Introduction

Cerebral revascularization remains an important technique in the armamentarium of vascular neurosurgeons. Originally described by Yasargil et al.¹ for the treatment of occlusive cerebrovascular disease, the indications for bypass surgery have expanded to include the management of fusiform or giant aneurysms and complex skull-based tumors.^2–6 The choice of bypass conduits has evolved concomitantly. The selection criteria for an appropriate conduit is dependent upon a number of factors, including the extent of blood flow augmentation required, graft length, recipient vessel size, long-term occlusion rates, and the ease, availability and safety of the graft’s harvest.^6–8 The radial artery is frequently an ideal candidate. Although extensively described for use in coronary circulation bypass, the evaluation of employing radial artery grafts for cerebral revascularization has not been paid a similar service. In this chapter, we focus on the preoperative evaluation and technical nuances of radial artery harvest for cerebrovascular bypass.

Radial artery bypass grafts

The use of the radial artery as a conduit for vascular bypass was first reported by Carpentier et al.⁹ in the treatment of coronary artery disease. In 1978, Ausman et al.¹⁰ described the use of a radial artery interposition graft for PICA distribution revascularization in the neurosurgical literature. However, its subsequent popularity as a bypass conduit diminished initially due to difficulties with vasospasm-induced graft failure and trauma during skeletonized harvesting.^11,12 Advancements in calcium channel blockade and the use of the pressure distention technique¹³ led to its resurgence in the late 1990s.¹⁴ The first large series introducing radial artery conduits as a viable alternative to saphenous vein and pedicled arterial grafts for EC-IC bypass was reported by Sekhar et al.¹³ in 2001. Following that landmark study, it remains the conduit of choice for high flow bypass at many institutions.

The radial artery is a medium-sized vessel with a diameter of ∼3.5 mm and a flow rate of 40 to 70 ml/min.¹⁵ It originates at the bifurcation of the brachial artery in the antecubital fossa, supplies muscular branches to the radial forearm, and terminates in the superficial and deep palmar arcades. Here it anastamoses with the ulnar artery to supply the hand (Figure 11–1). The radial artery is comparable in caliber to common sites of cerebrovascular attachment, namely the M2 and P1 segments, which facilitates anastamosis construction and prevents flow mismatch. It is most commonly used in high flow bypasses, following temporary or permanent occlusion of medium to large vessels, when significant flow augmentation is required. With time and demand, radial artery grafts have been shown to distend appreciably and achieve larger flow rates.³ In comparison to saphenous vein grafts, radial arteries have many advantages. They lack the valves and varices present in venous conduits, which increase the risk of graft thrombosis and induce directionality to the conduit. Arterial walls are thick and facilitate anastamosis construction, minimize kinking, and are physiologically designed to carry arterial blood flow and can adapt to changes in pressure and flow. Radial artery grafts have also been shown to better tolerate temporary occlusion and in the long-term are less susceptible to intimal hyperplasia and graft atherosclerosis than vein grafts.¹⁶ They also have improved long-term patency rates as compared to saphenous vein counterparts in both the cardiothoracic^17,18 and neurosurgical^19,20 revascularization literature. In addition, harvest of the radial artery is relatively straightforward given its constant anatomical location within the radial forearm. Donor site morbidity is minimal and includes vascular insufficiency, forearm or hand dysesthesias and weakness, and a low but measurable harvest site infection rate.²¹ The major limiting factor for the use of the radial artery is its length. Frequently, lengths of 20 to 25 cm are required to create a tension-free anastamosis from the neck to a cranial attachment. Poor preoperative screening for this limitation can lead to intraoperative abandonment of the graft. Contraindications for its use as a bypass conduit include forearm ischemia on preoperative testing, severe atherosclerosis or calcification within the graft, and dissection from prior cannulation.

Figure 11–1 Anatomical drawing of the radial artery and surrounding structures in the forearm.

Preoperative Diagnostic Evaluation for Radial Artery Harvest

The preoperative assessment of the intended radial harvest graft is crucial to ensure adequate length and prevent postoperative vascular insufficiency within the hand. The nondominant side is preferred; however, either forearm may be used. If appropriate, the arm contralateral to the planned craniotomy allows for multiple teams to proceed simultaneously in the operative room. Short arms may foreshadow a radial artery graft with an inadequate length, requiring more extensive evaluation (Figure 11–2).

Figure 11–2 Forearm length measured from the proximal transverse furrow of the wrist to the elbow crease overlying the antecubital fossa. A difference of nearly 6 cm was noted between the two subjects, which highlights the variability of graft length.

The first step in the evaluation of a potential radial artery graft should be an Allen’s test to confirm adequate ulnar collateral circulation to the palmar arcades. This test is inexpensive, easy to perform, and is highly reliable. It is conducted by occluding both the radial and ulnar arteries within the forearm until pallor is appreciated in the hand. The ulnar artery is then released and the pallor should be replaced by hyperemic rubor, which will gradually fade to normal hues. If the ulnar artery supply is insufficient, the hand pallor will remain. The modified Allen’s test employs the use of pulse oximetry while conducting the assessment.²² The probe is placed on the index finger, and a baseline amplitude is measured. Allen’s test is then conducted, and, if the amplitude of the curve is low or the value does not return to baseline within 10 seconds, the harvest of that artery is abandoned. In a prospective study in 2001, Meharwal and Trehan²¹ used this technique for the preoperative evaluation of 3977 radial artery grafts and reported no postoperative ischemic hand complications. Alternatively, pulse oximetry can be measured intraoperatively following temporary clipping of the radial artery. Digital plethysmography and duplex ultrasonography have also been reported as adjunct measures to assess the relative flow within the radial and ulnar arteries and confirm the patency of the palmar arcade.^23,24

There are many imaging modalities currently available to preoperatively evaluate the caliber and length of the intended radial artery. These include ultrasound, CT angiography, and intraluminal catheter based angiography (Figures 11-3 through 11-5).

Figure 11–3 Reconstructed CT angiograms of the radial artery in the forearm of the left upper extremity allowing for measurements of length.

Figure 11–4 Reconstructed CT angiogram of a cross-section of the radial artery in the left distal forearm.

Figure 11–5 Digitally subtracted angiogram of the forearm with a brachial artery injection showing flow in the radial artery, ulnar artery, and palmar arcade. Quantitative measurements can be carried out to determine adequate graft length and luminal diameter. In addition, qualitative analysis of collateral flow can also be assessed.

Once the decision to harvest a radial artery graft is made, swift evaluation and site selection should be made so that arterial lines and blood gas analysis are avoided on that side.

Technique for radial artery harvest and preparation

The technique for radial artery bypass harvest for cerebral revascularization is an evolving process. As detailed above, the preoperative evaluation process is imperative for graft selection and harvest safety profiles. The critical aspect of the procedure involves harvesting a graft of adequate length to enable the creation of a tension-free anastomosis. Various techniques have been reported, including open and endoscopic approaches.^25,26 We employ an open technique similar to those previously detailed;^8,13 however, over time our technique has evolved to include several new variations.

Following extensive preoperative evaluation as described above, the chosen forearm is prepared by the harvesting team. >Assisted by palpation and a Doppler probe, an incision overlying the artery is marked from the proximal transverse furrow of the wrist to the antecubital fossa. The arm is prepped and then opened in distal to proximal fashion, depending upon the length necessary for that particular case. Care is taken to avoid injury to the lateral cutaneous nerve of the forearm, which crosses the artery from lateral to medial near the distal end. The radial artery is identified in the deep fascia, between the brachioradialis and the flexor carpi ulnaris. Muscular branches are coagulated 1 mm from the parent vessel with the use of a harmonic scalpel (Starion CardioForceps, Sunnyvale, CA). Meticulous dissection is employed and care is taken not to injure or manipulate the radial artery. The venae comitantes is left attached to the artery, as it is thought that this may prolong the life of the graft.²⁷ We locate one or two large muscular branches at either end of the graft for use as outlets following anastamosis construction. These branches are divided ∼0.5 to 1 cm from the parent artery and preserved with placement of a temporary aneurysm clip. They will be used later for flushing out clot and evaluation of graft patency (Figure 11–6). The artery is then measured for length and marked on its superficial surface to prevent kinking and rotation during the tunneling process (Figure 11–7). The artery remains in situ, covered in wet gauze until the cranial and cervical dissections are complete. The artery is then stitch ligated proximally and distally, preserving the patency of the ulnar and common interosseous arteries. The most common anomaly of the brachial artery bifurcation is a high division of the radial artery, occurring in up to 15% of cases.²⁸ This can be advantageous for conduit harvest, allowing for increased graft length. The artery is then extracted and flushed with a cocktail of heparinized saline, papaverine, and a calcium channel blocker to remove clots and help prevent vasospasm.¹³ Various antispasmotic agents have been employed both directly and systemically for this purpose.²⁹ The pressure distention technique as described by Sekhar et al.¹³ is then undertaken. This involves blunt canalization of the radial artery graft followed by meticulous sequential dilation with heparinized saline. Any remaining patent side branches are cauterized closed with care to avoid injury to the parent vessel. The adventitia is then stripped from both ends of the graft for 1 to 1.5 cm to help facilitate a “clean” anastamosis. The graft remains in a heparinized saline bath until it is ready for implantation. Following proper hemostasis, the forearm is closed in two layers and the skin closed in subcuticular fashion with absorbable suture.

Figure 11–6 A, Diagram showing outlet branch for flushing radial artery prior to revascularization. B, Diagram showing the utility of an outlet branch in determining anastomosis patency.

Figure 11–7 In situ radial artery graft awaiting harvest. The anterior wall is marked to avoid torsion of the artery during tunneling.

Construction of the extra and intracranial anastomosis and tunneling of the graft are then undertaken with a technique discussed extensively elsewhere.³⁰ Prior to releasing the vascular clamps on the cranial circulation, the preserved large muscular side branches are used to flush out any remaining clot in a manner similar to that of the external carotid artery during a carotid endarterectomy to prevent showering embolic stroke (Figure 11–6A). This technique also ensures patency of the graft and donor vessels immediately following anastomosis construction and can allow for graft evaluation without the use of an additional arteriotomy should micro-Doppler analysis reveal poor flow (Figure 11–6B). Many flow probes exist for evaluation of the graft, including micro-Doppler ultrasonography (Mizuho America, Inc., Ithaca, NY, and Transonic Systems, Inc., Beverly, MA). Pulsatility within the graft is encouraging; however, it is not definitive for patency. On-lay vasodilators can assist in temporarily maximizing graft flow. Intraoperative angiography remains the gold standard to confirm flow. Additionally, ICG video-angiography has become an integral tool within the vascular neurosurgeon’s arsenal and has been shown to be a valid measure of bypass patency.³¹

Complications

Complications attributable to graft occlusion are limited in the neurosurgical literature. Sekhar and Kalavakonda published their series documenting a 15-year period. A total of 24 radial arteries were harvested in 50 patients with aneurysms and 83 patients with skull base tumors requiring a revascularization procedure. The overall graft patency rate in their cohort was 95.6% with the pressure distension technique credited with overcoming the problem of arterial vasospasm previously encountered with radial artery conduits.³² Similarly, Houkin et al. reported a 100% patency rate in 20 of 43 patients who underwent postoperative MRA and DSA following high-flow EC-IC bypass with radial artery conduits over a 10-year period.¹⁹ In the cardiothoracic literature, excellent results have been found with the use of radial artery grafts. The short-, mid-, and long-term patency rates are 96 to 100%, 94 to 97%, and 84 o 96%, respectively;²⁹ however, these grafts are typically shorter in nature and may be prone to striking differences in vascular microenvironments. Saphenous vein grafts patency rates are significantly lower in the neurosurgical literature with short-, mid-, and long-term patency rates of 86%, 82%, and 73%, respectively.³³ In 2004, Desai et al.¹⁷ enrolled 561 patients in 13 centers into a randomized controlled comparison of radial artery versus saphenous vein conduits for coronary bypass surgery. At 1 year, they found a 5.4% absolute lower occlusion rate in radial arteries as determined by follow-up angiography.

Intraoperative graft occlusions are often secondary to technical issues or hypercoagulable states and are addressed immediately. Postoperative occlusion can occur at any time point and have a number of etiologies related to anastomosis failure, donor or recipient vasculopathy, graft kinking, graft stenosis, flow mismatch, and graft spasm. In addition, intimal hyperplasia and disruption of the vaso vasorum have been implicated as possible etiologies of late occlusion.^11,27

The majority of literature describing donor site complications of employing the radial artery as a vascular conduit stems from the thoracic literature as it has been used extensively in coronary bypass surgery. General surgical complications such as wound infection, hematoma formation, and wound dehiscence are encountered rarely. Wound infection of radial artery donor sites range between 2% and 12% with a preoperative diagnosis of diabetes mellitus and a surgical duration of greater than 5 hours identified as independent risk factors.^21,34–36 The most common organism encountered is Staphylococcus aureus.³⁶ Postoperative hematomas were observed to develop in 0.3 to 6.4% of cases.^21,34,37,38 Wound dehiscence requiring resuturing is encountered very rarely with one study documenting the incidence at 0.16%.²¹ Neurologic or vascular compromise as a sequelae of radial artery harvest is a rare event and is typically self-limiting in nature. In one study of 560 patients, neurologic complications were reported in 30% with a sensory abnormality in 18% and decreased thumb strength observed in 5%. However, only 12% of patients reported sensory or motor abnormalities that did not improve within 8 months.³⁹ Another study employing the radial artery in 3477 patients found that initially 28% of patients complained of paresthesias and 12% complained of limitation of hand activity that decreased to 3% and 1.22%, respectively, after 6 months.²¹ In the majority of cases, it appears that the sensory deficits correlate to involvement of the lateral antebrachial cutaneous nerve.^37,40 Although hand ischemia has been described in the literature, the incidence of blood flow abnormalities postoperatively was not found to be significant in those patients who had a negative Allen’s test preoperatively.^21,40,41

Conclusion

The selection of an appropriate bypass graft is made on a case-by-case basis. The neurovascular surgeon should be facile with the harvest, implantation, and management of complications arising from the gamut of graft choices. Radial artery grafts offer a number of advantages and should be given thoughtful consideration when planning a bypass procedure. They have been proven to be safe, durable, and relatively straightforward to use. Thorough preoperative evaluation will aid in establishing graft length and minimizing an already low complication profile. The intraoperative outlet technique described above can aid in stroke prevention and graft evaluation should a thrombus arise.