OA-PICA Bypass

Published on 08/03/2015 by admin

Filed under Neurosurgery

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 2532 times

9 OA-PICA Bypass

Introduction

Yasargil et al.1 demonstrated the feasibility of EC-IC bypass surgery in 1967 after performing the first successful superficial temporal artery to middle cerebral artery anastomosis. Ausman and colleagues2 applied these techniques for vertebral artery (VA) occlusive disease and reported the first intracranial posterior circulation revascularization procedure in 1976 after performing an occipital artery (OA) to posterior inferior cerebellar artery (PICA) anastomosis. Posterior circulation bypass surgery has since been used to treat vertebrobasilar insufficiency, skull base tumors involving the VA and its branches, and complex or giant posterior circulation aneurysms. Current advances in endovascular techniques and the availability of intracranial stents have largely limited the need for posterior circulation bypass surgery for vertebrobasilar insufficiency. Occipital artery–PICA bypass surgery, however, remains a valuable tool in the management of fusiform aneurysms of the VA encompassing the PICA origin as well as giant and complex VA/PICA and PICA aneurysms that cannot be reconstructed with surgical clipping or endovascular coiling and require parent artery occlusion or trapping. In the context of VB insufficiency, the bypass may still have a role in the rare settings of poor endovascular access due to tortuosity or occlusion. This chapter reviews the relevant anatomy, surgical technique, perioperative care, and complications associated with OA-PICA revascularization.

Anatomy

Occipital Artery

The OA is an ideal choice for PICA revascularization as it can be encountered during the suboccipital approach and closely approximates the size of the PICA. The course of the OA can be divided into three segments.3

The first segment, also known as the digastric segment, extends from the origin of the OA from the external carotid artery to the point of emergence from the occipital groove of the mastoid process. The OA originates from the posterior or lateral wall of the external carotid artery at the level of the angle of the mandible. The artery ascends medial to the external carotid artery and lateral to the internal jugular vein to a point posterior and medial to the styloid process. The OA then courses posteriorly and laterally superficial to the rectus capitus lateralis muscle first and then superior oblique muscle.4 The artery is covered by the posterior belly of the digastric muscle laterally, hence known as the digastric segment. The artery then runs in the occipital groove or occasionally a true bony canal3 medial to the mastoid notch, in which the posterior belly of the digastric muscle arises.4

The second segment, also known as the suboccipital or horizontal segment, extends from the emergence of the OA from the occipital groove of the mastoid process to the superior nuchal line. The OA exits the occipital groove between the superior oblique muscle and posterior belly of the digastric and is covered by the splenium capitis and sternocleidomastoid. The artery courses medially in a horizontal plane either superficial or deep to the longissimus capitus muscle, depending on whether the occipital groove is absent or present. The OA then continues superficial to the semispinalis capitis muscle just below the superior nuchal line in the upper part of the posterior triangle. The artery then changes course and runs vertically upward, piercing the fascia connecting the cranial attachment of the trapezius and sternocleidomastoid muscles to the superior nuchal line.5 The suboccipital segment gives rise to ascending and descending muscular branches as well as transosseous branches to the dura of the posterior fossa. The diameter of the suboccipital segment ranges from 1.6 to 2.2 mm (mean 1.9 mm) and the length ranges from 75 to 85 mm (range 79.3 mm).6

The third segment, also known as the occipital or subgaleal segment, begins at the superior nuchal line after the OA pierces the fascial attachment of the trapezius and sternocleidomastoid muscles. In a cadaveric study, the OA was found to cross the superior nuchal line approximately 35 mm (±0.5 mm) lateral to the inion.3 The artery continues underneath the galea and above the occipitalis muscle before dividing into its terminal branches. The diameter of the OA at the superior nuchal line is approximately 1.4 mm (±0.3 mm).3

PICA Artery

Segments of the PICA

It is very convenient and appropriate to subscribe to the concept advanced by Lister et al.,7 that the PICA can be divided into five segments based on its relationship to the medulla and the cerebellum: anterior medullary, lateral medullary, tonsillomedullary, telovelotonsillar, and cortical.

The anterior medullary segment begins at the origin of the PICA from the VA anterior to the medulla oblongata. Congenital anomalies of the PICA include double origin, fenestration, or duplicated PICAs, a common anterior inferior cerebellar artery (AICA)–PICA configuration, a VA termination at the PICA, extradural origins at C1 and C2 levels (5% to 20%), origins at the hypoglossal, proatlantal, or posterior meningeal arteries, and at all points along the intradural VA.815 From its origin, the artery runs posteriorly around the medulla past the exit of the hypoglossal nerve rootlets from the anterior border of the inferior olivary complex to the boundary between the anterior and lateral surfaces of the medulla, which is marked by a rostrocaudal line through the most prominent part of the inferior olive. The lateral medullary segment begins at the point where the PICA passes the most prominent part of the inferior olive and extends posteriorly around the lateral aspect of the medulla to the origin of the glossopharyngeal, vagus, and accessory rootlets from the posterior border of the inferior olivary complex. The PICA continues medially as the tonsillomedullary segment between the lower half of the cerebellar tonsil and the posterior aspect of the medulla oblongata. The artery makes a caudally convex curve as it passes around the lower pole of the cerebellar tonsil known as the caudal or infratonsillar loop. The caudal loop is frequently used as a recipient during OA-PICA bypasses and its diameter ranges from 0.9 to 1.4 mm (mean 1.2 mm).6 After forming the caudal loop, the PICA ascends to the midlevel of the medial surface of the tonsil where it becomes the telovelotonsillar segment. The artery continues along the medial surface of the tonsil toward the roof of the fourth ventricle. The PICA then forms a rostrally convex curve referred to as the cranial or supratonsillar loop. This loop consists of proximal ascending and distal descending limbs and an apex that lies caudal to the fastigium at the center of the inferior medullary velum. The ascending limb runs posterior to the tela choroidea and inferior medullary velum toward the fastigium of the fourth ventricle. The descending limb runs posteriorly in the fissure between the vermis medially and the superomedial surface of the tonsil and cerebellar hemisphere laterally. The PICA emerges from the fissure and continues as the cortical segment. The artery divides into a smaller medial and a larger lateral trunk and subsequently gives rise to hemispheric, vermian, and tonsillar branches.

Branches of the PICA

The PICA gives rise to three types of branches:7

Preoperative evaluation

Preoperative evaluation includes standard preoperative labs, an electrocardiogram, a chest radiograph, and assessment of the patient’s general health. In contrast to dissection of the superficial temporal artery, harvesting the OA requires considerable muscle dissection and may lead to postoperative hematomas particularly if patients are on aspirin. We therefore do not give patients aspirin preoperatively and prefer to stop aspirin 1 week prior to surgery in patients on aspirin therapy for vertebrobasilar insufficiency.

Digital subtraction cerebral angiography with three-dimensional reconstruction as well as selective external carotid artery injections is essential. Aneurysm size, morphology, dome-to-neck ratio, and relation to the surrounding arteries are evaluated. The caliber and course of the OA as well as the caliber and configuration of the PICA vessel and its tributaries should be carefully studied. The course, position, and proximity of the tonsillomedullary segments of both PICAs need assessment, in order to have a side-to-side PICA-PICA bypass as an alternate plan. In cases of vertebrobasilar insufficiency, the caliber of both VAs and the presence and caliber of the posterior communicating arteries must be evaluated and if necessary an Alcock’s test should be performed (carotid compression during vertebral artery injection).

Physiological imaging modalities such as positron emission tomography, xenon computed tomography, single-photon emission computed tomography, computed tomography perfusion, and magnetic resonance perfusion commonly used to detect hemodynamic compromise in anterior circulation occlusive disease are less effective in assessing the posterior circulation as a result of their limited regional resolution.16 Furthermore, the validity of these imaging modalities in detecting posterior circulation hypoperfusion remains uncertain. Phase-contrast quantitative magnetic resonance angiography (QMRA) has become available in recent years and is capable of directly measuring volumetric blood flow (milliliters per minute) through the major vessels of the posterior (as well as anterior) circulation. The technique is now implemented and enhanced in commercially available software called the NOVA (Noninvasive Optimal Vessel Analysis) system (VasSol, Inc., Chicago, IL). Table 9–1 shows the mean blood flow values and ranges for posterior circulation vessels in 50 healthy patients.17 In a retrospective study of 47 patients with symptomatic vertebrobasilar disease, patients with >20% reduction of blood flow in the basilar artery (<120 cc/min) and PCAs (<40 cc/min) had a higher risk of stroke as compared to patients with preserved blood flow.17 The VERiTAS (Vertebrobasilar Flow Evaluation and Risk of Transient Ischemic Attack and Stroke) is an ongoing prospective multicenter observational study funded by the National Institutes of Health and aimed at determining the utility of QMRA in assessing patients with symptomatic vertebrobasilar occlusive disease of ≥50%. If predictive, QMRA evaluation could help identify high-risk patients who would benefit most from either surgical revascularization or endovascular angioplasty and stenting.

Table 9–1 The Mean Value and Ranges of Blood Flow for Posterior Circulation Vessels in 50 Healthy Volunteers.

Vessel Mean Flow (cc/min) Range* (cc/min)
BA 190 150–230
LPCA 72 50–94
RPCA 68 50–86
LVA 126 94–158
RVA 110 81–139

BA, indicates basilar artery; PCA, posterior cerebral artery; VA, vertebral artery; L, left; R, right.

* Mean ± standard deviation.

Surgical indications

Indications for cerebral revascularization procedures can be divided into flow augmentation and flow replacement.

Anesthetic technique and neuroprotection

The surgery is performed under general anesthesia. Hypotension must be avoided during the initial part of the procedure particularly in patients with marginal cerebral perfusion. Hyperventilation and alpha-adrenergic agents are not recommended because of their vasoconstrictive effects. We routinely use modest hypothermia (33°C) throughout the procedure as well as induced blood pressure elevation to 20% to 30% of baseline during the period of temporary PICA cross-clamping to augment collateral flow. Electrophysiologic monitoring using somatosensory and motor-evoked potentials and brainstem auditory-evoked potentials allows early detection of ischemia or excessive retraction and manipulation. Barbiturates have been used during bypass surgery to increase tolerance to cerebral hypoperfusion. The mechanisms of barbiturate neuroprotection are multifactorial and incompletely understood. It is believed that a reversible, dose-dependent depression of cerebral blood flow occurs, with subsequent reduction in cerebral metabolic rate and intracranial pressure.1821 Furthermore, vasoconstriction in normal areas of the brain may result in an inverse steal phenomenon with redistribution of cerebral blood flow to ischemic tissue.22 At a cellular level, barbiturates have been demonstrated to reduce ischemia-induced glutamate release,23 enhance GABA-ergic transmission,24,25 and reduce ischemia-induced intracellular calcium through inhibition of both voltage-gated calcium channels and NMDA receptors.26 In addition to the fore mentioned neuroprotective properties, barbiturates may also act as a scavenger of membrane-damaging free radicals. Despite these potential benefits, we do not routinely use barbiturate neuroprotection during OA-PICA bypass surgery due to problems associated with circulatory and respiratory depression as well as delayed postoperative wake-up. However, if technical difficulties during the anastomosis result in excessively prolonged temporary occlusion, we may consider using barbiturates.

Surgical technique

Figures 9-1 through 9-3 show preoperative imaging of a patient diagnosed with fusiform right VA-PICA origin large aneurysm.

The patient is either placed prone (more likely for a pure ischemia case), or more commonly in a three-quarter prone position (more likely for an aneurysm or tumor case) for a far lateral approach if access to the PICA origin is necessary. For the prone position, the head is placed in moderate flexion and secured in a three-pin head holder. Proper Mayfield pin placement is crucial, as it may hinder the procedure if placed improperly. The pins are placed so that the single pin is 2 cm superior and anterior to the pinna ipsilateral to the donor OA. The paired pins are positioned so that the posterior pin is 2 cm above the contralateral ear pinna. The head is positioned above the level of the heart to reduce cerebral venous congestion. In the three-quarter prone position, the head is positioned with four movements (Figure 9–4):

A portable transcutaneous Doppler probe is used to identify the course of the OA over the scalp from the mastoid to approximately 4 cm above the superior nuchal line (Figure 9–5). Scalp infiltration solutions containing vasoconstrictive agents should not be used. A hockey-stick incision is then made. The incision starts approximately at the level of the spinous process of C3 and extends superiorly in the avascular midline plane to approximately 2 cm above the superior nuchal line. The incision is then turned laterally parallel to the superior nuchal line. As this limb of the incision approaches the point at which the distal OA crosses the incision, a curved hemostat is used to dissect over and protect the OA. The incision is then continued over the OA to a point immediately superior to the mastoid process. Finally, the incision is curved inferiorly to end just inferior to the mastoid tip. The distal OA is then identified and dissected proximally in its muscular plane. Although this may be performed under the operative microscope, in our opinion loupe magnification is sufficient and more efficient. Dissection of the OA is generally the most difficult part of the procedure due to the tortuosity and adherence of the artery to the surrounding tissue. The OA is typically surrounded by a venous plexus and runs with the occipital nerve in a fascial sheath. A generous cuff of periadventitial tissue is left around the artery. Small side branches are carefully coagulated using low current bipolar forceps, so as not to cause thermal injury to the parent artery. The side branches are then sectioned at a distance from their origin from the OA trunk. It is important to dissect the artery as far proximally as the occipital groove to ensure adequate length of the graft. The OA is wrapped in a papaverine-soaked cottonoid to relieve spasms related to vessel manipulation, and is left in continuity until just before it is required for the anastomosis (Figure 9–6). The suboccipital musculature is swept laterally in a subperiosteal fashion to expose the occiput as far laterally as the mastoid process as well as the arch of C1. The skin and muscle flap are retracted inferolaterally and held in position by fish hooks. It is important to leave a muscle cuff attached to the superior nuchal line. This facilitates tight muscle closure at the end of the procedure, as a water-tight dural closure is not possible because of the necessity of creating an opening for passage of the OA.

An ipsilateral suboccipital craniotomy extending just across the midline and a C1 hemi-laminectomy are performed. The craniotomy may be extended to a far lateral transcondylar approach by exposure of the sigmoid sinus and resection of the posterior medial third of the occipital condyle if access to the PICA origin is required. Any opened mastoid air cells must be thoroughly sealed with bone wax to avoid postoperative cerebrospinal fluid leaks. The dura is opened in the midline at the level of C1 and extended in a curvilinear fashion to the superolateral extent of the exposure. An additional incision is made from the caudal end of the dural incision toward the C1/C2 joint, caudal to the vertebral artery ring. The dural flap is then sutured to the surrounding tissues with 4-0 neurolon traction sutures (Figure 9–7).

Under the operative microscope, the ipsilateral cerebellar tonsil is retracted superolaterally and the caudal loop of the PICA identified. We generally try to avoid using self-retaining brain retractors, as they may actually hinder the surgical exposure. However, if retraction is necessary in spite of extensive arachnoidal lysis, then a tapered self-retaining retractor should be used on the tonsil. The caudal loop is carefully dissected by sharply dividing the arachnoidal bands anchoring the artery to the dorsal surface of the medulla. A rubber dam is then placed deep to the dissected caudal loop of the PICA. The rubber dam may be sutured to the surrounding dura or soft tissue superiorly or inferiorly to carefully elevate the caudal loop, provided that there are no perforating vessels tethering the artery. A micromalleable self-suction device is placed in the vicinity of the anastomosis site to act as a constant drainage path for cerebrospinal fluid and blood.

The OA is then prepared for the anastomosis. A temporary clip is first applied to the proximal end of the artery, which is then cut distally at an appropriate length for the bypass. The occipital artery is notoriously tortuous and amenable to lengthening by “undoing” the various turns and loops. It is, however, critical to let the artery find its own natural contour after dissecting it, to avoid forcing an unnatural kink that might result in graft occlusion. The mean length of OA required for a PICA bypass is 58 mm (range 54 to 60 mm).6 The distal 1 cm of the artery is then stripped of its periadventitial layer. Although this step may be easier to perform while the OA is still in continuity, because of the benefit of countertraction, we prefer to first section the artery to the required length for the bypass to avoid having to repeat the periadventitial stripping if the artery proves too long. There are several ways that the distal OA—or any donor vessel, for that matter—can be prepared, but the three most common techniques are depicted in Figure 9–8, and are as follows:

The latter technique is our favored technique because it increases the cross-sectional area available for the anastomosis and provides redundancy of donor artery wall, thereby minimizing the possibility of stenosis at the site of anastomosis. We firmly believe that end-to-side anastomoses in general fare better if an “elephant foot design” is achieved, with the redundant edges of the donor allowing a flaring of the completed anastomosis and a reduced risk of stenosis or occlusion. Furthermore, the obliquity of the resulting construct allows flow to be preferentially directed toward the proximal PICA and its medullary branches. Using simple high school mathematics (Pythagorean theorem, circumference, and area formulas of circles), we can easily show in Figure 9–8 that the three different techniques result in three different anastomosis circumferences and areas at the suture line. As can be seen from the calculations, compared to the unmodified simple 90-degree, end-to-side technique (Technique 1), the 45-degree cut (Technique 2) “doubles” the cross-sectional area at the suture line, but—better yet—the simple one-sided fishmouthing (Technique 3) “quadruples” it! This, of course, comes at the cost of lengthening the needed arteriotomy in the recipient vessel by a factor of 2 (Technique 3 versus Technique 1). Surgeons often wonder what the arteriotomy length in the recipient vessel needs to be, and Figure 9–8 shows clearly that, for Technique 3, it needs to be twice the “flattened” diameter of the donor, which is about three times its “true” diameter. In summary, when focusing on the geometry of the anastomotic line, Technique 3, compared to Technique 1, results in four times the cross-sectional area and two times the circumference, at the cost of two times the recipient arteriotomy. These geometrical concepts have been clearly simplified and idealized for the purpose of approximation and illustration, and rely on the assumption that the sutured anastomosis will assume a close-to-circular configuration under flow conditions and that the vessel wall edges do not stretch significantly, both very reasonable assumptions. In addition, as a slight departure from Technique 3 as described, we often “round” (small tissue excision) very slightly the sharp corners of the 90-degree fishmouth angles to facilitate running the suture.

Regarding the arteriotomy performed on the side of the recipient artery, we prefer “incising” a straight cut, as opposed to other surgeons who actually “excise” an elliptical segment of vessel wall. Our preference is again based on the desire to maximize the circumference and orthogonal cross-sectional area of the recipient artery at the level of the anastomosis. Without getting into the mathematical details, the excision of an ellipse of tissue measuring “x” millimeters at its widest point will result in a corresponding decrease in the available recipient circumference by “x” millimeters, even though the final circumference—postanastomosis—will be augmented by the amount of the actual donor diameter “d.” For a donor artery of 2 mm in diameter, a recipient artery of 2 mm in diameter, and an excision of x = 1 mm, the “excision” technique, compared to the “incision” technique, results in a relative loss of circumference and area of 25%.

We routinely make detailed intraoperative measurements of blood flow in the donor and recipient arteries during bypass surgeries, using a microvascular ultrasonic flow probe (Charbel Micro-Flow Probe, Transonic Systems, Inc.). The “cut flow,” or the maximal flow carrying capacity of the donor vessel in the absence of downstream resistance, is first measured. The baseline flow in the recipient artery is also measured and gives an indication of the amount of flow that is desirable to replace or augment. Once the anastomosis has been completed, flow through the OA is measured and is known as the “bypass flow” (Figure 9–9). This measurement represents flow through the bypass and provides immediate verification of bypass patency. The ratio of the bypass flow measurement to the initial cut-flow measurement is known as the “cut-flow index” (CFI). Values greater than 0.5 have been found to be a sensitive predictor for acute and long-term postoperative bypass patency, at least in STA-MCA bypasses, primarily done for flow augmentation in ischemic cases.27

Going back to the technical details of the anastomosis, a temporary aneurysm clip is placed on the proximal OA, and its lumen cannulated with a blunt needle and flushed with heparinized saline to remove any blood and debris. Miniature slim-tapered temporary clips are placed on either side of the PICA segment selected for the anastomosis, generally at the telovelotonsillar segment. The patient’s blood pressure is raised by 20% to 30% above baseline during the period of temporary occlusion. An arteriotomy is initially made in the PICA using a beaver blade and subsequently extended using the most delicate and sharpened microscissors. A dull instrument will destroy the delicate vessel wall. A useful alternate technique is to puncture the recipient artery first with a 27-gauge needle—with its tip bent slightly—that is attached to an arterial transducer line, with the double benefit of providing a puncture hole for the beaver blade tip to initiate the arteriotomy, as well as providing a potentially useful measurement of intraluminal recipient artery pressure (particularly in ischemia cases). Heparinized saline is used to irrigate the PICA lumen throughout the course of the suturing, and is delivered through a syringe tipped with a 27-gauge angiocatheter.

The anastomosis may be performed in several ways. We prefer to first anchor the “heel” (the apex of the fishmouth) of the distal end of the OA to one apex of the recipient arteriotomy. On the other hand, some surgeons prefer to first anchor the “toe” (the non-fishmouthed corner) of the distal end of the OA. This latter technique allows the surgeon to extend the fishmouthed corner of the donor vessel if, during suturing, it is realized that the recipient vessel arteriotomy has been made too long and that the fishmouth needs to be extended. An alternative method is to anchor both the heel and toe ends of the distal end of the OA. This latter method is useful, as it prevents errors during suturing related to unequal distances between suture throws, and is probably easier for the “novice” bypass surgeon. Despite this advantage, we prefer not to anchor both ends simultaneously, as this decreases the space available between the donor and recipient vessel walls, and therefore makes visualization of the individual vessel walls while running the suture more difficult.

The anastomosis may be carried out using interrupted stitches or a continuous suture. Although interrupted stitches have a theoretical advantage of allowing future enlargement and maturation of the bypass, we do not believe this to be a real concern unless the vessels are particularly small (<0.8 mm), and we thus generally favor the running suture for its expediency. Furthermore, we suture the back and front wall of the anastomosis separately and thereby allow some enlargement of the bypass over time, if needed. Third, since we utilize Technique 3 in most cases, we believe we are already providing a four-fold increase in cross-sectional area (see Figure 9–8), negating any concerns about stenotic complications. The initial anchoring suture is passed through the wall of the donor OA at the “heel,” from outside the lumen to inside the lumen of the donor (colloquially called the “out-in” stitch), and then the needle is passed through the wall of the recipient PICA at the apex from inside the lumen to outside the lumen (colloquially called the “in-out” stitch) so that the knot ends up lying outside the lumen, naturally. The anastomosis proceeds by first suturing the more difficult (i.e., less accessible) back wall. Care must be taken to avoid grasping the luminal surface of the vessel wall with either the forceps or the needle holder, regardless of the delicacy of the instruments, as the intima may be injured. The forceps should be used to gently anchor the artery as the needle is passed through its wall. The sharp tip of the needle itself is a very useful “hook” that should be used to precisely manipulate, position, and pierce wall edges. The suture is kept loose between stitches, like the spiral binding of a book, until the entire back wall of the anastomosis has been sutured. There is a happy medium for how loose the loops should be: too loose, and they overlap on each other and interfere with each other and the passing of the next bite; too tight, and they are too close to the vessel wall and become harder to grasp later for the tightening phase. At the end of the run, the loops should ideally be equally spaced and lying parallel to each other, without interweaving. The suture is then tightened one loop at a time using two forceps starting from the initial anchoring suture. One forceps tightens the suture while the other handles the next loop, in a sequential manner, ensuring that steady tension is maintained constantly on the previously tightened suture. These forceps have to be relatively new or at least well maintained and preferably diamond-dusted, in order to provide a generous grasping surface area with less risk of slippage or of weakening the suture. The final loop is then tied to itself. Once the back wall has been sutured, the OA is rotated to expose the front wall. A new anchoring suture is placed and the front wall is then sutured in a similar manner. The lumen of the PICA is filled with heparinized saline prior to completion of the final stitch. The anastomosis is carried out using a 10-0 monofilament suture. We prefer a BV 75-3 (Ethicon, Johnson & Johnson, Somerville, NJ) tapered point needle. The needle forms three-eighths of a circle, and therefore minimizes the degree of wrist rotation required during suturing as compared to needles forming one-half of a circle. We do not use a silicon splint in the recipient, as some surgeons do.

Blood flow is restored after completion of the anastomosis by first removing the temporary clips off the recipient artery and re-measuring the recipient flow, prior to unclamping the donor OA. An intraluminal thrombus or stitch-occlusion would be detectable now. Only after this is ruled out and after adequate native recipient flow is re-established, do we then unclip the OA. Slight oozing along the suture line usually stops with a single layer of surgicel (Johnson & Johnson) and gentle pressure with a cottonoid applied over the arteriotomy. Occasionally it may be necessary to place an additional interrupted suture at a specific point if the bleeding does not abate with light pressure. A final assessment of vessel patency is made with the quantitative flow probe. Intraoperative angiography, but more likely and easily, indocyanine green (ICG) video-angiography may also be used.

Following completion of the bypass, any associated VA–PICA or PICA aneurysms should be ideally trapped in the same setting. Alternatively, if there are important perforators arising from the body of the aneurysm, or if access to one end of the aneurysm is not possible surgically or endovascularly, then proximal or distal occlusion may be considered.

The self-suction device and rubber dam are removed. The dura is then closed with or without a dural graft. As previously mentioned, a water-tight dural closure is not possible because of the necessity of creating an opening for passage of the OA. A sheet of DuraGen (Integra Life Sciences, Plainsboro, NJ) may be placed over the dural opening and reinforced with a tissue sealant. A small opening is made in the bone flap to allow an uncompromised passage of the occipital artery. The bone flap is then replaced and fixed with miniplates and screws. The wound is copiously irrigated with antibiotic solution. The muscle is reapproximated to the muscle cuff along the superior nuchal line. A multilayer meticulous closure of the muscle and fascia is crucial to avoid cerebrospinal fluid leaks and may be facilitated by taking the patient partially out of the flexed neck position. The skin is closed in two layers using 2-0 inverted vicryl stitches followed by a 3-0 running locking prolene suture. Sutures are left in place for 2 weeks. The patient is given an aspirin (300 mg) suppository in the operating room prior to extubation, and oral 325 mg daily. Aspirin is continued postoperatively indefinitely.

Postoperative care

Patients are monitored in an intensive care unit for 1 to 2 days (Figure 9–10). Neurological examinations and assessment of bypass graft patency with bedside Doppler are performed every hour. Patients are kept normotensive. Postoperative hypertension should be avoided particularly in patients with vertebrobasilar ischemia, as it may predispose to parenchymal hemorrhage or cerebral edema. Computed tomography angiography (CTA) or preferably a digital subtraction angiography (DSA) is routinely performed 1 to 2 days after surgery. Patients are generally placed on daily aspirin (325 mg) for life.

Complications and Their Avoidance

In addition to general postoperative complications such as wound, urinary, and respiratory infections; deep venous thrombosis; and pulmonary embolism, there are inherent risks to OA-PICA bypass surgery.

Wound healing problems secondary to scalp ischemia from OA diversion can be prevented by performing a wide-based flap sparing the posterior auricular artery. Furthermore, meticulous wound closure and avoidance of skin strangulation by the suture can prevent such problems. Wound bed, subdural, and epidural hematomas may result from leakage at the anastomosis or from muscle ooze related to dissection of the OA. Such complications can be minimized by careful surgical technique and hemostasis at all stages of the procedure. As mentioned above, water-tight dural closure is not possible and thus patients are prone to developing cerebrospinal fluid leaks and pseudomeningoceles. This can be prevented by thoroughly waxing any open mastoid air cells and by performing a careful multilayer closure of the muscles and fascia.

Complications related to the OA include damage during dissection, thrombosis with subsequent emboli formation or occlusion, overstretching, and kinking. Injury during dissection can be prevented by careful preoperative assessment of the course and tortuosity of the OA and meticulous surgical technique, including avoidance of thermal injury during coagulation of side branches. Overstretching or kinking of the artery can be prevented by dissecting the artery proximally to its exit from the occipital groove of the mastoid to ensure that there is adequate length to reach the anastomotic site. Damage or compromise of the OA during closure can be avoided by creating an adequate opening in the bone flap for unhindered passage of the OA as well as careful muscle closure to avoid inadvertent stitching of the artery.

The vertebral artery is at risk for injury during the exposure, and this may be prevented by careful sharp muscle dissection, stripping of the cervical musculature in a subperiosteal fashion, and avoiding the use of Bovie coagulation in the vicinity of the artery near the sulcus arteriosus at C1. In the event of inadvertent vertebral artery injury, it is much easier to repair a sharply made arteriotomy than that produced by a thermal injury. Preoperative knowledge of the anatomy and caliber of the contralateral vertebral artery is also helpful in guiding decisions regarding whether to repair or sacrifice the injured artery. The PICA is also vulnerable to injury at several stages of the procedure. Injuries may occur during the exposure, particularly if the PICA originates extradurally, as it may be mistaken for a muscular branch or for the posterior meningeal artery. The PICA and its perforators can also be injured or occluded intradurally during aneurysm dissection or clipping.

Inadvertent injury to the lower cranial nerves is a major cause of morbidity. Injuries usually result from manipulating the nerve rootlets during dissection and clipping of the aneurysm or inadvertent/deliberate inclusion of nerve rootlets in the clip. These cranial nerves are very sensitive to manipulation, necessitating very gentle retraction and sharp dissection. Lower cranial nerve injury may result in dysphagia, dysarthria, dysphonia, and inadequate airway protection; thus patients should be extubated and started on an oral diet only after a formal evaluation of lower cranial nerve function has been performed. Brainstem injury may result from either excessive retraction or vascular injury. Compromise of the PICA or its perforators may result in postoperative lateral medullary (Wallenberg) syndrome.

Failure of the bypass is usually the result of technical errors or poor patient selection. Technical mistakes may occur during vessel preparation and suturing of the anastomosis and can be minimized by meticulous surgical technique and intraoperative detection using ICG video-angiography and microvascular ultrasonic flow probe measurements. Poor patient selection in ischemic cases as a result of underestimation of collateral circulation will frequently lead to bypass failure due to an insufficient demand for the additional blood flow.

Common alternatives to oa-pica in posterior fossa revascularization

The following alternatives to OA-PICA bypass are performed by using a variety of donor and recipient vessels:

References

1 Yaşargil M.G. Anastomosis between Superficial Temporal Artery and a Branch of the Middle Cerebral Artery. Stuttgart: Georg Thieme Verlag, 1969.

2 Ausman J.I., Lee M.C., Klassen A.C., et al. Stroke: what’s new? Cerebral revascularization. Minn Med. 1976;59(4):223-227.

3 Alvernia J.E., Fraser K., Lanzino G. The occipital artery: a microanatomical study. Neurosurgery. 2006;58(Suppl 1):ONS114-ONS122.

4 Rhoton A.L.Jr. The far-lateral approach and its transcondylar, supracondylar, and paracondylar extensions. Neurosurgery. 2000;47(Suppl 3):S195-S209.

5 Kawashima M., Rhoton A.L.Jr, Tanriover N., et al. Microsurgical anatomy of cerebral revascularization. Part II: posterior circulation. J Neurosurg. 2005;102(1):132-147.

6 Ateş O., Ahmed A.S., Niemann D., et al. The occipital artery for posterior circulation bypass: microsurgical anatomy. Neurosurg Focus. 2008;24(2):E9.

7 Lister J.R., Rhoton A.L.Jr, Matsushima T., et al. Microsurgical anatomy of the posterior inferior cerebellar artery. Neurosurgery. 1982;10(2):170-199.

8 Lesley W.S., Rajab M.H., Case R.S. Double origin of the posterior inferior cerebellar artery: association with intracranial aneurysm on catheter angiography. AJR Am J Roentgenol. 2007;189(4):893-897.

9 Fine A.D., Cardoso A., Rhoton A.L.Jr. Microsurgical anatomy of the extracranial-extradural origin of the posterior inferior cerebellar artery. J Neurosurg. 1999;91:645-652.

10 Ahuja A., Graves V.B., Crosby D.L., et al. Anomalous origin of the posterior inferior cerebellar artery from the internal carotid artery. AJNR Am J Neuroradiol. 1992;13:1625-1626.

11 Lasjaunias P., Vallee B., Person H., et al. The lateral spinal artery of the upper cervical spinal cord. Anatomy, normal variations, and angiographic aspects. J Neurosurg. 1985;63:235-241.

12 Manabe H., Oda N., Ishii M., et al. The posterior inferior cerebellar artery originating from the internal carotid artery, associated with multiple aneurysms. Neuroradiology. 1991;33:513-515.

13 Margolis M.T., Newton T.H. The posterior inferior cerebellar artery. Newton T.H., Potts D.G., editors. Radiology of the Skull and Brain, vol. 2. St. Louis: Mosby, 1974;1719-1775. bk. 2

14 Ogawa T., Fujita H., Inugami A., et al. Anomalous origin of the posterior inferior cerebellar artery from the posterior meningeal artery. AJNR Am J Neuroradiol. 1991;12:186.

15 Tanaka A., Kimura M., Yoshinaga S., et al. Extracranial aneurysm of the posterior inferior cerebellar artery: case report. Neurosurgery. 1993;33:742-744.

16 Haase J., Magnussen I.B., Ogilvy C.S., et al. Evaluating patients with vertebrobasilar transient ischemic attacks. Surg Neurol. 1999;52(4):386-392.

17 Amin-Hanjani S., Du X., Zhao M., et al. Use of quantitative magnetic resonance angiography to stratify stroke risk in symptomatic vertebrobasilar disease. Stroke. 2005;36(6):1140-1145.

18 Howe J.R., Kindt G.W. Cerebral protection during carotid endarterectomy. Stroke. 1974;5:340-343.

19 Michenfelder J.D., Milde J.H., Sundt T.M.Jr. Cerebral protection by barbiturate anesthesia. Use after middle cerebral artery occlusion in Java monkeys. Arch Neurol. 1976;33:345-350.

20 Gross C.E., Adams H.P.Jr, Sokoll M.D., et al. Use of anticoagulants, electroencephalographic monitoring, and barbiturate cerebral protection in carotid endarterectomy. Neurosurgery. 1981;9:1-5.

21 Imparato A.M., Ramirez A., Riles T., et al. Cerebral protection in carotid surgery. Arch Surg. 1982;117:1073-1078.

22 Feustel P.J., Ingvar M.C., Severinghaus J.W. Cerebral oxygen availability and blood flow during middle cerebral artery occlusion: effects of pentobarbital. Stroke. 1981;12:858-863.

23 Amakawa K., Adachi N., Liu K., et al. Effects of pre- and postischemic administration of thiopental on transmitter amino acid release and histologic outcome in gerbils. Anesthesiology. 1996;85:1422-1430.

24 Buggy D.J., Nicol B., Rowbotham D.J., et al. Effects of intravenous anesthetic agents on glutamate release: a role for GABAA receptor-mediated inhibition. Anesthesiology. 2000;92:1067-1073.

25 Bieda M.C., MacIver M.B. Major role for tonic GABAA conductances in anesthetic suppression of intrinsic neuronal excitability. J Neurophysiol. 2004;92:1658-1667.

26 Zhan R.Z., Fujiwara N., Endoh H., et al. Thiopentone inhibits increases in [Ca2+] induced by membrane depolarization, NMDA receptor activation and ischemia in rat hippocampal and cortical slices. Anesthesiology. 1998;89:456-466.

27 Ashley W.W., Amin-Hanjani S., Alaraj A., et al. Flow-assisted surgical cerebral revascularization. Neurosurg Focus. 2008;24(2):E20.