What email address or phone number would you like to use to sign in to Docs.com?
If you already have an account that you use with Office or other Microsoft services, enter it here.
Or sign in with:
Signing in allows you to download and like content, and it provides the authors analytical data about your interactions with their content.
Embed code for: Spinal Cord Blood Supply and Its Surgical Implications
Select a size
Review Article Spinal Cord Blood Supply and Its Surgical Implications Abstract The blood supply to the spine is based on a predictable segmental vascular structure at each spinal level, but true radiculomedullary arteries, which feed the dominant cord supply vessel, the anterior spinal artery, are relatively few and their locations variable. Under pathologic conditions, such as aortic stent grafting, spinal deformity surgery, or spinal tumor resection, sacrifice of a dominant radiculomedullaryvesselmayormaynotleadtospinalcordischemia, depending on dynamic autoregulatory or collateral mechanisms to compensate for its loss. Elucidation of the exact mechanisms for this compensationrequiresfurther studybutwill beaidedby preoperative, intraoperative, and postoperative comparative angiography. Protocols in place at our center and others minimize the risk of spinal cord ischemia during planned radiculomedullary vessel sacrifice. General Structure of Spinal Cord Blood Supply The spinal cord is nourished by a structure of arterial supply that fol- lowsagenerallypredictablepathfrom the great vessels to the parenchyma of the white and gray matter throughout the spinalcord(Figure 1). Segmental vessels originate from the great vessels of the neck, thorax, and abdomen at each segmental level and, with few exceptions, are paired bilaterally. Although the segmental vessel typically divides into an anterior and posterior ramus, the posterior ramus is the dominant vessel and divides further into a muscular branch and a spinal branch. The spinal branch becomes an anterior radicular artery and a posterior radicular artery as it traverses the neuroforamen alongside the segmental nerve root. There are 31 paired radicular arteries, one for each segmental level, but relatively few of them contribute meaningfully to the spinal cord. Most end within the nerve root, dura, or pial plexus. The radicular arteries that do con- tribute to the longitudinally oriented single anterior spinal artery (ASA), and therefore to the anterior two thirdsofthespinalcordparenchyma, are named anterior radiculomedul- lary vessels. Gao et al1 demonstrated the presence of 72 of these vessels across 20 human specimens, for an average of 3.6 per specimen. Con- versely, Martirosyan et al2 cited an average of 10 (range, 2 to 17) throughout the spinal cord. Anterior radiculomedullary arteries are usu- ally not paired at any given level. Their end recipient vessel, the ASA, is mostly in continuity throughout the entire cord3,4 and experiences both anterograde and retrograde flow, depending on the functional demand and location of dominant radiculomedullary feeder vessels. The ASA gives off many central ar- teries, which project and often bifurcate in the sagittal plane, October 2015, Vol 23, No 10 581 Matthew W. Colman, MD Francis J. Hornicek, MD, PhD Joseph H. Schwab, MD From the Department of Orthopaedic Surgery, Massachusetts General Hospital, Boston, MA. J Am Acad Orthop Surg 2015;23: 581-591 http://dx.doi.org/10.5435/ JAAOS-D-14-00219 Copyright 2015 by the American Academy of Orthopaedic Surgeons. Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. heading to the center of the cord matter. These central arteries are most dense in the lumbosacral region, followed by the cervical and thoracic regions. These terminate in centrifugally oriented, inside-to- outside capillary beds, which are five times as dense in gray as in white matter.5 The posterior one third of the cord parenchyma is supplied by paired posteriorspinalarteries(PSAs),which are fed by the posterior radicular ar- teries. This arterial supply is more similartoanarterialplexusthantothe often-envisioned two longitudinal paired arteries.6 Generally, these ar- teries are much smaller than the ASA andarelessconsequentialinregardto ischemic functional deficits when disrupted. As opposed to the terminal branches of the anterior system, the PSAs supply the cord matter in a centripetal pattern. Most anatomic studies have not identified direct anastomotic connections between the anterior and posterior systems,2 whereas others have observed a con- nection in certain anatomic areas, suchasaround the conus medullaris.1 In any case, significant overlap and redundancy does occur between the terminalsupply branches of these two systems within the parenchyma of the spinal cord itself. Structure Unique to the Cervical Spine TheASAinthecervicalspinetakesits origin from two intervertebral ar- teries at the very upper end of the cervical spine, where it is largest, and slowly tapers to a constant diameter for the remainder of the cervical region and into the thoracic region.7 In the cervical spine, segmental ves- sels originate from the vertebral ar- teries in the upper portion and, in the remainder of the cervical spine, from the vertebral arteries and deep cervi- cal, costocervical, or ascending cer- vical branches. The segmentals then give off anterior and posterior radic- ulomedullary arteries, the former of which supply the ASA, as in other regions of the spine. There are typi- cally from one to several dominant unilateral anterior radiculomedullary vessels in the cervical spine.1,2 Although the blood supply to the cervical spinal cord is generally thought to be redundant and well- collateralized, significant variation exists in the number of dominant supply vessels to the ASA, and their number and distribution should be well documented before considering surgical disruption. Structure Unique to the Thoracic Spine In the thoracic spine, the segmental vessels come from the aorta or the subclavian artery and continue on as intercostal arteries. The number of radiculomedullary arteries in the thoracic spine is fewer than in other areas (average, one to four), and they are more spread out.3 There is poor collateral potential in this region, and there is virtually no direct communication between the central (anterior) and peripheral (posterior) systems.4 The ASA of the thora- columbar spine is fed by one or two anterior radiculomedullary vessels, the most dominant of which is termed the artery of Adamkiewicz (AA).8 This feeder vessel of the ASA can be up to 1.3 mm in diameter and occurs on the left side between T9 and T12 in 75% to 80% of cases.9,10 The AA gives off a dominant de- scending branch in a “hairpin” configuration and a much smaller ascending branch as it joins the ASA. The ASA is typically continuous throughout the thoracic spine, but its caliber is significantly attenuated, especially as it approaches the AA.9,11-13 These factors are thought to be responsible for the sensitivity of the thoracic region to ischemic insult Figure 1 Illustration of typical spinal cord blood supply under normal conditions. One radiculomedullary vessel is shown, which supplies the anterior spinal artery through a hairpin loop. Dr. Hornicek or an immediate family member has received research or institutional support from Stryker; has received nonincome support (such as equipment or services), commercially derived honoraria, or other non-research–related funding (such as paid travel) from Biomet; and serves as a board member, owner, officer, or committee member of the American Association of Tissue Banks and the International Society of Limb Salvage. Dr. Schwab or an immediate family member is a member of a speakers’ bureau or has made paid presentations on behalf of Synthes and Stryker Spine; serves as a paid consultant to and serves as a board member, owner, officer, or committee member of Biom’up; and has received nonincome support (such as equipment or services), commercially derived honoraria, or other non-research– related funding (such as paid travel) from Globus Medical and Stryker. Neither Dr. Colman nor any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article. Spinal Cord Blood Supply and Its Surgical Implications 582 Journal of the American Academy of Orthopaedic Surgeons Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. of the anterior part of the spinal cord. Structure Unique to the Lumbar and Sacral Spine The tip of the conus medullaris may be nourished by the descending ter- minationoftheASAdirectly,butthere also exists a highly collateralized anastomotic network at this location termed the anastomotic loop of the conus.7 The lower lumbar and sacral radicular arteries contribute to this highly collateralized anastomotic network rather than to the more cephalad ASA directly. Thus, although ligating a single artery or multiple sacral or lumbar radicular arteries usually does not have ische- mic consequences for the conus or the cephalad spinal cord, these vessels do supply end arterioles to these regions. In fact, particulate matter clogging these end arterioles has resulted in lower extremity paraparesis, such as with inadvertent intravascular par- ticulate steroid injection in the lower lumbar or sacral spine.14 Spinal Cord Ischemia During Pathologic Conditions Spinal cord perfusion pressure is simply a function of the distal aortic perfusion pressure (or mean arterial pressure) less the extrinsic cord pres- sure(orcerebrospinalfluidpressure). The blood supply to the spinal cord may be interrupted iatrogenically during many different types of surgi- cal intervention, including aortic aneurysm repair, tumor resection, and spinal deformity surgery. The mechanism of spinal cord injury may involveglobalhypoperfusion,aswith aortic cross clamping or systemic hypotension; selective ischemia from ligation of dominant segmental ves- sels; or secondary insult, such as with reperfusion injury. The clinical presentation of cord ischemia ranges from complete paraparesis to asymp- tomatic transient abnormalities detectable only by neuromonitoring. The key clinical question driving much of the anatomic research in spinal cord ischemia during surgical interventions is which key vessels must be preserved to maintain spinal cord perfusion pressure and therefore spinal cord blood flow to avoid func- tional deficits. Thoracoabdominal Aortic Aneurysm Repair Although thoracoabdominal aortic aneurysm (TAAA) repair does not have direct relevance for the ortho- paedic surgeon, its related literature is the most robust repository available for reports of ischemic spinal cord damage resulting from vascular inter- ruption. In fact, TAAA repairs have been described as “Russian roulette for the vascular surgeon”15 because of a reported high incidence of spinal cord ischemia, with rates of 16% to 32%, depending on the location and extent of the aneurysm.16,17 The mechanism of cord ischemia is multifactorial but involves the extent of disruption of segmental vessels during aortic mobilization and aneurysm repair. Several studies have highlighted the importance of preserving the domi- nant anterior radiculomedullary artery to the thoracolumbar spine or the AA.18,19 In one study of open TAAA repairs, preoperative angiographic visualization of the AA outside the repair or aortic clamp area resulted in nospinalcordinjury.Inaddition,when the AA was located within the clamped section, microrevascularization of the AA during the open procedure resulted in a 5% rate of cord ischemia com- pared with 50% when revasculariza- tion was unsuccessful.18 However, because of the confound- ing effect of global hypoperfusion during aortic cross clamping, studies involving open TAAA repair may not be the best model for examining the anatomic causes of spinal cord ische- mia. Endovascular techniques, such as thoracic endovascular aortic repair, which selectively excludes specific seg- mental vessels by way of intra-aortic stents, may be a better model for examiningtheeffectofsegmentalvessel sacrifice on cord ischemia. In 71 pa- tients undergoing TAAA endovascular repair, Kawaharada et al20 observed less spinal cord injury in a group in whomtheAAwasnot occluded bythe stent (none) compared with the group in whom the AA was occluded (10%), but the difference was not significant, and the overall event rate was low (3.6%). Schurink et al21 analyzed 13 patients undergoing stent grafting of aneurysms at or below T8, 8 of whom had four or more segmental levels occluded by the stent, and 6 of whom had the segmental artery feeding the AA itself occluded. The authors observed two cases of intraoperative transient motor evoked potential abnormality correctable by raising the mean arterial pressure, and no cases of postoperative paraplegia.21 Although these and other studies indicate that the AA can be safely and permanently occluded at its more proximal segmental supply, it is important to remember that the aortic and microvascular structures of pa- tients with TAAA are chronically degenerativeinnatureandthusarenot normal. Chronic occlusion of seg- mental vessels may allow time for the development of collateral systems that supply the radiculomedullary vessels, thereby making sudden occlusion less likely to cause a catastrophic ischemic event.22,23 A more rigorous model of critical cord blood supply may be found in other clinical scenarios. Spinal Deformity Study of the neurologic effects of seg- mental vessel sacrifice demonstrates Matthew W. Colman, MD, et al October 2015, Vol 23, No 10 583 Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. that segmental vessels must be ligated for exposure during anterior thora- columbar spine exposure and defor- mity correction; early reports raised concerns that sacrifice of segmental vessels supplying dominant radicu- lomedullary vessels can lead to para- paresis. Apel et al24 reported on three cases of congenital kyphoscoliosis in whichcompletelossofsomatosensory- evoked potential (SSEP) tracings was observed within 5 minutes after liga- tion of segmental vessels around the curve apex from T3 to T9. Each case resulted in postoperative paraparesis. This prompted the authors to institute a protocol in 44 patients whereby segmental vessels being considered for ligation were temporarily occluded before permanent ligation. In seven cases, SSEP amplitude changes of .50% were observed within 5 min- utes of temporary occlusion of a seg- mental vessel at the curve apex; all of these returned to baseline within 5 to 19 minutes after removal of the occlusive clamp and preservation of the vessel. No cases of postoperative paraparesis were observed.24 A temporary occlusion protocol before permanent segmental vessel division has been advocated by other authors.25 However, the importance of this has also been brought into question.4,26,27 Bassett et al4 studied 15 patients undergoing anterior spi- nal surgery for kyphoscoliosis. Using angiography, the authors identified 32 total dominant radiculomedul- lary vessels supplying the thora- columbar ASA in their population. Nine of these vessels in eight patients were within the surgical field and were temporarily occluded with no SSEP changes. However, in seven of these eight patients, other radicu- lomedullary vessels outside the sur- gical field were not occluded. The eighth patient had no changes after occlusion of the one dominant seg- mental vessel supplying the lone radiculomedullary vessel, but it was located at L1, an area known for its rich perimedullary collateral supply.4 Another study retrospectively ana- lyzed 1,197 patients who underwent anterior spinal surgery for kyphosco- liosis without using a regular tempo- rary occlusion protocol before dividing segmentals. There were no cases of postoperative paraplegia. However, in one patient, revision sur- gery on the convex side of the defor- mity was undertaken after prior surgery on the concave side. In this patient, the authors did temporarily occlude the segmental vessels in the surgical field, and for one vessel, there were immediate SSEP changes; thus, thevesselwaspreserved.26 In addition tostandardneuroprotectivemeasures, such as avoidance of intraoperative hypotension, these and other authors advise dividing segmentals as far from the neuroforamen as possible, divid- ing segmentals unilaterally on the convex side of the deformity only (which may play a lesser role in vas- cularization of the cord as a result of chronic vessel tension), and consider- ing temporary occlusion or “soft clamping” when a thoracic-only deformity with intraspinal abnormal- ity at the same levels is encountered. Oncologic Resection of Intraspinal Neoplasms Resection of malignant neoplasms of the spinal axis using the technique of total en bloc spondylectomy presents a distinct challenge with regard to maintenance of spinal cord blood supply. Typically, a more extensive, often 360 exposure is required over at least two spinal motion segments to isolate the resection level for uni- vertebral disease. Also, the tumor extent may make it impossible to safely preserve certain vessels, even if there is preoperative concern for their critical role in cord blood supply. Based on an early report that up to four unilateral nerve roots and their accompanying radicular vessels can be sacrificed in dogs without cord ischemia,28 Nambu et al29 and Ueda et al30 undertook extensive experi- mental work investigating the importance of preserving cord- supplying vasculature during total en bloc spondylectomy. In several initial reports, these authors showed only modest decreases (15% to 25%) in spinal cord blood flow without change in motor evoked potentials (MEPs) or spinal cord evoked potentials (SCEPs) after bilateral three-level segmental vessel ligation at T11-T13.29,30 This area in dogs corresponds to the thoracic watershed area in humans where the AA is most often located. In follow- up studies that sought to define the limits of segmental vessel sacrifice in this area without effect on cord health, these authors observed blood flow to the cord dropping to less than half of normal only in dogs that had bilateral five-level (44%) and seven-level (25%) segmental vessel sacrifice; in addition, only in dogs with five or more bilateral levels ligated were any MEP, SCEP, or neurologic examination abnormali- ties detected.31 In a refinement, the group studied segmental sacrifice over multiple levels that specifically included the canine equivalent of the AA based on preoperative angio- gram; only dogs with greater than three-level bilateral segmental vessel ligations including the AA were abnormal with regard to MEP, SCEP, and neurologic examination.32 It is on the basis of these studies that this group currently advocates for safe, bilateral three-level segmental sacri- fice, including the AA, in humans. These authors have retrospectively reported on their experience with 180 cases oftotalen bloc spondylectomyin humans;15hadtheAAinvolvedatthe resection level and underwent target vertebra preoperative embolization and three-level segmental vessel sacri- fice, including the AA. There were no Spinal Cord Blood Supply and Its Surgical Implications 584 Journal of the American Academy of Orthopaedic Surgeons Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. cases of neurologic deterioration or paraplegia after surgery.33 Other au- thors have also suggested that ligating the AA may be safer than previously thought and have demonstrated post- operative development of new collat- erals to the ASA after AA ligation.34,35 The explanations given for the somewhat counterintuitive finding of normal spinal cord function after ligation of the dominant anterior ra- diculomedullary vessel to the ASA are theoretical. Tomita et al36 claim that there must be anastomotic commu- nication between the posterior spinal artery and anterior spinal artery sys- tems through the intercanal, dural, or pial plexus (Figure 2, C). Some au- thors have corroborated this in a lim- ited fashion in anatomic dissections by demonstrating communication at the conus medullaris level,1 whereas othershaveclaimedthatnofunctional communication exists.2 Kawahara et al37 have demonstrated that spinal shortening during total en bloc spondylectomy may increase blood flow to the cord. Other anatomic explanations are that proximal liga- tion of a segmental artery or even of an anterior radiculomedullary artery itself does not disrupt the afferent supply to the ASA because of preex- isting, more distal contributions to the critical anterior radiculomedullary artery (Figure 2, B). These more distal contributions may not be seen on a preoperative angiogram because they are dynamically recruited only after the pathologic condition is initiated. Finally, it is possible that anterograde or retrograde flow dy- namics through the ASA itself from distant cephalad or caudad anterior radiculomedullary vessels compensate fora given anterior radiculomedullary division (Figure 2, D). Monitoring and Timing Following the sacrifice of key spinal cord supply vasculature, the options for intraoperative monitoring of def- icit include SSEPs, continuous elec- tromyographic monitoring (EMG), MEPs, and the intraoperative Stag- nara wake-up test. SSEP monitoring remains the preferred method and is unlikely to be affected by anesthetics, but because it predominantly mon- itors the dorsal columns, it is at best an indirect measure of the lateral corticospinal tracts and the motor neurons, which are supplied by the ASA. False-negative readings can occur, with rates as high as 9%.38,39 Free-running or continuous EMG monitoring, like SSEP monitoring, is a good measure of individual nerve root irritation during surgery and has the benefit of detecting sponta- neous nerve root activity in response to stretch or compression. In con- trast to both, transcranial MEPs directly measure the anterior cord motor pathways. However, they are susceptible to the anesthetic effects of halogenated compounds, nitrous Figure 2 A through D, Illustration of disrupted spinal cord supply following ligation of a key segmental artery (shown in gray) (A) with three possible compensatory mechanisms for reconstitution of the anterior spinal artery (ASA). Without direct supply to the ASA via the typical flow from the segmental artery to the radiculomedullary artery (RMA), the ASA may be reconstituted by collaterals emanating from an adjacent segment radicular artery (B), communication between the posterior spinal arterial system and the ASA system via the pial plexus and areas of spinal cord parenchymal overlap (C), or compensatory dynamic reversal of flow in the ASA itself using supply from distant RMAs or the anastomotic loop of the conus (D). Matthew W. Colman, MD, et al October 2015, Vol 23, No 10 585 Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. oxide, or neuromuscular blockade and are associated with complica- tions from muscle contraction, intraoperative movement, and aber- rant electrical signals affecting the brain or cardiac systems.40 Most authors feel that a combined approach using SSEP and MEP is the safest configuration, with a sensitivity and specificity for detecting cord ischemia of 100% and 91%, respec- tively.41 The Stagnara wake-up test is effective but used less commonly today because a relatively long lag time is required to reverse anesthesia and make an accurate determination of cord function clinically.42 Many authors have commented on the timing of neurophysiologic mon- itoring changes after maneuvers that affect cord blood flow, but the duration is variable. MEP and SEP changeshavebeenreportedtooccur within 5 to 10 minutes after tem- porary or permanent occlusion of key segmental vessels and to return to baseline in reversible scenarios, suchas“soft clamping,” within 5 to 19 minutes of clamp removal.24,43 However, although combined neurophysiologic monitoring is thought to be an adequate proxy for clinical motor function, the correlation is not perfect. Irrevers- ible SSEP/MEP changes may indeed be accompanied by postoperative paraplegia, but not always, and the reverse is true, as well, whereby false-negative readings occur.39 The report by Svensson et al16 of 1,509 TAAA procedures had a 16% clinical postoperative neurologic deficit rate, but only 21% were observed in the immediate postoperative period; 32% were not observed for several days after surgery, and the range of onset was 1 to 21 days.16 Neurologic damage from altered blood supply is complex and may be related to local ischemia, global perfusion dynamics, Figure 3 Midsagittal T2-weighted (A) and axial T1-weighted (B) postcontrast magnetic resonance images of the spine in a 58-year-old woman with a history of small- cell osteosarcoma of the right hemipelvis depicting the tumor extent, with pathologic collapse of the vertebra and extraosseous epidural tumor extension. She presented with a solitary L1 metastasis 3 years after local and systemic treatment of her pelvic disease. Figure 4 A and B, Spinal arterial angiography images with overlying radiograph of the same patient shown in Figure 3, depicting the left-sided dominant thoracolumbar radiculomedullary artery, or artery of Adamkiewicz (black arrow). Filling of the anterior spinal artery is also seen (white arrow). Figure 5 Spinal arterial angiography image with overlying radiograph depicting the comparatively minor contribution of a second radiculomedullary artery located on the left side at T10 in the samepatientshowninFigures3and4. Spinal Cord Blood Supply and Its Surgical Implications 586 Journal of the American Academy of Orthopaedic Surgeons Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. reperfusion injury, or a combination of the three. Protective Maneuvers Standard neuroprotective maneuvers for any surgical procedure involving the spinal cord include maintenance of hemostasis, maintenance of mean arterial pressure, maintenance of end-tissue oxygenation, and avoid- ance of mechanical stressors such as distraction or rapid deformation. However, because spinal cord perfu- sion pressure is simply a function of mean arterial pressure minus the cerebrospinal fluid pressure, other methods have been developed to balance this equation in a favorable way. One example is controlled cerebrospinal fluid drainage via lumbarcannulationoftheintrathecal space. A recent Cochrane review of three large trials investigating cere- brospinalfluiddrainageinthesetting of TAAA repairs found evidence, albeitlimited,thatcerebrospinalfluid drainage improves neurologic out- comeinopenTAAA.44 Although less germane to tumor resection or deformity surgery, another neuro- protective maneuver used during TAAA repair involves distal aortic perfusion during cross-clamping of the aorta to maintain distal aortic perfusion pressure. A Cochrane review found no high-quality ran- domized studies to support this technique but did highlight several observational studies that suggest a neuroprotective effect with improved outcomes.45 In addition, other authors have reported low rates of neurologic complications during TAAA using a protocol that involves reimplantation of segmental vessels previously divided for expo- sure;46-49 however, this protocol has been called into question by others,19 who have reported low rates of neurologic compromise without re- implantation, and it is not commonly Figure 6 Postoperative PA (A) and lateral (B) radiographs depicting the spinal reconstruction in the same patient shown in Figures 3 through 5, following total spondylectomy of L1, with partial spondylectomies of T12 and L2, including posterior segmental fixation, femoral allograft interbody reconstruction, Kaneda- type rod fixation of the anterior vertebrae, and linkage of the posterior and anterior systems via a crosslink device. Figure 7 A, Axial short tau inversion recovery (STIR) magnetic resonance image of the cervical spine in a 30-year-old woman who presented with a 2-year history of left periscapular shoulder pain demonstrating tumor encasement of the left vertebral artery and extensive left neuroforaminal involvement. B, Left-sagittal STIR magnetic resonance image demonstrating the tumor extent across multiple segments in the midcervical spine. Matthew W. Colman, MD, et al October 2015, Vol 23, No 10 587 Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. done at our center. Other potentially clinically useful maneuvers include epidural or systemic cooling;50 also, recent experimental evidence in ani- mals has demonstrated a neuro- protective effect of intravascular injection of a direct free-radical scavenger agent.51 Aside from the limited support available for the use of cerebrospinal fluid drainage, evi- dence for most modalities is lacking. Common Protocols at the Authors’ Center Currently, to minimize the risk of acute or postoperative ischemic injurytothespinalcordduringtumor resection, our group uses several methods in the preoperative, intra- operative, and postoperative stages. Preoperatively, we use angiography to identify critical anterior radicu- lomedullary vessels, and we favor a multidisciplinary approach to the preoperative planning involving tho- racic surgery, vascular surgery, and neurointerventional radiology. We frequently use preoperative emboli- zation of the tumor levels, depending on the presence of visualized collat- eral supply to the cord. Intraoperatively and postoper- atively, we maintain the mean arterial pressure at .90 mm Hg; however, relatively higher pressures must be balanced with intraoperative bleeding risk. We use combined SSEP/MEP monitoring for every case. In high-risk patients, such as those with comorbid risk factors for ischemia (eg, old age, peripheral vascular disease, kidney disease) or with disease-related risk factors (eg, planned sacrifice of major radiculomedullary vessel or verte- bral artery), we occasionally insert prophylactic cerebrospinal fluid drains, which are initially clamped and used simply to monitor cerebro- spinal fluid pressure. Depending on the patient risk profile and any neu- rologic changes that occur intra- operatively or postoperatively, we initiate strict maintenance of cere- brospinal fluid pressures ,10 mm Hg52,53 with continued maintenance of the mean arterial pressure at .90 mm Hg. In the case of a postoperative neurologic change, we also advise the use of axial imaging to rule out other Figure 8 Standard (A) and reversed (B) spinal angiogram images in the same patient shown in Figure 7, demonstrating the lone dominant C5 radiculomedullary artery and its contribution to the anterior spinal artery. Figure 9 Postoperative AP (A) and lateral (B) radiographs of the cervical spine reconstruction, in the same patient shown in Figures 7 and 8. Spinal Cord Blood Supply and Its Surgical Implications 588 Journal of the American Academy of Orthopaedic Surgeons Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. causes of neurologic change, such as postoperative hematoma. Intraoperatively, when dividing segmental vessels or nerve roots containing radicular vessels, we attempttodivideonlythevesselsthat are necessary for exposure and safe resection of the tumor, and we divide them as far laterally as possible. We regularly use a “soft clamping” pro- tocol, in which key segmental or radicular vessels to be divided are temporarily clamped and the neu- rophysiologic monitoring observed for 10 minutes before permanent division. If neuromonitoring changes are observed during this “soft clamping,” the clamp is removed and attempts to spare the segmental vessel in question are undertaken. Often, the vessel in question cannot be safely spared, given the tumor extent, and although attempts at revascularization can be considered, a difficult decision must at times be made regarding paraplegia risk versus oncologic resection margin. This is discussed preoperatively with every patient. Advances in the techni- ques of radiotherapy delivery both in and out of the operating room have helped address close or intentionally positive margins, but the clinical sce- nario remains challenging. Case Examples Thefollowingcaseexamplesillustrate sacrifice of dominant radiculomedul- lary vessels without postoperative paraplegia. Case 1 A neurologically normal 58-year-old woman with a history of small-cell osteosarcoma of the right hemipelvis presented with a solitary L1 metas- tasis 3 years after local and systemic treatmentofherpelvicdisease(Figure 3). Based on a necrosis rate of 50%, her tumor responded poorly to sys- temic chemotherapy. Aside from the spinal disease, there was no evidence ofothermetastasisorlocalrecurrence in the pelvis. She underwent pre- operative radiotherapy with 19.8 Gy to the L1 tumor site. She underwent standard preoperative staging and imaging procedures, including pre- operative thoracolumbar angiogra- phy. Angiography demonstrated a dominant anterior radiculomedul- laryvesseloriginating fromthe leftL1 segmental vessel and supplying the ASA (Figure 4). There was a minor contribution to the ASA via a second radiculomedullary artery on the left at T10 (Figure 5). With a thorough understanding of the neurologic risks, the patient elec- ted to undergo en bloc resection of the isolated metastasis at L1. This was performed as a two-stage pro- cedure, with the first stage involving posterior decompression and partial transpedicular osteotomies at T12 and L2 and posterior segmental instrumentation and fusion with iliac crest autograft from T10 to L4. T12 and L1 nerve roots along with their radicular vessels were ligated bilat- erally, given the extent of the tumor. The second stage was performed 2 days later and consisted of left-sided thoracotomy and en bloc excision of L1, along with partial segments of T12 and L2, with femoral strut allograft reconstruction and Kaneda- typeanteriorinstrumentation(Figure 6). The patient did receive 10 Gy of intraoperative P32 dural plaque radiation, as well as a postoperative boost with protons, for a total of 68.4 Gy. Intraoperative margins were microscopically positive at the dural margin, as planned. Given the L1 nerve root sacrifice, the patient had postoperative hip flexor weak- ness, but there was no evidence of global paraparesis, as might be pre- dicted following sacrifice of the AA at L1. One explanation for this is that the minor T10 radiculomedul- lary artery compensated for the sacrificed dominant AA. Case 2 A neurologically normal 30-year-old woman presented with a 2-year his- toryofleftperiscapularshoulderpain. She had imaging and biopsy studies consistent with a conventional-type chordoma involving the C4 and C5 cervical segments, as well as the left C4-5 neuroforamen (Figure 7). The tumor extent was predominantly left- sided with encasement of the left vertebral artery and was not metas- tatic. She underwent preoperative cerebral angiography, which demon- strated a co-dominant vertebral artery system with adequate contra- lateral reflux through the circle of Willis and the presence of a single dominant radiculomedullary vessel emanating from the left vertebral arteryatC5(Figure8).There wereno identifiable tumor supply vessels for embolization. She underwent pre- operative combinatorial proton- and photon-based radiation to 50.4 Gy. With a thorough understanding of the neurologic risks, the patient elec- ted to undergo en bloc resection of her tumor. This was performed as atwo-stageprocedure.Thefirststage involved posterior decompression, partial posterior osteotomies at C3 and C6, posterior instrumentation fromC2toC7,andligationoftheleft vertebral artery at C3 and C6. In addition, the C4 and C5 nerve roots were ligated, along with the domi- nant radicular vessel. Prior to liga- tion, however, temporary vascular clips were used for 10 minutes to test for neurologic potential changes, withnoeventsobserved.Throughout the remainder of the first-stage oper- ation, there were no significant changesfrombaselineintheSSEPsor MEPs. A wake-up test was not performed. The second stage of the resection involvedcompletionosteotomiesand a hemivertebrectomy of the most caudad portion of C3, the entirety of C4 and C5, and the most cephalad Matthew W. Colman, MD, et al October 2015, Vol 23, No 10 589 Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. portion of C6. The right vertebral artery, including segmental vascula- ture, was not dissected free because it was not involved by tumor. The tumor specimen was removed en bloc, and the defect was re- constructed with free autogenous fibula graft and an anterior plate (Figure 9). The postoperative course was not significant for any sensori- motor neurologic deficit except for weakness (3/5) in the left deltoid and subtle weakness (4/5) in the left biceps, given the oncologically scheduled root sacrifice at C5. By 8 months postoperatively, the motor function in the left biceps and deltoid muscle groups had normalized. Summary The blood supply to the spine is com- plex and may have dynamic autor- egulatory or anastamotic mechanisms tocompensateforlossofcriticalvessels that feed the ASA during surgical pro- cedures. Elucidation of the exact mechanisms for this compensation re- quiresfurtherstudybutwillbeaidedby preoperative, intraoperative, and post- operative comparative angiography. Relatively little study has been dedi- cated to the cervical spine, and this region in particular could be an area of future investigation. References References printed in bold type are those published within the past 5 years. 1. Gao L,WangL,SuB,WangP,YeJ,ShenH: Thevascularsupplytothespinalcordandits relationship to anterior spine surgical approaches. Spine J 2013;13(8):966-973. 2. Martirosyan NL, Feuerstein JS, Theodore N, Cavalcanti DD, Spetzler RF, Preul MC: Blood supply and vascular reactivity of the spinal cord under normal and pathological conditions. J Neurosurg Spine 2011;15(3):238-251. 3. Dommisse GF: The blood supply of the spinal cord: A critical vascular zone in spinal surgery. J Bone Joint Surg Br 1974; 56(2):225-235. 4. Bassett G, Johnson C, Stanley P: Comparison of preoperative selective spinal angiography and somatosensory-evoked potential monitoring with temporary occlusion of segmental vessels during anterior spinal surgery. Spine (Phila Pa 1976) 1996;21(17):1996-1999. 5. Turnbull IM: Microvasculature of the human spinal cord. J Neurosurg 1971;35 (2):141-147. 6. Gillilan LA: The arterial blood supply of the human spinal cord. J Comp Neurol 1958; 110(1):75-103. 7. Tveten L: Spinal cord vascularity: III. The spinal cord arteries in man. Acta Radiol Diagn (Stockh) 1976;17(3):257-273. 8. Adamkiewicz AA: “Die Blutgefässe des menschlichenRückenmarkes: II. Die Gefässe der Rückenmarksoberfläche”. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften. Mathematisch- Naturwissenschaftliche Classe 1882;85: 101-130. 9. Koshino T, Murakami G, Morishita K, Mawatari T, Abe T: Does the Adamkiewicz artery originate from the larger segmental arteries? J Thorac Cardiovasc Surg 1999; 117(5):898-905. 10. Lazorthes G, Gouaze A, Zadeh JO, Santini JJ, Lazorthes Y, Burdin P: Arterial vascularization of the spinal cord: Recent studies of the anastomotic substitution pathways. J Neurosurg 1971;35(3): 253-262. 11. Svensson LG, Klepp P, Hinder RA: Spinal cord anatomy of the baboon: Comparison with man and implications for spinal cord blood flow during thoracic aortic cross- clamping. S Afr J Surg 1986;24(1):32-34. 12. Biglioli P, Spirito R, Roberto M, et al: The anterior spinal artery: The main arterial supply of the human spinal cord. A preliminary anatomic study. J Thorac Cardiovasc Surg 2000;119(2):376-379. 13. Morishita K, Murakami G, Fujisawa Y, et al: Anatomical study of blood supply to the spinal cord. Ann Thorac Surg 2003;76 (6):1967-1971. 14. Kennedy DJ, Dreyfuss P, Aprill CN, Bogduk N: Paraplegia following image- guided transforaminal lumbar spine epidural steroid injection: Two case reports. Pain Med 2009;10(8):1389-1394. 15. Doppman JL: Paraplegia after surgery for thoracoabdominal aneurysms: Russian roulette for the vascular surgeon. Radiology 1993;189(1):27-28. 16. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ: Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg 1993;17(2): 357-368. 17. Crawford ES, Crawford JL, Safi HJ, et al: Thoracoabdominal aortic aneurysms: Preoperative and intraoperative factors determining immediate and long-term results of operations in 605 patients. J Vasc Surg 1986;3(3):389-404. 18. Kieffer E, Richard T, Chiras J, Godet G, Cormier E: Preoperative spinal cord arteriography in aneurysmal disease of the descending thoracic and thoracoabdominal aorta: Preliminary results in 45 patients. Ann Vasc Surg 1989;3(1):34-46. 19. Etz CD, Halstead JC, Spielvogel D, et al: Thoracic and thoracoabdominal aneurysm repair: Is reimplantation of spinal cord arteries a waste of time? Ann Thorac Surg 2006;82(5):1670-1677. 20. Kawaharada N, Morishita K, Kurimoto Y, et al: Spinal cord ischemia after elective endovascular stent-graft repair of the thoracic aorta. Eur J Cardiothorac Surg 2007;31(6):998-1003. 21. Schurink GW, Nijenhuis RJ, Backes WH, et al: Assessment of spinal cord circulation and function in endovascular treatment of thoracic aortic aneurysms. Ann Thorac Surg 2007;83(2):S877-S881. 22. Akasaka J, Sakurai M, Takase K, Kumagai K, Tabayashi K: Another blood supply to Adamkiewicz’s artery. Jpn J Thorac Cardiovasc Surg 2004;52(9): 432-434. 23. Jacobs MJ, de Mol BA, Elenbaas T, et al: Spinal cord blood supply in patients with thoracoabdominal aortic aneurysms. J Vasc Surg 2002;35(1):30-37. 24. Apel DM, Marrero G, King J, Tolo VT, Bassett GS: Avoiding paraplegia during anterior spinal surgery: The role of somatosensory evoked potential monitoring with temporary occlusion of segmental spinal arteries. Spine (Phila Pa 1976) 1991;16(suppl 8):S365-S370. 25. Leung YL, Grevitt M, Henderson L, Smith J: Cord monitoring changes and segmental vessel ligation in the “at risk” cord during anterior spinal deformity surgery. Spine (Phila Pa 1976) 2005;30 (16):1870-1874. 26. Winter RB, Lonstein JE, Denis F, Leonard AS, Garamella JJ: Paraplegia resulting from vessel ligation. Spine (Phila Pa 1976) 1996;21(10):1232-1233. 27. Tsirikos AI, Howitt SP, McMaster MJ: Segmental vessel ligation in patients undergoing surgery for anterior spinal deformity. J Bone Joint Surg Br 2008;90(4): 474-479. 28. Woodard JS, Freeman LW: Ischemia of the spinal cord: An experimental study. J Neurosurg 1956;13(1):63-72. 29. Nambu K, Kawahara N, Kobayashi T, Murakami H, Ueda Y, Tomita K: Interruption of the bilateral segmental arteries at several levels: Influence on Spinal Cord Blood Supply and Its Surgical Implications 590 Journal of the American Academy of Orthopaedic Surgeons Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. vertebral blood flow. Spine (Phila Pa 1976) 2004;29(14):1530-1534. 30. Ueda Y, Kawahara N, Tomita K, Kobayashi T, Murakami H, Nambu K: Influence on spinal cord blood flow and function by interruption of bilateral segmental arteries at up to three levels: Experimental study in dogs. Spine (Phila Pa 1976) 2005;30(20):2239-2243. 31. Fujimaki Y, Kawahara N, Tomita K, Murakami H, Ueda Y: How many ligations of bilateral segmental arteries cause ischemic spinal cord dysfunction? An experimental study using a dog model. Spine (Phila Pa 1976) 2006;31(21): E781-E789. 32. Kato S, Kawahara N, Tomita K, Murakami H, Demura S, Fujimaki Y: Effects on spinal cord blood flow and neurologic function secondary to interruption of bilateral segmental arteries which supply the artery of Adamkiewicz: An experimental study using a dog model. Spine (Phila Pa 1976) 2008;33(14): 1533-1541. 33. Murakami H, Kawahara N, Tomita K, Demura S, Kato S, Yoshioka K: Does interruption of the artery of Adamkiewicz during total en bloc spondylectomy affect neurologic function? Spine (Phila Pa 1976) 2010;35(22):E1187-E1192. 34. Boriani S, Bandiera S, Gasbarrini A: The role of cord vascularization in planning spine oncologic surgery. ArgoSpine 2008; 18(1):40. 35. Toribatake Y: The effect of total en bloc spondylectomy on spinal cord circulation. Nihon Seikeigeka Gakkai Zasshi 1993;67 (11):1070-1080. 36. Tomita K, Kawahara N, Murakami H, Demura S: Total en bloc spondylectomy for spinal tumors: Improvement of the technique and its associated basic background. J Orthop Sci 2006;11(1): 3-12. 37. Kawahara N, Tomita K, Kobayashi T, Abdel-Wanis ME, Murakami H, Akamaru T: Influence of acute shortening on the spinal cord: An experimental study. Spine (Phila Pa 1976) 2005;30(6):613-620. 38. Dawson EG, Sherman JE, Kanim LE, Nuwer MR: Spinal cord monitoring: Results of the Scoliosis Research Society and the European Spinal Deformity Society survey. Spine (Phila Pa 1976) 1991;16 (suppl 8):S361-S364. 39. Deutsch H, Arginteanu M, Manhart K, et al: Somatosensoryevokedpotentialmonitoringin anterior thoracic vertebrectomy. J Neurosurg 2000;92(suppl 2):155-161. 40. Stecker MM: A review of intraoperative monitoring for spinal surgery. Surg Neurol Int 2012;3(suppl 3):S174-S187. 41. Pelosi L, Lamb J, Grevitt M, Mehdian SM, Webb JK, Blumhardt LD: Combined monitoring of motor and somatosensory evoked potentials in orthopaedic spinal surgery. Clin Neurophysiol 2002;113(7): 1082-1091. 42. Owen JH: The application of intraoperative monitoring during surgery for spinal deformity. Spine (Phila Pa 1976) 1999;24 (24):2649-2662. 43. Mizrahi EM, Crawford ES: Somatosensory evoked potentials during reversible spinal cord ischemia in man. Electroencephalogr Clin Neurophysiol 1984;58(2):120-126. 44. Khan SN, Stansby G: Cerebrospinal fluid drainage for thoracic and thoracoabdominal aortic aneurysm surgery. Cochrane Database Syst Rev 2012;10:CD003635. 45. Hsu CC, Kwan GN, van Driel ML, Rophael JA: Distal aortic perfusion during thoracoabdominal aneurysm repair for prevention of paraplegia. Cochrane Database Syst Rev 2012;3: CD008197. 46. Jacobs MJ, Mess W, Mochtar B, Nijenhuis RJ, Statius van Eps RG, Schurink GW: The value of motor evoked potentials in reducing paraplegia during thoracoabdominal aneurysm repair. J Vasc Surg 2006;43(2):239-246. 47. LeMaire SA, Miller CC III, Conklin LD, Schmittling ZC, Coselli JS: Estimating group mortality and paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 2003;75(2): 508-513. 48. Acher CW, Wynn MM, Hoch JR, Popic P, Archibald J, Turnipseed WD: Combined use of cerebral spinal fluid drainage and naloxone reduces the risk of paraplegia in thoracoabdominal aneurysm repair. J Vasc Surg 1994;19(2):236-246. 49. Griepp RB, Ergin MA, Galla JD, et al: Looking for the artery of Adamkiewicz: A quest to minimize paraplegia after operations for aneurysms of the descending thoracic and thoracoabdominal aorta. J Thorac Cardiovasc Surg 1996;112(5): 1202-1213. 50. Cambria RP, Davison JK: Regional hypothermia with epidural cooling for spinal cord protection during thoracoabdominal aneurysm repair. Semin Vasc Surg 2000;13(4):315-324. 51. Herlambang B, Orihashi K, Mizukami T, et al: New method for absolute spinal cord ischemia protection in rabbits. J Vasc Surg 2011;54(4):1109-1116. 52. Svensson LG, Hess KR, D’Agostino RS, et al: Reduction of neurologic injury after high-risk thoracoabdominal aortic operation. Ann Thorac Surg 1998;66(1): 132-138. 53. Coselli JS, LeMaire SA, Köksoy C, Schmittling ZC, Curling PE: Cerebrospinal fluid drainage reduces paraplegia after thoracoabdominal aortic aneurysm repair: Results of a randomized clinical trial. J Vasc Surg 2002;35(4):631-639. Matthew W. Colman, MD, et al October 2015, Vol 23, No 10 591 Copyright ª the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. using a dog model. Spine (Phila Pa 1976) 2008;33(14): 1533-1541. 33. Murakami H, Kawahara N, Tomita K, Demura S, Kato S, Yoshioka K: Does interruption of the artery of Adamkiewicz during total en bloc spondylectomy affect neurologic function? Spine (Phila Pa 1976) 2010;35(22):E1187-E1192. 34. Boriani S, Bandiera S, Gasbarrini A: The role of cord vascularization in planning spine oncologic surgery. ArgoSpine 2008; 18(1):40. 35. Toribatake Y: The effect of total en bloc spondylectomy on spinal cord circulation. Nihon Seikeigeka Gakkai Zasshi 1993;67 (11):1070-1080. 36. Tomita K, Kawahara N, Murakami H, Demura S: Total en bloc spondylectomy for spinal tumors: Improvement of the technique and its associated basic background. J Orthop Sci 2006;11(1): 3-12. 37. Kawahara N, Tomita K, Kobayashi T, Abdel-Wanis ME, Murakami H, Akamaru T: Influence of acute shortening on the spinal cord: An experimental study. Spine (Phila Pa 1976) 2005;30(