Establishing the Diagnosis

Inherent to surgical intervention, the vast majority of neurosurgical consultations involve diagnostic neuroimaging. In most cases, referral to a neurosurgeon occurs as a consequence of abnormal imaging results, already creating a unique patient-physician relationship. The neurosurgeon is often confronted with the situation of deciding whether the patient’s symptoms or signs, if any, correlate with the neuroimaging findings. This is in stark contrast to neurological practice, in which certain diagnoses such as headache and seizures may not receive diagnostic imaging initially or even at all. Although the finding of a neuroimaging abnormality is not a prerequisite for neurosurgical intervention (such as movement disorder surgery), neuroimaging is required nevertheless (Britz et al., 1995; Kent et al., 1994; and Nijeholt et al., 1998).

The reliance upon diagnostic neuroimaging has changed neurosurgical residency training, necessitating that neurosurgeons achieve a high level of diagnostic neuroradiological competence. For the neurosurgeon, the diagnostic process (and surgical planning) typically hinges upon the imaging, and therefore a considerable amount of time is devoted to carefully, personally reviewing the imaging studies. As a result, when referring a patient for neurosurgical evaluation, it is important to emphasize to the patient that the actual imaging study be physically available for review at the time of the neurosurgical consultation.

It is important to understand the difference between standard diagnostic imaging and the high-resolution neuroimaging required for neurosurgical decision making. For example, routine diagnostic magnetic resonance imaging (MRI) studies of the brain are performed with relatively thick 5-mm slices, with 2.5-mm skip between slices. Although generally adequate to detect lesions 4 mm in size or larger, partial volume averaging can obscure fine anatomical detail such as the relationship of arteries, veins, and cranial nerves to the lesion. For example, although the diagnosis of hydrocephalus may be obvious on routine diagnostic imaging, establishing the etiology of the ventriculomegaly may require high-resolution constructive interference in steady state (CISS) imaging (Fig. 46.1). In this case, a diagnosis of occult aqueductal stenosis would dramatically change the treatment, with endoscopic third ventriculostomy being the preferred treatment rather than a ventriculoperitoneal shunt.

Fig. 46.1 Imaging aqueductal stenosis. Left, Midsagittal noncontrast T1-weighted magnetic resonance image (MRI) showing enlarged lateral ventricle, bowing of corpus callosum, and normal-sized fourth ventricle. Cerebral aqueduct appears patent. Right, Corresponding constructive interference in steady state (CISS) MRI reveals a distal cerebral aqueduct occlusive membrane (white arrow). Premesencephalic cistern has favorable anatomy with a posteriorly placed basilar artery bifurcation (black arrow) for endoscopic third ventriculostomy.

(From Bergsneider, M., Miller, C., Vespa, P.M., et al, 2008. Surgical management of adult hydrocephalus. Neurosurgery 62, 643-660.)

With routine “diagnostic” imaging, the number of different MRI sequences is often limited by the concept of the study being a screening exam rather than one honing in on a certain differential diagnosis. For example, atypical arachnoid cysts can be differentiated from epidermoid cysts if repeat imaging includes a diffusion-weighted imaging (DWI) study (Tsuruda et al., 1990). In this case, an arachnoid cyst may not require referral to a neurosurgeon, whereas the epidermoid cyst should. Consultation with a neuroradiologist, if needed, could lead to proper imaging studies being performed prior to possible neurosurgical consultation.

In other cases, routine diagnostic imaging may be adequate for establishing a diagnosis but inadequate for neurosurgical decision making. In some cases, additional imaging modalities may be required, such as computed tomography (CT) imaging of bony anatomy to supplement abnormal soft-tissue findings on MRI. In other cases, special high-resolution or unique imaging sequences may be required (Barnett et al., 1993; Curtin and Hirsch, 1992). Although beyond the scope of this chapter, suffice it to state that a brief discussion with the neurosurgeon—prior to referral—may in some cases lead to essential imaging studies being obtained prior to the consultation. This may make for a much more productive and useful consultation, obviating the need to have the patient return for an otherwise unnecessary follow-up visit owing to the lack of key imaging studies at the first visit.

Patient Selection for Surgery

Patient selection for surgery is under the purview of the neurosurgeon in most cases, although for many disorders, the neurologist plays an integral and key role. As noted previously, the increasing sophistication and complexity of surgical intervention mandates intimate collaboration between neurologists, neurosurgeons, and other subspecialists. Knowing when to refer to multidisciplinary specialty centers is obviously a function of patient diagnosis and proximity to such centers. In general, community neurosurgeons are well versed in treating a wide range of neurosurgical disorders, although increasingly, many are limiting their practices to mainly treating spinal disorders. Increasing evidence suggests that for specific diagnoses, surgical outcomes are a function of surgical volume load and experience. For example, dedicated brain tumor neurosurgeons performing over 50 brain tumor operations per year generally have better patient outcomes than those only occasionally performing these surgeries (Cowan et al., 2002; and Cowan et al., 2003; D’Agostino et al., 2009; and Shahlaie et al., 2010). Similar findings have been reported for cerebral aneurysm treatment, epilepsy, movement disorders, and pediatric disorders. Most academic centers and some large group private practices have dedicated subspecialty neurosurgeons (as well as the necessary ancillary team) to comprehensively and optimally manage these patients.

Surgeon experience, however, is only part of the equation. Whereas it is highly possible a community neurosurgeon can perform an adequate surgical removal of a brain tumor, neurosurgeons affiliated with a brain tumor center may be able to offer patients clinical studies they might not otherwise be aware of. For some trials, fresh pathological tissue is required, and therefore the operation has to be performed at the site of the trial. Understanding the practice of the neurosurgeons to whom you routinely refer may therefore indirectly improve the care your patients receive.

Patient selection for surgery has become increasingly complex, and new treatment modalities and alternatives have arisen. Just as surgical volume load and experience varies among neurosurgeons, it may be important that the referring physician understand other important factors in patient selection for surgery. An example is vestibular schwannoma (acoustic neuroma) management. As already noted, surgical experience may play a major role with regard to patient outcome. It is important to recognize, however, that fewer patients are choosing surgery for small acoustic neuromas, and instead are undergoing stereotactic radiosurgery or radiotherapy treatment (Andrews et al., 2001; Flickinger et al., 2001; Lunsford et al., 2005; and Meijer et al., 2000). Referring to a neurosurgeon whose practice is located in an area without focal beam radiation capabilities may inadvertently lead to an emphasis on surgical alternatives. A similar situation exists with intracranial aneurysms, in which endovascular coiling has increasingly become the treatment of choice for selected patients. Typically, high-volume neurovascular surgeons are closely affiliated with (or perform themselves) endovascular treatment, and therefore this issue is a moot point (D’Agostino et al., 2009).

Neurosurgery is increasingly subspecialized (as is neurology), and certain diagnoses may require a higher level of expertise. In addition to skull base tumors and neurovascular disorders, others include intractable epilepsy, adult hydrocephalus, movement disorders, and rare conditions such as colloid cysts. A knowledgeable neurologist making the proper initial referral may be a key factor in outcome.

In some cases, the neurologist is the primary determinant of patient selection for surgical intervention. The management of movement disorders is an obvious example. In only a minority of patients is surgical intervention such as deep brain stimulation appropriate Olanow and Koller, 1998). Referral to a movement disorder neurologist first may be appropriate instead of initial referral to a functional neurosurgeon. Other examples include carotid endarterectomy (versus stenting) and cerebrovascular bypass procedures.

To some extent, patient selection for surgery begins prior to neurosurgical evaluation. There are no hard-and-fast guidelines as to what types of neurological diagnoses should be referred for neurosurgical consultation, but in most cases, the decision to refer for neurosurgical consultation is obvious. Symptomatic patients with obvious abnormalities seen on imaging, such as a large cerebral aneurysm, require neurosurgical consultation. There are other cases, however, that may not require neurosurgical consultation and can be managed well and appropriately by the neurologist. Although it may seem obvious and trite, it is important to understand that a neurosurgical practice is a surgical practice. With the increasing use of neuroimaging in our society, so-called incidentalomas are being noted by radiologists (Fainstein et al., 2004; Molitch, 2008; and Molitch, 2009). In some cases, this has led to reflexive referrals to neurosurgeons based solely on the radiology report, without the personal review of the imaging study. Referral of these nonoperative lesions to the neurosurgeon first raises the expectation of the patient that surgical intervention is advised and/or pending, and second takes up a significant amount of the neurosurgeon’s clinic time explaining why the lesion does not require surgery. Examples follow.

Arachnoid Cyst

The vast majority of arachnoid cysts are small and asymptomatic and consequently, nonsurgical. A typical location is the anterior middle fossa where the anterior temporal lobe is displaced posteriorly (Fig. 46.2). Small arachnoid cysts that do not produce significant mass effect (midline shift, brainstem compression) are rarely a cause of headaches (Pierre-Kahn et al., 1990; Pradilla and Jallo, 2007). If it is equivocal whether the lesion is large enough to warrant surgical intervention, a quick film review by the neurosurgeon is often all that is needed to adequately triage the patient.

Fig. 46.2 Transaxial computed tomography image demonstrating bilateral middle fossa arachnoid cysts. Although there is displacement of anterior temporal lobes, note the absence of clinically relevant mass effect (i.e., no effacement of basal cisterns). In most cases, reassurance and follow-up imaging is all that is needed.

Intervertebral Disk Bulges and Annular Tears

Here, the correlation of neuroimaging findings and neurological symptoms and signs is key (Fujiwara et al., 1999) (Fig. 46.3). Clinically insignificant disk bulges are so common in our aging society that a neurosurgical spine specialist could easily be inundated with nonsurgical consultations, stealing time from a practice tailored to helping appropriately referred patients (Stabler et al., 1996; Stabler et al., 1997). Admittedly, inappropriate patient referral to neurosurgeons for asymptomatic disk bulges is infrequent by neurologists, given their training and understanding of neural anatomy.

Fig. 46.3 Sagittal T2-weighted magnetic resonance imaging of lumbar spine showing a disk bulge at L5-S1 level. Referral to a neurosurgeon (or other spine specialist) would be indicated if corresponding neurological symptoms existed.

In some cases, certain diagnoses may not require immediate neurosurgical referral. A prototypical example is pseudotumor cerebri (idiopathic intracranial hypertension). It is generally accepted that surgical intervention is one of the last treatment options for this condition (Acheson, 2006; Kelman et al., 1992; Rosenberg et al., 1993). Although many, if not most, neurologists primarily manage this condition, some initiate a referral to neurosurgery for consideration for a cerebrospinal fluid (CSF) shunt procedure. Again, this sometimes raises patient expectations and leads to frustration and delay of (medical) treatment. As diuretics and weight loss are the primary treatment modalities, neurosurgical consultation may never be necessary. However, with rapid visual deterioration, optic nerve fenestration or CSF diversion may be indicated. Also, consideration of intracranial venous occlusive disease should be made for non-obese pseudotumor patients (Mokri et al., 1993; Owler et al., 2003; Owler et al., 2005). A telephone conversation with the neurosurgical colleague could direct the completion of a contrast MR venogram or CT venogram to rule out venous sinus abnormalities.

Equally important to averting unnecessary neurosurgical consultation is avoiding delayed referral. Again, the scope of this chapter does not allow a detailed list of possible diagnoses, but a few examples are given for demonstrable purposes.

Preoperative Surgical Planning

The initial patient encounter marks the beginning of surgical planning. As discussed earlier, high-resolution three-dimensional (3D) multimodality neuroimaging is increasingly used on a routine basis. There are several factors the neurosurgeon takes into account for each case. Many are specific to the type of pathology. For intracranial lesions, there is increasing use of 3D reconstructions depicting the relationship between the pathology and surrounding structures.

Vascular lesions are often best characterized by digital subtraction angiography, in some cases augmented by 3D angiography techniques. This modality provides functional and temporal information of the vasculature and is most helpful to characterize arteriovenous malformations (AVMs), complex aneurysms, and various tumors. Magnetic resonance angiography (MRA), however, is rarely helpful to characterize AVMs and aneurysms and therefore is a waste of medical resources if ordered. In some cases, CT angiography (CTA) is adequate for presurgical planning for straightforward intracerebral aneurysms (Papke et al., 2007) (Fig. 46.4).

Fig. 46.4 Three-dimensional (3D) reconstruction of a bilobed right internal carotid artery bifurcation aneurysm (arrow) visualized using computed tomographic angiography (CTA). The 3D anatomy gives the neurovascular surgeon important insight into aneurysm configuration and orientation.

Brain tumor surgical planning can be quite complex. The location of the tumor is often the defining variable. Again, technology has significantly advanced the surgical treatment of brain tumors. Surgical planning for skull base tumors such as meningiomas and vestibular schwannomas may involve consideration of several different surgical corridors, each with inherent advantages and risks. Deep-seated intraaxial tumors that require transcortical approaches ideally should be performed with intraoperative stereotactic neuronavigation that incorporates functional imaging (fMRI), diffusion tensor imaging, white-matter tractography, and sometimes MR spectroscopy (MRS). Fiber tractography (Fig. 46.5) can help avoid damage to axonal structures like the optic radiations and arcuate fasciculus, whereas MRS can aid in biopsy site choice (Dowling et al., 2001; Kamada et al., 2005; Kamada et al., 2007). These imaging modalities are not universally available and require experienced technicians and complex postprocessing. For the neurologist, awareness of these considerations and possibilities should play a role in deciding where to refer patients. High-volume neurosurgical subspecialists who routinely use this technology should be sought for appropriate cases.

Fig. 46.5 Diffusion tensor magnetic resonance imaging (MRI) can be used to construct white-matter fiber representations (tractography). Combined with anatomical cues and sometimes functional MRI (fMRI) data, the relationship of specific key white-matter bundles can be determined relative to brain tumor location. This data, and often other data such as MR spectroscopy and/or positron emission tomography imaging, can be fused and imported into the surgeon’s stereotactic neuronavigation unit. Brain mapping has revolutionized brain tumor surgery, allowing safer, more effective surgery.

Epilepsy surgery requires a different set of diagnostic and surgical decision-making tools. When the location of the epileptogenic focus is not clear or may be attributable to numerous lesions, studies to help determine the culprit lesion are necessary. Cortical surface electrode electroencephalography (EEG), magnetoencephalography (MEG), or depth electrodes may be needed to identify the source of seizures; such modalities are generally only available at dedicated epilepsy centers (Collura et al., 1990; Matsumoto, 1990).

Spinal pathology adds additional diagnostic challenges to differentiating causes of pain and weakness. Dynamic studies such as flexion-extension spinal films can determine instability or ligamentous injury (Elsig and Kaech, 2006). Epidural or facet joint steroid injections can be both diagnostic and therapeutic in the treatment of back pain and radicular pain, directing surgical planning (Cuckler et al., 1985; Koes et al., 1995). Electromyography (EMG) and nerve conduction studies can help pinpoint the source(s) of peripheral nerve pathology (Levin et al., 1996; Tullberg et al., 1993).

The goals of surgery must be clear and direct prior to entering the operative suite. Surgical goals depend on the underlying pathology. These goals include establishing a diagnosis, decompression of neural structures, stabilization, reconstruction, and often cure. The neurosurgeon is ethically mandated to discuss nonsurgical options with the patient or family. The preoperative discussion is essential to emphasize goals and expectations during and after surgery. Expected and unexpected neurological outcomes are discussed, as well as the inpatient and outpatient postoperative course. A common misconception is that surgery will be curative, and patients will be able to return to work immediately. A preoperative discussion can be helpful to provide more realistic expectations.

Similar to medical management of diseases, surgical management may require multiple procedures to achieve the goals of care. Such examples include performing a surgical biopsy prior to surgical resection, approaching a spinal deformity anteriorly prior to a posterior approach, or resecting a lesion prior to placing a ventriculoperitoneal shunt. Advising patients and their families of future events will help prepare them for the treatment process and instill confidence in the surgical team.

“Half of the surgery is done before an incision is made!” This statement is most applicable in the field of neurosurgery. It is not uncommon for the pre-incision activities to take as long as the surgery itself. Attention to specific details such as airway management, intracranial pressure (ICP) management, neuromonitoring, image guidance, brain retraction, and incision planning help ensure safe surgery and plan for untoward events.

As important as a patient’s decision to have surgery is the surgeon’s decision to offer surgery. The “big picture” must be evaluated for each patient. For example, surgical resection of a small meningioma may be indicated in a healthy young adult but may not be suitable for the vasculopathic, emphysematous, octogenarian smoker. An appropriate preoperative evaluation should also be performed by medical specialists to evaluate for cardiac function, pulmonary capacity, infectious risk, and endocrinopathies. This can help stratify risk for the patient and assist the anesthesia and surgical teams to anticipate intraoperative and postoperative complications.

Surgical Technique

Neurosurgery entails several key considerations. The type of anesthesia is tailored to the patient’s pathology and surgical plan. For example, patients with cervical stenosis and myelopathy cannot tolerate direct laryngoscopy and cervical manipulation because of the risk of cervical spinal cord injury. Therefore, awake fiberoptic intubation is performed to avoid cervical hyperextension and allow the patient to control their cervical musculature (Fuchs et al., 1999). Sometimes it is the neurologist, who has the time to comprehensively assess patients preoperatively, who diagnoses conditions such as spinal stenosis, even if it is not the primary condition requiring surgical intervention. Another example is the patient who is dependent on optimal cerebral perfusion pressure, such as in carotid stenosis or compressive cervical myelopathy, and does not tolerate the significant hypotension that is otherwise forgivable with general anesthesia. To combat this, invasive arterial blood pressure monitors are placed prior to induction, and mean arterial pressure (MAP) is kept normotensive throughout the surgery (Epstein, 1988).

Intraoperative neuromonitoring is routinely performed in a variety of neurosurgical operations. Modalities include EEG to monitor for vascular compromise, somatosensory evoked potential (SSEP) and motor evoked potential (MEP) monitoring for cortex through spinal cord viability, brainstem auditory evoked response (BAER) monitoring for acoustic nerve and brainstem injury, and direct nerve stimulation with EMG for cranial nerve and peripheral nerve monitoring. If a change from baseline is encountered in any of these monitoring modalities, surgical or anesthetic adjustments are performed (Minahan, 2002). Examples include repositioning an aneurysm clip if there is a decrease in SSEPs, EEGs that raise concerns of cortical ischemia, tailoring the extent of spinal cord tumor resection if there is a decrease in SSEP and MEP signals, or adjusting spinal pedicle screw trajectories if there is EMG conduction through the neural foramina (Dehdashti et al., 2006). Medical interventions can include increasing MAP for patients exhibiting ischemia, or inducing cortical burst suppression on EEG as a neuroprotective measure (Lavine et al., 1997).

One major advancement of modern neurosurgery is intraoperative stereotactic neuronavigation. The technology entails registering (co-aligning) the patient’s head to a preoperative 3D high-resolution image (MRI or CT). The location of the tip of a pointer is then depicted on the imaging monitor (Fig. 46.6). This technology has proven helpful in determining the extent of tumor resection, especially if intraoperative CT or MRI is performed (Knauth et al., 1999). Stereotactic biopsy of lesions has become increasingly accurate (Moriarty et al., 2000). On the other hand, neuronavigation can help the surgeon avoid eloquent cortical regions such as the Broca area, Wernicke region, the corona radiata of the primary motor and sensory tracts, and the arcuate fasciculus. Concomitant fMRI and diffusion tensor imaging (DTI) with postprocessing is required to implement this technology (Wu et al., 2007). Lastly, this technology allows smaller incisions and craniotomies compared to those of the past.

Fig. 46.6 Screen capture taken from an intraoperative neuronavigation system. Virtual pointer displayed on image corresponds to tip of an actual pointer being used by the surgeon. Neuronavigation allows for precise localization of craniotomies and delineation of tumor (lesion) boundaries that are sometimes not apparent by direct vision.

Along with utilizing complex technology to avoid neurological complications is the more straightforward concept of passive brain retraction and venous drainage. Much intraoperative morbidity can be prevented by patient positioning. For intracranial surgeries, the head position can be adjusted to let gravity retract the brain from the lesion (Rolighed Larsen et al., 2002). The use of the falx cerebri as a natural retractor is helpful for interhemispheric lesions (Stone et al., 1990). Although these technical nuances may not be relevant to a neurologist’s daily practice, again it underscores the responsibility of the referring physician to guide patients to specialists who are highly experienced and use modern techniques that minimize operative morbidity.

In some highly specialized neurosurgical procedures, the neurologist plays an integral intraoperative role. One example is awake craniotomies for cortical mapping of eloquent language areas (Costello and Cormack, 2004). Another is epilepsy lesionectomy procedures in which the interpretation of intraoperative corticography is required.

Postoperative Management

The postoperative period can be critical to surgical success. In many larger hospitals, the neurointensive care unit incorporates a neurologist as the neurointensivist. This hospitalist provides minute-to-minute care that cannot be provided by a neurosurgeon while in the operating room. In general, prevention of complications is ideal. Early neurological examination establishes a postoperative neurological baseline. Improvement or decline from preoperative examination should be noted. Early identification of neurological deficits can help diagnose a hematoma, ischemia, or edema that at times can be reversible. Therefore, thorough, frequent, and sometimes continuous neuromonitoring is essential for quick identification and intervention of surgical complications (Bhatia and Gupta, 2007). Many modern intensive care units offer continuous EEG monitoring, again highlighting the natural marriage of neurointensive care with neurology (Guerit, 1999; Guerit et al., 2009).

Postoperative seizure prophylaxis is a controversial topic but is typically initiated intraoperatively and maintained up to 1 week following surgery using established antiepileptic (AED) agents (Horwitz, 1989). Although this has not been proven to decrease the incidence of postoperative seizures or epilepsy, AED use is justified to prevent catastrophic seizures that may lead to increased morbidity and mortality (Glantz et al., 2000). The neurologist can play a key role in postoperative management in this regard, helping determine the best AED agent and duration of treatment.

Another key role neurologists should play is in the long-term management and follow-up of patients with neurosurgical problems. This includes recovery from neurotrauma, intracerebral hemorrhage, hydrocephalus, brain tumor, and epilepsy surgery. Management of chronic neurological conditions is intrinsic to neurological training, and the neurologist is in many ways ideally poised to offer this care. The management of certain neuromodulatory devices such as deep brain stimulators and vagal nerve stimulators is ideally performed by trained neurologists who have a detailed understanding of the patient’s neurological status and can titrate stimulator settings much like medications (Okun et al., 2005). Although all neurosurgeons desire to be updated on the long-term sequelae of their intervention, neurosurgical practice is tailored to providing the best perioperative management of patients. It is important to understand that a minor percentage of a neurosurgeon’s practice involves outpatient encounters, and therefore this precious clinic time should be largely dedicated to evaluating and treating surgical conditions.

Summary

The neurologist can play a key role in the surgical management of patients. This includes making the diagnosis, ordering appropriate preoperative imaging, referring to the best neurosurgeon for that given patient, educating the patient with regard to treatment options and pitfalls, and actively participating in the long-term management of the patient.