Bendok, Bernard R. MD*,‡,§; El Tecle, Najib E. MD*; El Ahmadieh, Tarek Y. MD*; Koht, Antoun MD¶; Gallagher, Thomas A. MD‡; Carroll, Timothy J. PhD‡,‖; Markl, Michael PhD‡,‖; Sabbagha, Randa PhD#; Sabbagha, Asma PhD#; Cella, David PhD**; Nowinski, Cindy MD, PhD**; Dewald, Julius P.A. PhD*,‡‡; Meade, Thomas J. PhD§§; Samson, Duke MD¶¶; Batjer, H. Hunt MD¶¶
Section Editor(s): Bendok, Bernard R. MD; Levy, Elad I. MD
Arteriovenous malformations (AVMs) of the brain are one of neurosurgery's most fascinating and challenging pathologies. Although relatively rare, their potentially devastating effects as a leading cause of hemorrhagic stroke in young individuals have raised their importance to a level that goes beyond what their epidemiology alone justifies. Described by Professor Yaşargil to be as treacherous as Medusa's Head, the ability to manage these lesions successfully has dramatically improved over the past 4 decades. Although advances in microsurgery are the focus of this review, advances in radiosurgery and endovascular techniques have also played important roles in the evolution of AVM management. All 3 modalities play important and potentially synergistic roles. Microsurgery, however, carries the highest immediate cure rate and, when used appropriately, is a very safe and effective therapy for what could otherwise be a potentially devastating disease.
To better understand innovations and future directions in AVM microsurgery, we begin with a brief historical overview. We then analyze and present factors and technologies that have advanced or have the potential to advance AVM surgery. To put these factors into an organized, logical scheme, we have divided the discussion into the following important phases of success with almost any AVM operation: (1) case selection; (2) preoperative planning and management for the surgeon, patient, and team; (4) intraoperative care, technique, and technologies; and (4) postoperative neuro-rehabilitation. Finally, we comment on promising future directions in AVM treatment and on promising innovations and inventions.
HISTORY OF AVM MICROSURGERY
AVMs were initially described in the mid-1800s by Luschka1 (in 1854) and Virchow2 (in 1863). Three decades later, Giordano performed the first surgical exposure of an AVM in 1889.3 In the same year, Péan performed the first successful extirpation of an AVM.3 Multiple neurosurgeons attempted to treat AVMs in the early 20th century. Krause, for example, was the first to attempt a surgical elimination of an AVM by ligating its arterial feeders, without great success, however.4 Multiple techniques that were considered innovations at a certain point in the 20th century might look bizarre to the contemporary neurosurgeon. In 1951, Bilsland5 published a report in which he reviewed the use of preoperative exsanguination to decrease bleeding during intracranial surgery; of particular interest were AVMs because it was thought that exsanguination might help control intraoperative bleeding. In 1954, Brown6 supported this technique by illustrating a series of 133 patients. He stated that “[exsanguination] can be readily controlled more so than when specific vasodilator drugs are used to reduce bleeding.”
Like many other innovations in neurosurgical techniques, the major historical bend in AVM treatment had to wait for the development of microsurgery during the 1970s. In 1976, Yaşargil et al7 published their first series of 10 AVM patients treated with microsurgical techniques with no mortality and minimal morbidity. As microsurgery became increasingly accepted as a management option for AVMs, identifying those AVMs that can be resected safely became a priority. To address this issue, Spetzler and Martin8 published their famous grading scale in 1986. The 1970s and 1980s were marked by a number of other innovations in AVM management. As multiple neurosurgeons attempted to embolize AVMs with different materials, Leksell9,10 started his earliest trials of radiosurgery for these lesions. Both AVM embolization and radiosurgery had to wait for major technical advances before gaining wide acceptance as possible treatment options for AVMs.
Innovations in AVM surgery also occurred as a function of increasingly vivid insight into their biology and hemodynamics. The initial model, described by Luschka and Virchow, of a simple static abnormal connection between the arterial and venous systems progressively evolved into a more elaborate concept. In 1987, Yaşargil was the first to suggest that AVMs are dynamic in nature. In 1996, Mullan10a suggested that AVMs develop after early stages of pregnancy, and in 1997, Lasjaunias10b presented a new description of AVMs as a biological dysfunction at the venocapillary junction, which might occur secondary to angiogenesis after ischemia or hemorrhage. In 2009, Kim et al11 implicated biological factors such as matrix metalloproteinases and antigenic and inflammatory factors in AVM rupture mechanisms.
CASE SELECTION: TOWARD PRECISION MEDICINE
In 2011, the US National Research Council Committee on a Framework for Developing a New Taxonomy of Disease published their report, Toward Precision Medicine: Building a Knowledge Network for Biomedical Research and a New Taxonomy of Disease.12 The report conceived a new multilayer framework built on the basic sciences of discoveries, molecular characterization, electronic health records, and clinical discoveries. The committee agreed that precision medicine might hold the key to targeted treatment of the future, which is equivalent to delivering patient-specific, lesion-specific therapeutics to patients. The concept of precision medicine is particularly attractive for lesions as unique and varied as AVMs and will have to begin with a greater understanding of natural history, hemodynamics, and biology, ideally on an individual level.
AVMS are highly heterogeneous in terms of their angioarchitecture and relationships to surrounding brain structures. Case selection for microsurgery depends on a sophisticated understating of natural history and risk factors and an appraisal of the likelihood of achieving an outcome that is devoid of unacceptable neurological injury. Case selection should increasingly be thought of in terms of precision medicine. Precision medicine moves beyond population science to factor in individual features. The future of natural history assessment will likely go beyond the current thinking of average numbers and factor in individual anatomic, imaging, hemodynamic, and biological factors. An appraisal of surgical risk must factor in surgeon and center experience and a thorough understanding of what can be achieved with all modalities of care individually and in combination. A strong argument can be made for centralization of care for these complex lesions.
Studies focused on understanding natural history of AVMs date back to the early 1980s. In 1983, Graf et al13 reported a series of 191 patients followed up over an average of 3.6 years. In their series, they found that patients with small nidus size (< 3 cm), female patients, and patients with right-sided lesions had a higher risk of initial hemorrhagic presentation. In 1990, Ondra et al14 evaluated a series of 166 untreated patients over a period of 24 years and determined the rate of rebleeding to be 4% per year and the mortality rate to be 1.0% per year. In 2006, Stapf et al15 published their series of 622 AVM patients who were prospectively followed up. They concluded that hemorrhagic presentation, increasing age, deep brain location, and exclusive deep venous drainage are independent predictors for AVM hemorrhage. A more recent study by Hernesniemi et al15 evaluated the risk of hemorrhage in 238 AVM patients. The conclusion they reached was similar to the conclusion of most previous studies: Previously ruptured AVMs and infratentorial and deep AVMs have the highest risk of subsequent hemorrhage. da Costa et al16 in 2009 followed up 678 patients and concluded that brain AVMs presenting with hemorrhage, deep venous drainage, or associated aneurysms have an approximately 2-fold greater likelihood of future hemorrhage. A Randomized Trial of Unruptured Brain AVMs (ARUBA) is a randomized study that aimed to compare intervention and observation in AVM management. ARUBA was halted in May 2013 after an interim analysis showed increased mortality in the intervention group. However, the results of ARUBA have been severely criticized for possible internal and external validity issues. As the available data related to AVM natural history are expanding, the need for decision-making processes accounting for the complexity of AVM hemodynamics and the biology of these lesions is expanding. Some evidence exists to suggest that anatomic factors such as venous stenosis and associated aneurysms may influence natural history17,18; there is also evidence that biological factors such as matrix metalloproteinases and interleukin-6 might be involved in modulating the natural evolution of AVMs.19-21
In addition to understanding the natural history of untreated AVMs, the neurosurgeon must understand the natural history of AVMs treated with other modalities. With regard to radiosurgery, it is important to understand the risk of bleeding during the latency period and the management of treatment failures and complications. With regard to embolization, it is important to understand the implications of various embolization strategies and the potential deleterious effects of partial embolization without a defined end point.22 It is in the best interest of the patient to have a neurosurgeon who understands the nuances of microsurgery, endovascular therapy, and radiosurgery and how these modalities may interplay.
WHO SHOULD HAVE SURGERY: GRADING SCALES
In 1986, Spetzler and Martin8 introduced the first grading system for AVMs to estimate the risk of surgery based on size, location, and venous drainage that defined a 5-point scale. The scale suggests that asymptomatic grade 4 and 5 patients not be treated, but this view is not universally accepted. More recently, Spetzler and Ponce23 introduced a simplified and consolidated 3-tier classification in which AVMs are divided into 3 classes: class A, including the grade I and II Spetzler-Martin AVMs; class B, including the grade III Spetzler-Martin AVMs; and class C, including the grade IV and V Spetzler-Martin AVMs. Lawton et al24 in 2010 introduced another classification scheme that accounts for additional parameters that are likely to affect outcomes of AVM surgery: patient age, hemorrhagic presentation, nidal diffuseness, and deep perforating arterial supply. A full multivariable model and a supplementary model were proposed. The full multivariable model was shown to have the highest predictive accuracy.
Despite the above grading systems, case selection in AVM patients remains challenging. Many variables previously thought to be random or not amenable to measurement such as hemodynamics, shear wall stress, and expression of particular inflammatory markers are suspected to affect AVM natural history and possibly surgical outcomes.25 A highly predictive, large, multivariate model might in the future replace the current grading scales to ensure precision medicine and tailored treatment delivery. Future grading scales may incorporate imaging features. It should be noted that hemorrhage may allow access to AVMs that could not be safely accessed otherwise by creating routes of access.26,27 This was the case in an 18-year-old male patient who presented to our institution with spastic hemiparesis resulting from a third bleed from an AVM. Hemorrhage created a corridor of access to this deep AVM, as shown in Figure 1A. Additionally, in the treatment of this patient, we secured a deep feeder endovascularly with n-butyl-2-cyanoacrylate before taking the patient to surgery (Figure 1B and 1C).
Advances in imaging have progressively allowed a greater understanding of AVM anatomy, AVM hemodynamics, and the nature of surrounding brain tissue. From the well established to the promising, imaging studies have become an invaluable tool in AVM evaluation preoperatively, intraoperatively, and postoperatively.
Although computed tomography (CT), magnetic resonance (MR) imaging (MRI), and invasive angiography have been essential tools in the workup of AVM patients for the past 2 decades, recent advances in each modality have enhanced clinicians' insight into these lesions.
CT and CT angiography are particularly valuable in emergencies. CT can quickly assess the extent and nature of hemorrhage, whereas CT angiography allows rapid assessment of AVM angioarchitecture and sometimes allows the identification of associated aneurysms that could represent the rupture site. CT angiography can also define the relationship between a hematoma and an AVM.
MRI and MR angiography are non--radiation-dependent techniques, which can provide greater insight into the surrounding brain tissue. Fluid-attenuated inversion-recovery sequences, for example, may reflect an inflammatory biological process in an AVM.28 Essig et al28 demonstrated that fluid-attenuated inversion-recovery imaging is superior to T2-weighted fast spin-echo imaging in the assessment of intralesional and perilesional gliosis. Fluid-attenuated inversion-recovery sequences could theoretically be an important guide regarding the extent of surgical resection of a potential epileptogenic focus. Another MRI modality used to assess AVMs is time-resolved imaging of contrast kinetics MR angiography. This modality was shown to be a reliable, noninvasive tool for the evaluation of the feeding arteries, the nidus, and the draining veins of AVMs (see Video, Supplemental Digital Content 1 and 2, http://links.lww.com/NEU/A601, http://links.lww.com/NEU/A602, demonstrating time-resolved MR angiography).29-32 On an experimental basis, MR perfusion is being used to noninvasively assess perfusion of the brain surrounding the AVM.33
Digital subtraction angiography (DSA) remains the diagnostic gold standard for AVMs. The spatial and time resolutions of DSA are superior to those of any other modality. Cone-beam CT angiography and enhancements in 3-dimensional angiography have recently enhanced the value of DSA in AVM workup.34
AVM hemodynamic variables and their impact on AVM natural history and management remain subject to debate and controversy.35 Hemodynamic factors such as blood velocity in different regions of the AVM, the angioarchitecture of the feeding arteries, and the draining veins, as well as the presence of turbulence and flow-induced aneurysms, can possibly affect the natural history and management of AVMs. Little is known about how such variables interact and what their impact is, mostly because acquiring such data in vivo has proven difficult.
Using MRI to assess AVM hemodynamics has emerged as a new and promising tool.34,36,37 It is possible that understanding AVM hemodynamics could allow practitioners to dichotomize AVMs into high- and low-risk lesions.31,34 Figure 2 illustrates the potential of 4-dimensional flow MRI for the comprehensive evaluation of pre-embolization and postembolization cerebral hemodynamics in a case of a 43-year-old male patient with a Spetzler-Martin grade 3 AVM. A video illustration is also provided (see Video, Supplemental Digital Content 3, http://links.lww.com/NEU/A603, illustrating 4-dimensional MRI visualization of an AVM with color-coded blood velocity).
MR INNOVATIONS FOR IMAGING THE SURROUNDING BRAIN
Understanding the surrounding brain harboring an AVM is another key objective of AVM imaging. Diffusion tensor imaging is a technology that applies gaussian diffusion model “ellipsoids” that are sculpted into the principal directions of diffusion in the brain by eigenvalues and eigenvectors of a diffusion “tensor.”38 The tensor itself is populated with diffusion coefficients (measured by MRI in at least 6 noncollinear directions) and resolved to deliver eigenvalue and eigenvector information, the principle directions. Anisotropy determined namely by intact, myelinated axons constrains diffusion along the white matter tracts in the brain, allowing resolution of direction and segmentation of different functional tracts at risk.39-42 Tractography methods, including the very commonly used FACT algorithm, (fiber assignment by continuous tracking), grow streamlines by connecting dominant eigenvectors across many voxels, allowing 3-dimensional (3-D) visualization of many white matter tracts. Although diffusion tensor imaging is effective, it remains a model-based technique, and as a consequence of its reliance on a tensor, it cannot resolve intravoxel crossing fibers.43,44 Tracking algorithms may get stuck where white matter tracts mingle, an important classic example being the intersection of superiorly projecting corticospinal/corticobulbar tracts with anteriorly-posteriorly directed components of the superior longitudinal fasciculus.43,44 Multiple technical improvements in high-angular-resolution white matter imaging, including diffusion spectrum imaging and q-ball methods, are relatively “model free.” They are becoming more clinically feasible and allow resolution of intravoxel crossing fibers and construction of dramatic images of unimpeded white matter tracts and their complex intersections (Figures 3 and 4).45 Such techniques are currently used in the Connectome project, which is aiming to understand and map intricate functional connectivity in the human brain.46,47 The literature regarding applications of tractography in AVM patients is limited to few case reports and small series; therefore, evidence of a positive impact of tractography on outcomes of AVM surgery is pending.48-52 It is likely that better definition and mapping of vital tracts at risk may refine treatment approaches and outcomes.
Whereas tractography allows assessment of important white matter tracts, blood oxygen level--dependent functional MRI (fMRI) enables functional mapping of at-risk cortex in the vicinity of an AVM (Figure 5A and 5B). Pretreatment knowledge of such eloquent areas (motor, language) is particularly relevant in AVM patients, given the potential occurrence of unconventional functional organization around such lesions.53
Because preserving eloquent cortical areas has been shown to have a positive impact on AVM patients' surgical outcomes,54,55 fMRI coupled with intraoperative mapping might allow optimal planning of the surgical exposure and preservation of nearby eloquent cortex (Figure 6).56 Obtaining reliable blood oxygen level--dependent fMRI signal, which itself is derived from endogenous hemodynamic fluctuations in oxyhemoglobin during brain active and resting states,57-59 can be challenging in proximity to the tumultuous hemodynamic environment and potential hemorrhage associated with AVMs.60,61 This invites fMRI methods such as cerebrovascular reserve (response to hypercapnia, breath hold) that provide a baseline assessment of where reliable blood oxygen level--dependent fMRI activations can be seen and where they may be compromised, which prove essential to interpretation. The addition of fMRI, diffusion tensor imaging, diffusion spectrum imaging, and q-ball methods to pretreatment planning may help optimize the treatment strategy and outcome for the AVM patient, particularly as we learn more about brain connectivity and functional networks. A 3-D case illustration in which functional imaging and intraoperative mapping were used was previously published and can be found at56 https://www.youtube.com/watch?v=WKneTeMP5Fo.
FUTURE DIRECTIONS IN IMAGING
Advances in chemistry, biology, and image acquisition have allowed the development of newer imaging modalities that might change the way AVMs are managed. Molecular imaging is a newly developed technique that takes advantage of specific antibodies combined with detectable atoms such as gadolinium or gold.62-64 These techniques have been used to image particular receptors on tumors and could noninvasively detect biological markers in the wall of AVM vessels (Figure 7).65,66
A truly unique renaissance is underway regarding the design and development of the next generation of molecular imaging probes. From small molecular reporters to multimodal labeled nanoparticles, imaging agents have had a significant impact on our fundamental understanding of in vivo biological mechanisms. For example, in the last decade, entirely new classes of MRI agents have been developed and tested that are conditionally activated. Similar to prodrugs, these agents are activated by a number of physiological events, including enzymes, ions such as Ca2+ and Zn2+, pH, and other metabolites. Although translation of these new probes to the clinical setting is yet embryonic, it has begun.
If an appropriate biomarker for AVM is discovered, it will be a relatively short period of time before a highly selective imaging probe is produced. The impact of such an advance on the undiagnosed population with AVM would be revolutionary.
The cognitive and emotional impact of having an AVM is poorly understood. The next wave of AVM research will most likely increasingly focus on these areas. Very little is known regarding how having an AVM can affect quality of life and how treatment may or may not affect quality of life. Another psychological dimension is the need to prepare a patient for treatment. All these issues have created the need to incorporate a cognitive psychology team into the care of AVM patients. Assessing quality of life before and after therapy is an increasingly important dimension in surgery in general and in neurosurgery and cerebrovascular surgery in particular. The Neurology Quality of life Measurement System Neuro-QOL and the NIH Toolbox for the assessment of neurological and behavioral function are recently developed tools that allow clinicians to get a handle on these issues on the clinical and research fronts. Together, these 2 measurement systems provide the clinician or researcher with a broad range of options from which he/she can choose the specific measures that will best facilitate treatment planning, clinical decision making, and achievement of study goals.
MANAGEMENT OF THE ACUTELY RUPTURED AVM
Although AVM resection in the acute phase after rupture is usually not advisable, a number of strategies can help improve outcomes and reduce the risk for rehemorrhage. Craniectomy with or without partial hematoma evacuation may enhance outcomes by reducing intracranial pressure and protecting tissue at risk around the AVM. External ventricular drainage may be needed in certain cases to manage intraventricular hemorrhage and hydrocephalus. Aggressive management of seizures with continuous electroencephalographic monitoring in a neurological intensive care unit is an important consideration in selected patients, particularly those who are comatose. Although we do not favor early embolization for most patients with AVM rupture, strategic embolization to seal a rupture site when one can be seen and accessed safely has the potential to protect the patient from early rehemorrhage until definitive treatment can be undertaken. We recently reported a case in which this strategy was successfully used, followed by intraventricular tissue-type plasminogen activator administration to clear severe intraventricular hemorrhage.67,68 Data support aggressive management of patients who present with ruptured AVMS because excellent outcomes can be achieved in the long term even for patients who present in coma.69
Preoperative management is key to a successful surgery and to optimize recovery. Preoperative management potentially includes strategic use of embolization and adequate psychological and emotional preparation.
Advances in embolization techniques have enhanced microsurgical outcomes in the past 2 decades. The innovations have come in 3 categories: greater insight into embolization strategies; improvements in catheter, wires, and embolic agents; and advances in intraprocedural imaging. Most grade 1 and 2 AVMs can be resected safely without embolization. For large AVMs, a staged approach with the goal of gradual flow reduction may reduce the risk of postoperative hemorrhage and brain edema.70 Embolizing difficult-to-access deep feeders may significantly simplify the operative approach and avoid deep bleeding (Figure 1A–1C). Occluding proximal aneurysms, which pose risk for hemorrhage, can enhance surgical safety (Figure 8).
SURGERY AFTER RADIOSURGERY
Recent published series and our experience suggest that prior radiosurgery facilitates surgical resection.71 It is conceivable that radiosurgery can be used to convert inoperable AVMs to operable ones by eliminating the high-risk components of the nidus.
Dramatic evolutions have occurred on multiple fronts to make AVM surgery safer. These advances have occurred in the areas of neuroanesthesia and neuromonitoring, image guidance, microsurgical instrumentation, and the creation of hybrid operating rooms that allow MRI, CT, and angiography without closing or leaving the operative environment.
Neuroanesthesia and Neuromonitoring
Although not supported by Level 1 evidence, a strong argument can be made for a dedicated subspecialized neuroanesthesia team for AVM surgery. Neuroanesthesia progressively evolved from a service that delivered anesthesia to neurosurgery patients to a discipline in charge of optimizing intraoperative parameters, which can enhance neurosurgical outcome. Neurosurgical anesthesia adds a third dimension to the neurosurgical management. Several challenges are encountered during the management of AVMs such as the management of blood pressure, blood flow, bleeding, and coagulation abnormalities and neurophysiological monitoring. Other challenges include AVMs near eloquent areas that may require awake craniotomy and AVMs during pregnancy. Such challenges require an added ability and vigilance of the anesthesia team. Special anesthetic alterations are needed during neurophysiological monitoring that include minimizing inhalation agents to 0.5 minimum alveolar concentration and limiting muscle relaxants for intubation, if motor evoked potentials are used. These adjustments may lead to 2 important and feared complications: surgical recall and sudden movements, which may require further adjustments of the anesthetics. Special attention to the cardiovascular status should be maintained during the early stages of surgery to avoid potential bleeding as a result of hypertension or ischemia caused by hypotension.
Maintaining close attention to the surgical progress, blood loss, and coagulation parameters helps improve surgical outcome. Blood analyses are done periodically and corrected as needed. If temporary clips are to be used, it is important for communication to occur between the surgeon and the neuroanesthesia team. Oxygen consumption can be decreased by pharmacological means and blood pressure should be raised during this period to enhance collateral circulation.
Recently, adenosine asystole has become a useful tool at our institution during aneurysm surgery complicated by rupture. We have started to use adenosine in select AVM cases when deep feeders are difficult to coagulate and placing AVM clips proves difficult. Close cooperation between the neurosurgeon and the anesthesiologist is very important when adenosine is used.
In addition to the above points, the neuroanesthesiologist can contribute to the optimization and management of neurophysiological monitoring, which is used to alert the team about potential neurological deficits. Both amplitude and latency of sensory and motor evoked potentials can be helpful during AVM resection, especially if provocative testing is used. Optimization of anesthesia to obtain the best recordings while maintaining minimal movements is essential. In large, when dealing with AVMs that require a big surgical flap, the stimulation and recording needles may not be placed in the proper positions, which results in less-than-optimal and doubtful responses. Adjustment can be done with sterile needles or stimulating leads that are placed by the surgeon over the motor cortex, thus allowing continuous monitoring. Evoked potential changes can be the result of technical, physiological, pharmacological, positional, or surgical causes. Identifying the cause is an important step toward management. The anesthesiologist can play an important role at this stage by maintaining higher blood pressure, decreasing oxygen consumption, and increasing oxygen supply while the differential diagnosis is being established or if surgical insult already occurred. Controlling blood pressure is important to minimize the dangers of the hyperemia, which can lead to possible ischemia or bleeding to the surrounding brain tissue.
SURGICAL PLANNING AND IMAGE GUIDANCE
Image Guidance and Designing the Skin and Bone Flap
Designing the skin and bone flaps appropriately is an important step in microsurgical removal of AVMs. In our experience, large craniotomies are preferred for AVM excision, especially when cortical mapping is needed. Large craniotomies allow better control of the surrounding vasculature and can avoid burr hole placement over an AVM. Video 3 (see Video, Supplemental Digital Content 4, http://links.lww.com/NEU/A604, showing AVM resection via a large craniotomy) illustrates this concept. Image guidance continues to evolve and allows accurate design of bone flaps. As imaging modalities have advanced, so has image guidance. DSA and tractography can now be incorporated into an image guidance system, giving the surgeon greater insight during surgery. Incorporation of image guidance into 3-D viewing systems has recently been introduced (Figure 9A, and 10, 9B). This has the potential to dramatically enhance a surgeon's understanding of the complex anatomy of an AVM before and during surgery.72
ADVANCES IN SURGICAL MICROSCOPY AND FIELD ILLUMINATION
Since Yaşargil pioneered the use of the operating microscope in the 1960s, microscopes have improved dramatically with regard to optics, illumination, and ease of use. Perhaps the most significant recent advance has been the seamless incorporation of indocyanine green angiography. Killory et al73 published their review of 10 consecutive AVM surgeries during which they used indocyanine green angiography. Intraoperative indocyanine green angiography was able to detect 1 of 2 residual AVMs confirmed by intraoperative DSA. We routinely use indocyanine green angiography at various stages of resection to assess the status of arteriovenous shunting and the overall transit time through the AVM. This advance is likely the tip of the iceberg. It is possible that contrast imaging techniques will be used to give the surgeon dynamic information on brain perfusion and arteriovenous shunting through the microscope. The incorporation of image guidance and 3-D imaging will also likely merge with the microscope, giving the surgeon a heads-up display of important anatomic information. This has the potential to dramatically facilitate the surgeon's understanding of the complex 3-D architecture of an AVM. The ability of a surgeon to preserve en passant vessels could be greatly aided by such an advance. Additionally, micro-Doppler and flow probe technologies may allow the surgeon to gauge hemodynamic changes throughout the operation as well.74
ADVANCES IN MICROINSTRUMENTATION
Over the past decade, multiple enhancements in microinstrumentation have been geared at AVM surgery. One challenge with resecting AVMs is having the bipolar tips become adherent to fragile AVM tissue. Recent advances in coating technologies have made this less of an issue.75 Adding illumination to bipolar instrumentation has also been recently described. Additionally, AVM clips have been enhanced over the past decade.
HYBRID OPERATING ROOMS
The physical separation of high-end radiology equipment from the operating room has always been a barrier to image-guided surgery. This has been particularly true for surgeries such as AVM resection in which postoperative imaging can dictate a return to the operating room. This led to a growing need to integrate CT, MR, and DSA technologies into the operating theater, which has culminated in the creation of hybrid operating rooms. One such operating room suite is the IMRIS, which was recently developed by J.M. Keckler Medical Co (Figure 9A and 9B). Such operative rooms will enhance surgical precision and safety and reduce returns to the operating room.
ENHANCING SURGICAL EXPOSURES
Tailored cranial base approaches and endoscopic techniques may play a role in the future of AVM management.76 The orbitozygomatic approach, various transtemporal approaches, and the far lateral approach may enhance the safety of exposure of deep-seated lesions.77,78 Less reliance on retractors may also enhance outcomes.79 It is important here to distinguish brain compression by a retractor and the strategic use of a retractor to hold brain, which is already relaxed out of the way.
Improved Microsurgical Skills: The Role of Simulation
Despite all advances in tools and techniques, outcomes of AVM surgeries remain operator dependent. Simulation, if integrated into a curriculum with strong mentorship, may be one mechanism by which skills can be improved.80 The complexity of AVMs has made simulation difficult, but advances in 3-D printing and holographic technologies with haptic feedback may overcome challenges. Holography, which is the 3-D projection of the AVM into space, might help the surgeon understand the anatomic and hemodynamic characteristics of the AVM in all directions. Holographic projection is still in its very early developmental stage; medical applications remain restricted only to few case reports, and applications even outside medicine remain scarce.81,82 Currently, 3-D printing is used to print 3-D biological scaffolds.83 Other applications, although rare, include printing of DICOM (digital imaging and communications in medicine) files.84 Despite being a very new technology, 3-D printing might in the future provide the surgeon with the ability to practice a surgical operation as many times as deemed necessary before doing it on a patient.
Specialized Neurointensive Care
Recent data suggest that neurosurgical patient outcomes can be enhanced by specialized neurological intensive care units where physicians and nurses have significant experience with neurosurgical pathologies.85 Blood pressure control is important in the postoperative period after AVM resection because surrounding brain tissue may be prone to hemorrhage until autoregulation has been restored. Postresection hyperemia is a unique complication caused by loss of autoregulation, and care should be taken to meticulously ensure hemostasis during surgery. Additionally, it is important to have a comprehensive strategy for deep venous thrombosis prophylaxis and medical complication avoidance. To reduce radiation exposure, we have recently shifted to using MRI preferentially for postoperative monitoring rather than CT. A postoperative angiogram is essential to rule out any residual AVM, which could be a risk for rehemorrhage.
The Need for Early Science Underpinned Neuro-rehabilitation
Most deficits which occur after AVM surgery are likely to improve over time.86 Aggressive science-based rehabilitation is likely to enhance the odds of an excellent outcome.87 Recent innovations in rehabilitation include more sophisticated quantitative assessments of sensorimotor dysfunction using impairment based neuro-robotics.89 Robotic and mechatronic technologies are also increasingly being used to customize therapies (Figure 11) and precisely measure improvement.90-93 For example, losses of corticofugal projections, as can occur following AVM surgery and quantified with DTI, results in weakness and losses of independent joint control.93 This causes increases in obligatory coupling between shoulder abduction and elbow, wrist and finger flexion as the limb's weight is increased from weightlessness to beyond limb weight. Both weakness and the loss of independent joint control can be measured and significantly reduced following interventions using robotic and mechatronic devices.91,92
AVMs are an important cause of hemorrhagic stroke in young individuals. Surgery remains the most versatile and curative option for many patients with AVMs. AVM microsurgery is a vibrant field that has seen exciting advances in the past 2 decades, with many advances yet to come. Continued innovations in the science of precision medicine, imaging, intraoperative technologies, simulation, neuroanesthesia, rehabilitation, and the science of outcomes assessment will likely enhance outcomes and refine indications. Concurrent advances in radiosurgery and endovascular techniques are creating new opportunities to treat more AVMs with microsurgery with lower morbidity and better outcomes.
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
1. von Luschka H. Die Adergeflechte des menschlichen Gehirnes: eine Monographie. G. Reimer, 1855.
2. Virchow R. Handbuch der speciellen Pathologie und Therapie. Vol. 6, No. 2. Enke, 1856.
3. Yaşargil MG. Microneurosurgery: Thieme Classics. New York, NY: Thieme Medical Publishers; 1987.
4. Olivecrona H, Riives J. Arteriovenous aneurysms of the brain, their diagnosis and treatment. Arch Neurol Psychiatry. 1948;59(5):567–602.
5. Bilsland WL. Controlled hypotension by arteriotomy in intracranial surgery. Anaesthesia. 1951;6(1):20–25.
6. Brown AS. Pre-operative exsanguination in intracranial surgery. Anaesthesia. 1954;9(1):17–20.
7. Yaşargil MG, Jain KK, Antic J, Laciga R. Arteriovenous malformations of the splenium of the corpus callosum: microsurgical treatment. Surg Neurol. 1976;5(1):5–14.
8. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65(4):476–483.
9. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry. 1983;46(9):797–803.
10. Steiner L, Leksell L, Forster DM, Greitz T, Backlund EO. Stereotactic radiosurgery in intracranial arterio-venous malformations. Acta Neurochir (Wien). 1974;(suppl 21):195–209.
10a. Mullan S, Mojtahedi S, Johnson DL, Macdonald RL. Embryological basis of some aspects of cerebral vascular fistulas and malformations. J Neurosurg. 1996;85(1):1–8.
10b. Lasjaunias P. A revised concept of the congenital nature of cerebral arteriovenous malformations. Interv Neuroradiol. 1997;3(4):275–281.
11. Kim H, Pawlikowska L, Chen Y, Su H, Yang GY, Young WL. Brain arteriovenous malformation biology relevant to hemorrhage and implication for therapeutic development. Stroke. 2009;40(3 suppl):S95–S97.
12. Committee on a Framework for Development a New Taxonomy of Disease; National Research Council. Toward Precision Medicine: Building a Knowledge Network for Biomedical Research and a New Taxonomy of Disease. Washington, DC: The National Academies Press; 2011.
13. Graf CJ, Perret GE, Torner JC. Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg. 1983;58(3):331–337.
14. Ondra SL, Troupp H, George ED, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg. 1990;73(3):387–391.
15. Hernesniemi JA, Dashti R, Juvela S, Vaart K, Niemela M, Laakso A. Natural history of brain arteriovenous malformations: a long-term follow-up study of risk of hemorrhage in 238 patients. Neurosurgery. 2008;63(5):823–829; discussion 829-831.
16. da Costa L, Wallace MC, Ter Brugge KG, O'Kelly C, Willinsky RA, Tymianski M. The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke. 2009;40(1):100–105.
17. Hademenos GJ, Massoud TF. Risk of intracranial arteriovenous malformation rupture due to venous drainage impairment: a theoretical analysis. Stroke. 1996;27(6):1072–1083.
18. Viñuela F, Nombela L, Roach MR, Fox AJ, Pelz DM. Stenotic and occlusive disease of the venous drainage system of deep brain AVM's. J Neurosurg. 1985;63(2):180–184.
19. Gaetani P, Tartara F, Messina A, Tancioni F, Schiavo R, Grazioli V. Metalloproteases and intracranial vascular lesions. Neurol Res. 1999;21(4):385–390.
20. Chen Y, Pawlikowska L, Yao JS, et al.. Interleukin-6 involvement in brain arteriovenous malformations. Ann Neurol. 2006;59(1):72–80.
21. Starke RM, Komotar RJ, Hwang BY, et al.. Systemic expression of matrix metalloproteinase-9 in patients with cerebral arteriovenous malformations. Neurosurgery. 2010;66(2):343–348.
22. Loh Y, Duckwiler GR, Onyx Trial I. A prospective, multicenter, randomized trial of the Onyx liquid embolic system and N-butyl cyanoacrylate embolization of cerebral arteriovenous malformations: clinical article. J Neurosurg. 2010;113(4):733–741.
23. Spetzler RF, Ponce FA. A 3-tier classification of cerebral arteriovenous malformations: clinical article. J Neurosurg. 2011;114(3):842–849.
24. Lawton MT, Kim H, McCulloch CE, Mikhak B, Young WL. A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery. 2010;66(4):702–713.
25. Zhao Y, Li P, Fan W, et al.. The rs522616 polymorphism in the matrix metalloproteinase-3 (MMP-3) gene is associated with sporadic brain arteriovenous malformation in a Chinese population. J Clin Neurosci. 2010;17(12):1568–1572.
26. Arnaout OM, Gross BA, Eddleman CS, Bendok BR, Getch CC, Batjer HH. Posterior fossa arteriovenous malformations. Neurosurg Focus. 2009;26(5):E12.
27. Bendok BR, Getch CC, Frederiksen J, Batjer HH. Resection of a large arteriovenous fistula of the brain using low-flow deep hypothermic cardiopulmonary bypass: technical case report. Neurosurgery. 1999;44(4):888–890; discussion 890-891.
28. Essig M, Wenz F, Schoenberg SO, Debus J, Knopp MV, Van Kaick G. Arteriovenous malformations: assessment of gliotic and ischemic changes with fluid-attenuated inversion-recovery MRI. Invest Radiol. 2000;35(11):689–694.
29. Razek AA, Gaballa G, Megahed AS, Elmogy E. Time resolved imaging of contrast kinetics (TRICKS) MR angiography of arteriovenous malformations of head and neck. Eur J Radiol. 2013;82(11):1885–1891.
30. Eddleman CS, Jeong H, Cashen TA, et al.. Advanced noninvasive imaging of spinal vascular malformations. Neurosurg Focus. 2009;26(1):E9.
31. Eddleman CS, Jeong HJ, Hurley MC, et al.. 4D radial acquisition contrast-enhanced MR angiography and intracranial arteriovenous malformations: quickly approaching digital subtraction angiography. Stroke. 2009;40(8):2749–2753.
32. Cashen TA, Jeong H, Shah MK, et al.. 4D radial contrast-enhanced MR angiography with sliding subtraction. Magn Reson Med. 2007;58(5):962–972.
33. Guo WY, Wu YT, Wu HM, et al.. Toward normal perfusion after radiosurgery: perfusion MR imaging with independent component analysis of brain arteriovenous malformations. AJNR Am J Neuroradiol. 2004;25(10):1636–1644.
34. Ansari SA, Schnell S, Carroll T, et al.. Intracranial 4D flow MRI: toward individualized assessment of arteriovenous malformation hemodynamics and treatment-induced changes. AJNR Am J Neuroradiol. 2013;34(10):1922–1928.
35. Illies T, Forkert ND, Saering D, et al.. Persistent hemodynamic changes in ruptured brain arteriovenous malformations. Stroke. 2012;43(11):2910–2915.
36. Hope MD, Purcell DD, Hope TA, et al.. Complete intracranial arterial and venous blood flow evaluation with 4D flow MR imaging. AJNR Am J Neuroradiol. 2009;30(2):362–366.
37. Chang W, Loecher MW, Wu Y, et al.. Hemodynamic changes in patients with arteriovenous malformations assessed using high-resolution 3D radial phase-contrast MR angiography. AJNR Am J Neuroradiol. 2012;33(8):1565–1572.
38. Basser PJ, Mattiello J, LeBihan D. Estimation of the effective self-diffusion tensor from the NMR spin echo. J Magn Reson B. 1994;103(3):247–254.
39. Moseley ME, Cohen Y, Kucharczyk J, et al.. Diffusion-weighted MR imaging of anisotropic water diffusion in cat central nervous system. Radiology. 1990;176(2):439–445.
40. Moseley ME, Kucharczyk J, Asgari HS, Norman D. Anisotropy in diffusion-weighted MRI. Magn Reson Med. 1991;19(2):321–326.
41. Beaulieu C, Allen PS. Water diffusion in the giant axon of the squid: implications for diffusion-weighted MRI of the nervous system. Magn Reson Med. 1994;32(5):579–583.
42. Beaulieu C, Allen PS. Determinants of anisotropic water diffusion in nerves. Magn Reson Med. 1994;31(4):394–400.
43. Mori S, Crain BJ, Chacko VP, van Zijl PC. Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol. 1999;45(2):265–269.
44. Conturo TE, Lori NF, Cull TS, et al.. Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci U S A. 1999;96(18):10422–10427.
45. Tuch DS, Reese TG, Wiegell MR, Wedeen VJ. Diffusion MRI of complex neural architecture. Neuron. 2003;40(5):885–895.
46. Toga AW, Clark KA, Thompson PM, Shattuck DW, Van Horn JD. Mapping the human connectome. Neurosurgery. 2012;71(1):1–5.
47. Wedeen VJ, Wang RP, Schmahmann JD, et al.. Diffusion spectrum magnetic resonance imaging (DSI) tractography of crossing fibers. Neuroimage. 2008;41(4):1267–1277.
48. Itoh D, Aoki S, Maruyama K, et al.. Corticospinal tracts by diffusion tensor tractography in patients with arteriovenous malformations. J Comput Assist Tomogr. 2006;30(4):618–623.
49. Yamada K, Kizu O, Ito H, et al.. Tractography for arteriovenous malformations near the sensorimotor cortices. AJNR Am J Neuroradiol. 2005;26(3):598–602.
50. Okada T, Miki Y, Kikuta K, et al.. Diffusion tensor fiber tractography for arteriovenous malformations: quantitative analyses to evaluate the corticospinal tract and optic radiation. AJNR Am J Neuroradiol. 2007;28(6):1107–1113.
51. Tanei T, Takebayashi S, Nishihata T, Nakahara N, Wakabayashi T. Removal of an arteriovenous malformation near the pyramidal tract using the neuronavigation system: a case report [in Japanese]. No Shinkei Geka. 2010;38(8):745–750.
52. Koga T, Shin M, Maruyama K, et al.. Integration of corticospinal tractography reduces motor complications after radiosurgery. Int J Radiat Oncol Biol Phys. 2012;83(1):129–133.
53. Alkadhi H, Kollias SS, Crelier GR, Golay X, Hepp-Reymond MC, Valavanis A. Plasticity of the human motor cortex in patients with arteriovenous malformations: a functional MR imaging study. AJNR Am J Neuroradiol. 2000;21(8):1423–1433.
54. Stapleton S, Kiriakopoulos E, Mikulis D, et al.. Combined utility of functional MRI, cortical mapping, and frameless stereotaxy in the resection of lesions in eloquent areas of brain in children. Pediatr Neurosurg. 1997;26(2):68–82.
55. Latchaw RE, Hu X, Ugurbil K, Hall WA, Madison MT, Heros RC. Functional magnetic resonance imaging as a management tool for cerebral arteriovenous malformations. Neurosurgery. 1995;37(4):619–626.
56. El Ahmadieh TY, Aoun SG, Adel JG, Rosenow JM, Koht A, Bendok BR. Microsurgical treatment of a premotor arteriovenous malformation: 3-dimensional illustration. Neurosurgery. 2013;72(1 suppl operative):1.
57. Ogawa S, Tank DW, Menon R, et al.. Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci U S A. 1992;89(13):5951–5955.
58. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A. 1990;87(24):9868–9872.
59. Ogawa S, Lee TM, Nayak AS, Glynn P. Oxygenation-sensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn Reson Med. 1990;14(1):68–78.
60. Lehéricy S, Biondi A, Sourour N, et al.. Arteriovenous brain malformations: is functional MR imaging reliable for studying language reorganization in patients? Initial observations. Radiology. 2002;223(3):672–682.
61. Thickbroom GW, Byrnes ML, Morris IT, Fallon MJ, Knuckey NW, Mastaglia FL. Functional MRI near vascular anomalies: comparison of cavernoma and arteriovenous malformation. J Clin Neurosci. 2004;11(8):845–848.
62. Major JL, Meade TJ. Bioresponsive, cell-penetrating, and multimeric MR contrast agents. Acc Chem Res. 2009;42(7):893–903.
63. Schultz-Sikma EA, Meade TJ. Supramolecular chemistry in biological imaging in vivo. In: Steed JW, Gale PA, ed. Supramolecular Chemistry: From Molecules to Nanomaterials: New Jersey: John Wiley & sons, Ltd; 2012.
64. Thibon A, Pierre VC. Principles of responsive lanthanide-based luminescent probes for cellular imaging. Analytic Bioanalytic Chem. 2009;394(1):107–120
65. Matosziuk LM, Leibowitz JH, Heffern MC, Macrenaris KW, Ratner MA, Meade TJ. Structural optimization of Zn(II)-activated magnetic resonance imaging probes. Inorg Chem. 2013;52(21):12250–12261.
66. Strauch RC, Mastarone DJ, Sukerkar PA, Song Y, Ipsaro JJ, Meade TJ. Reporter protein-targeted probes for magnetic resonance imaging. J Am Chem Soc. 2011;133(41):16346–16349.
67. Pollock GA, Shaibani A, Awad I, Batjer HH, Bendok BR. Intraventricular hemorrhage secondary to intranidal aneurysm rupture-successful management by arteriovenous malformation embolization followed by intraventricular tissue plasminogen activator: case report. Neurosurgery. 2011;68(2):E581–E586; discussion E586.
68. Morgan T, Awad I, Keyl P, Lane K, Hanley D. Preliminary report of the clot lysis evaluating accelerated resolution of intraventricular hemorrhage (CLEAR-IVH) clinical trial. Acta Neurochir Suppl. 2008;105:217–220.
69. Lawton MT, Du R, Tran MN, et al.. Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations. Neurosurgery. 2005;56(3):485–493; discussion 485-493.
70. Spetzler RF, Wilson CB, Weinstein P, Mehdorn M, Townsend J, Telles D. Normal perfusion pressure breakthrough theory. Clin Neurosurg. 1978;25:651–672.
71. Sanchez-Mejia RO, McDermott MW, Tan J, Kim H, Young WL, Lawton MT. Radiosurgery facilitates resection of brain arteriovenous malformations and reduces surgical morbidity. Neurosurgery. 2009;64(2):231–238; discussion 238-240.
72. Kockro RA, Reisch R, Serra L, Goh LC, Lee E, Stadie AT. Image-guided neurosurgery with 3-dimensional multimodal imaging data on a stereoscopic monitor. Neurosurgery. 2013;72 suppl 1:78–88.
73. Killory BD, Nakaji P, Gonzales LF, Ponce FA, Wait SD, Spetzler RF. Prospective evaluation of surgical microscope-integrated intraoperative near-infrared indocyanine green angiography during cerebral arteriovenous malformation surgery. Neurosurgery. 2009;65(3):456–462; discussion 462.
74. Amin-Hanjani S, Alaraj A, Charbel FT. Flow replacement bypass for aneurysms: decision-making using intraoperative blood flow measurements. Acta Neurochir (Wien). 2010;152(6):1021–1032; discussion 1032.
75. Vellimana AK, Sciubba DM, Noggle JC, Jallo GI. Current technological advances of bipolar coagulation. Neurosurgery. 2009;64(3 suppl):ons11–ons18; discussion ons19.
76. Kassam AB, Thomas AJ, Zimmer LA, et al.. Expanded endonasal approach: a fully endoscopic completely transnasal resection of a skull base arteriovenous malformation. Childs Nerv Syst. 2007;23(5):491–498.
77. Lopez Flores G, Fernandez-Melo R, Guerra-Figueredo E, et al.. Direct approach to carotid cavernosa fistula type of arterio venous dural malformation: case report and review of the literature [in Spanish]. Rev Neurol. 2002;34(3):204–207.
78. Pannu Y, Shownkeen H, Nockels RP, Origitano TC. Obliteration of a tentorial dural arteriovenous fistula causing spinal cord myelopathy using the cranio-orbito zygomatic approach. Surg Neurol. 2004;62(5):463–467; discussion 467.
79. Spetzler RF, Sanai N. The quiet revolution: retractorless surgery for complex vascular and skull base lesions. J Neurosurg. 2012;116(2):291–300.
80. Ganju A, Aoun SG, Daou MR, et al.. The role of simulation in neurosurgical education: a survey of 99 United States neurosurgery program directors [published online ahead of print November 24, 2012]. World Neurosurg. doi:10.1016/j.wneu.2012.11.066.
81. Mischkowski RA, Bongartz J, Giel D, Frey S, Thelen A, Hering P. Holographic face models as planning tool in maxillofacial surgery. Int J Comput Dent. 2004;7(4):339–345.
82. Blanche PA, Bablumian A, Voorakaranam R, et al.. Holographic three-dimensional telepresence using large-area photorefractive polymer. Nature. 2010;468(7320):80–83.
83. Lu Y, Chen S. Projection printing of 3-dimensional tissue scaffolds. Methods Mol Biol. 2012;868:289–302.
84. Tam MD, Laycock SD, Bell D, Chojnowski A. 3-D printout of a DICOM file to aid surgical planning in a 6 year old patient with a large scapular osteochondroma complicating congenital diaphyseal aclasia. J Radiol Case Rep. 2012;6(1):31–37.
85. Diringer MN, Edwards DF. Admission to a neurologic/neurosurgical intensive care unit is associated with reduced mortality rate after intracerebral hemorrhage. Crit Care Med. 2001;29(3):635–640.
86. Heros RC. Multimodality treatment of cerebral arteriovenous malformations: modern treatment of cerebral arteriovenous malformations [published online ahead of print March 16, 2013]. World Neurosurg. doi:10.1016/j.wneu.2013.03.025.
87. Qu X, Su L, Wang M, Fan X. Two-year follow-up of osseointegration and rehabilitation in a patient with oral and maxillofacial arteriovenous malformations. Int J Oral Maxillofac Surg. 2013;42(9):1079–1082.
88. Ren Y, Kang SH, Park HS, Wu YN, Zhang LQ. Developing a multi-joint upper limb exoskeleton robot for diagnosis, therapy, and outcome evaluation in neurorehabilitation. IEEE Trans Neural Syst Rehabil Eng. 2013;21(3):490–499.
89. Dewald JPA, Ellis MD, Acosta AM, McPherson JG, Stienen HA. Implementation of impairment-based neurorehabilitation devices and technologies following brain injury. In: Neurorehabilitation Technology. Volker D, Nef T, Rymer WZ (eds). New York, NY: Springer. 2012.
90. Sukal TM, Ellis MD, Dewald JPA. Shoulder abduction-induced reductions in reaching work area following hemiparetic stroke: neuroscientific implications. Exp Brain Res. 2007;183(2):215–253.
91. Ellis MD, Sukal-Moulton T, Dewald JP. Progressive shoulder abduction loading is a crucial element of arm rehabilitation in chronic stroke. Neurorehabil Neural Repair. 2009;23(8):862–869.
92. Ellis MD, Sukal-Moulton TM, Dewald JPA. Impairment-based 3-D robotic intervention improves upper extremity work area in chronic stroke: targeting abnormal joint torque coupling with progressive shoulder abduction loading. IEEE Trans Robot. 2009;25(3):549–555.
93. Miller LC, Dewald JP. Involuntary paretic wrist/finger flexion forces and EMG increase with shoulder abduction load in individuals with chronic stroke. Clin Neurophysiol. 2012;123(6):1216–1225.