A 23-year-old man presents with a syncopal event and during evaluation was found to have a tumor in the posterior fossa. He underwent surgical resection of a tumor filling the fourth ventricle extending into the cerebellopontine angle and upper cervical spinal canal (Figure 16.1A). Pathology showed WHO grade II ependymoma. Following surgery the patient had dysarthria, dysphagia, and left hemi-ataxia. On the second day of surgery he underwent a reexploration with removal of an extradural hematoma in the posterior fossa and recovered well. A week later, he unexpectedly became gradually more stuporous, developed a new hemiparesis, and was transferred back to the neurosciences intensive care unit. On examination, he was not following simple commands, looked about, was mute and grinded his teeth. He had an increased snout reflex and bilateral grasp reflexes. His brainstem reflexes were intact. He had considerable right arm weakness barely overcoming gravity. CT scan showed new appearance of multiple hemispheric hypodensities, largely in the posterior frontal lobes (Figure 16.1B). His laboratory studies were normal. Cerebral angiogram shows diffuse cerebral vasospasm in both anterior and posterior circulation (Figure 16.1C).
FIGURE 16.1 A) MRI shows large ependymona in the 4th ventricle. B) CT scan about 10 days after surgery shows multiple hypodensities. C) The cerebral angiogram (sample of multiple series) shows cerebral vasospasm in the anterior circulation.
What do you do now?
In order to detect deterioration patients undergoing elective brain tumor surgery are commonly observed overnight in an intensive or intermediate care unit. Most patients will leave the intensive care unit without any complications. When neurointensivists are asked by neurosurgeons to become involved in postoperative care it is because the current clinical condition is unexplained and unusual, because seizures have occurred, or because a major systemic complication needs very close attention. Complications after craniotomy are relatively uncommon but may be more frequent after extensive and complex neurosurgery. To give some sense of what to consider, Table 16.1 lists several causes of deterioration after a craniotomy.
Any patient with an early deterioration may be having seizures, but most patients who are stuporous or comatose from seizures will have already shown focal twitches that then became more generalized and evolved into a nonconvulsive status epilepticus. The cause may not be clear or simply related to removal of brain tumor tissue. Best treatment options in these patients include intravenous levetiracetam loading with 1500–2000 mg or intravenous loading with 20 mg per kg of (fos) phenytoin.
Postoperative hemorrhage in the operative bed may or may not be symptomatic. When mass effect occurs, patients are more likely to decline. A more recently identified cause of neurologic deterioration is the appearance of a hematoma remote from the surgical site. These surprising venous hemorrhages may be in the opposite hemisphere or in the cerebellum in patients with surgery of cerebral hemispheres. It may also occur after drainage of an acute subdural hematoma. The mechanism is therefore most likely mechanical shift of the brain (“sagging”) after reducing intracranial pressure. These remote hemorrhages can become clinically relevant and because they are lobar in nature may actually present with new seizures. Hemorrhage in the cerebellar peduncles have produced new onset slurred speech, tremor, cerebellar ataxia, and nystagmus. Most of the time, these hemorrhages resolve on their own, and the impact on outcome is not substantial.
TABLE 16.1 Causes of Deterioration after Craniotomy
Seizures (partial or generalized) and status epilepticus
Postoperative hemorrhage (operative bed or remote)
Cerebral infarction (sacrifice of an arterial branch or cerebral vein)
Postoperative cerebral edema
Diffuse cerebral vasospasm
Medical complications (e.g., hyponatremia or hypernatremia after pituitary surgery)
Cerebral infarction can occur after craniotomy when there is sacrifice of an arterial or venous branch, and the typical example is a sizable meningioma extirpation with necessary sacrifice of large venous tributaries.
Some neurosurgical procedures have the potential for more specific complications in the postoperative days. These include patients with epilepsy surgery who have hemorrhagic complications associated with depth electrode placement and patients after pituitary surgery with major adrenal or thyroid deficiencies. Hyponatremia may occur after several days and may be profound. Hypernatremia may follow hyponatremia, and even a triphasic postoperative course (hypernatremia–hyponatremia–hypernatremia) may occur. Large amounts of fluids and marked decline in vasopressin levels may contribute. (see chapter 26)
Our patient had a postoperative complication that is well described, probably more common than appreciated, but rarely considered. Obviously, cerebral vasospasm is not recognized unless a cerebral angiogram is performed. Early on, even an MR angiogram may not be sufficiently sensitive to demonstrate vasospasm that starts in smaller arterial branches. The onset of new neurologic signs, particularly when not easily grouped into a single syndrome and the appearance of new onset multifocal ischemia on CT scan should point toward this possibility. In prior reported patients the diffuse cerebral vasospasm occured several days and up to a week after surgery, and its development was associated with clear clinical deterioration. Patients may have fluctuating neurologic deficits that could point toward the diagnosis. Cerebral vasospasm may be more prevalent in certain types of neurosurgical procedures and in particular after skull base surgery. A commonly implicated surgery is pituitary adenoma resection via transcranial or transsphenoidal approach.
The pathophysiology of cerebral vasospasm after tumor surgery is not known. Removal of tumor adjacent to the basal cisterns could release vasoactive substances, but this remains speculative. Intraoperative hemorrhage and postoperative blood on the subarachnoid cisterns were not prominent in reported cases and have not been factors in the cases we have seen. Mechanical manipulation with extensive mobilization of medium size arteries during surgery might be another cause. Yet, this does not explain the diffuse distribution of the vasospasm or its development after a prolonged interval. Therefore, neither what we know about delayed vasospasm following aneurysmal subarachnoid hemorrhage nor what we know about reversible cerebral vasoconstriction syndrome appears to apply well to these cases of vasospasm after brain tumor surgery.
After the diagnosis is established there is uncertainty on how to best approach and treat this condition. If the situation allows, the treatment should be 1) increase cerebral perfusion by opening up the larger arteries through angioplasty and 2) increase cerebral perfusion pressure with traditional methods of hemodynamic augmentation.
In our patient, there was significant diffuse cerebral vasospasm that warranted immediate balloon angioplasty and intra-arterial infusion of verapamil in multiple arterial segments. He was additionally treated with aggressive hemodynamic augmentation using vasopressors and albumin, and with calcium channel blockers. He recovered gradually and eventually achieved an acceptable functional outcome.
Even if unusual, it is important to consider cerebral vasospasm after brain surgery in any patient with unexplained deterioration because endovascular intervention is indicated and successful if done early.
KEY POINTS TO REMEMBER REGARDING STUPOR AFTER BRAIN SURGERY
· Neurological worsening after craniotomy for tumor surgery can be due to cerebral vasospasm. The cerebral vasospasm can be diffuse and not only in the surgical field and may only be documented by cerebral angiogram.
· Worsening can also occur because of hemorrhage in the surgical bed, remote hemorrhage, cerebral edema, or ischemic stroke from sacrifice of a large vein or artery.
· Postoperative seizures may present with focal seizures, which may evolve into partial or generalized status epilepticus.
Almubaslat M, Africk C. Cerebral vasospasm after resection of an esthesioneuroblastoma: case report and literature review. Surg Neurol 2007; 68:322–328.
Amini A, Osborn AG, McCall TD, Couldwell WT. Remote cerebellar hemorrhage. AJNR Am J Neuroradiol 2006; 27:387–390.
Bejjani GD, Sekhar LN, Yost AM, Bank WO, Wright DC. Vasospasm after cranial base tumor resection: pathogenesis, diagnosis, and therapy. Surg Neurol 1999; 52:577–583.
Brockmann MA, Groden C. Remote cerebellar hemorrhage: a review. Cerebellum. 2006;5:64–68.
LeRoux PD, Haglund MM, Mayberg MR, Winn HR. Symptomatic cerebral vasospasm following tumor resection: report of two cases. Surg Neurol 1991; 36:25–31.