [19], lower rates of arterial occlusion [20], and easier catheterization due to fewer vascular loops and decreased tortuosity [21], though these cardiovascular findings require validation in neurovascular applications. In contrast to extensive cardiovascular literature on TUA, neurovascular investigations are limited. These preliminary studies all report successful cerebral angiography via ulnar access alone and recommend further research to define indications, benefits, and comparative efficacy versus other arterial access sites [22-26]. We seek to broaden the scope of ulnar arterial access for cerebral angiography by describing its use specifically in patients positioned prone for concomitant craniotomy or laminectomy.
While angiography is unobstructed in procedures treating intracranial neurovascular disease with the patient positioned supine, patients requiring far-lateral suboccipital exposure, occipital-transtentorial exposure, craniocervical junction exposure, and suboccipital exposure are often positioned prone, and therefore obstruct TFA and TRA [2,27]. In addition to physical obstruction to arterial access, prone positioning also results in poor sterility at the access site and poses possible skin injury, thrombus formation, and other catheter-related complications as a result of patient habitus resting on the femoral artery access sheath and associated tubing [28]. To overcome the obstacles posed by prone positioning with traditional TFA, several modified and alternative approaches have been previously suggested and investigated, including 1) closure of the surgical wound followed by supine angiography [29], 2) prone angiography through the femoral artery using an extended femoral sheath [30], 3) transradial arterial access [28], and 4) transpopliteal arterial access [31].
TUA could be ergonomically favorable when patients are in the prone position, as the palmar surface of the hand is naturally exposed, providing an unobstructed and sterile surgical field for catheter manipulation. While our case series included patients with arterial access established prior to prone positioning, TUA could theoretically be delayed and established later when necessary, in the surgical procedure. In many hybrid operating rooms with the necessary equipment for intraoperative angiography, prone patients have the left side facing the surgeon, which may enhance the ease of catheter manipulation as ulnar access is usually established in the left hand, reflecting the left-dominant vertebral circulation in the majority of patients [32].
The choice of arterial access for prone intraoperative angiography requires consideration of the advantages and limitations of each available approach. Traditional TFA, while familiar to most operators, presents several challenges in the prone position. These include risk of catheter kinking during patient positioning, potential sheath thrombosis during prolonged procedures, difficulty maintaining sterile field integrity at the groin access site, and risk of skin injury or vascular compression from the patient's weight resting on the access site and tubing [28,33]. Several authors have described modifications such as extended femoral sheaths to mitigate some of these issues, though challenges remain. TRA, while increasingly popular for neurointerventional procedures in the supine position, presents ergonomic limitations when patients are prone. With the wrist typically positioned palmdown in prone cases, accessing and maintaining stable catheter position through the radial artery becomes more challenging compared to the supine position [33]. Furthermore, TRA may be unsuitable when bilateral vertebral artery access is required, catheter navigation to the contralateral vertebral artery may not be feasible, and radial artery anatomical variations may result in increased technical complexity [33,34]. Popliteal access has been reported as an alternative for prone cases, offering the advantage of avoiding upper extremity positioning issues and maintaining familiar lower extremity vascular access [31,35]. However, this approach requires accessing a deeper vessel in close proximity to the tibial nerve, may be unfamiliar to many operators, and can present challenges for accessing intracranial targets. Each approach has distinct advantages and limitations, and the optimal choice may depend on specific patient anatomy, lesion location, operator experience, and institutional resources. TUA represents one option among these alternatives, with potential ergonomic advantages when the prone position naturally exposes the ulnar artery on the palmar surface of the hand.
TUA in the prone position has limitations. The deeper course of the ulnar artery compared to the radial artery as it traverses the distal forearm may result in decreased compressibility against bony support, potentially presenting greater challenges for achieving hemostasis [21,36,37]. As a result, ensuring adequate hemostasis is even more critical, as ulnar bleeding may not be visible superficially; though complications are rare, providers must remain vigilant in monitoring for signs of forearm hematoma. Ulnar artery hematoma may result in compression of the ulnar nerve, causing paresthesia and either temporary or permanent sensory loss [21]. Additionally, ultrasonography should be utilized to protect against inadvertent damage to the ulnar nerve, which, unlike the radial nerve, runs directly lateral to its respectively named artery [21]. As with any new surgical technique, in our institutional experience the primary surgeon witnessed a subjective learning curve through these 10 consecutive surgeries; during the later operations, compared with earlier operations, there was virtually no disturbance to the typical operating room workflow while obtaining primary arterial access through the ulnar artery in the prone position. However, the learning curve for both surgeons and operating room staff requires further quantification and characterization in larger series.
This study has several limitations. The sample size of 10 patients is small and restricts generalizability. As a descriptive case series, it lacks comparison with transfemoral or TRA, preventing conclusions about relative advantages or complication rates. All procedures were performed by a single experienced vascular neurosurgeon at a high-volume center, which may limit reproducibility in other settings. Follow-up was limited to the immediate perioperative period, without long-term clinical or imaging assessment to detect delayed complications such as arterial occlusion, pseudoaneurysm, hand ischemia, or ulnar nerve injury, leaving their true incidence unknown. Although cardiovascular literature suggests potential benefits of TUA, these advantages require reassessment in neurovascular patients and in the prone position.
CONCLUSION
We demonstrate the technical feasibility of TUA for prone intraoperative cerebral angiography in 10 patients, with no immediate access-site complications. Findings are preliminary, limited by small sample size and lack of comparison, and warrant validation in larger prospective studies with longterm follow-up.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
