Portal de Periódicos da CAPES

Predicting Symptomatic Cerebral Vasospasm After Aneurysmal Subarachnoid Hemorrhage

2011; Lippincott Williams & Wilkins; Volume: 69; Issue: 2 Linguagem: Inglês

10.1227/neu.0b013e31821b7ed1

1524-4040

Edson Bor‐Seng‐Shu, Marcelo de Lima Oliveira, Manoel Jacobsen Teixeira, Ronney B. Panerai,

Cerebrovascular and Carotid Artery Diseases

To the Editor: Carrera et al1 recently addressed an important issue in clinical neurosurgery: the prediction of neurological deterioration in the setting of aneurysmal subarachnoid hemorrhage (SAH). They found that data from transcranial Doppler (TCD) blood flow velocities failed to predict delayed cerebral ischemia (DCI) in patients with aneurysmal SAH; the sensitivity of mean TCD blood flow velocity of more than 120 cm/s to predict DCI was only 63%, with a positive predictive value of 22% among Hunt-Hess grade I to III patients and 36% among Hunt-Hess grade IV to V patients. These results deserve consideration. It is believed that the prediction of future events in complex biological systems is quite difficult. DCI (clinical signs and symptoms of cerebral ischemia) results from a multifaceted interaction of a number of factors that influence cerebral blood flow (CBF) and metabolism such as: Severity of bleeding into the subarachnoid space routinely graded by the Fisher scale: the greater the SAH on computed tomography (CT) scan, the greater the risk of DCI. Patient's neurological condition frequently evaluated by Hunt-Hess classification scale: the worst the neurological condition, the greater the risk of DCI. Large- and small-vessel atherosclerosis: patients with severe atherosclerosis in cerebral arteries tend to have a higher risk of cerebral ischemia in cases of SAH vasospasm. Functional capacity of cerebrovascular collateral pathways. Cerebral microvascular reserve: progressive loss of cerebrovascular carbon dioxide reactivity identifies patients at high risk of DCI, and persistently normal reactivity implies a low risk. The sensitivity and specificity of impaired cerebrovascular reactivity for subsequent DCI is 91% and 49%, respectively.2 Moreover, impairment of cerebral pressure autoregulation was also demonstrated to be associated with cerebral vasospasm.3 Severity of cerebral vasospasm (degree of arterial narrowing, extension of arterial narrowing, number of arteries and arterial segments affected by spasm in different vascular territories or in the same vascular territory, and presence of tandem lesions). Degree of decrease in CBF: the lower the CBF, the closer the CBF reaches threshold level for cerebral ischemia; in patients with DCI, CBF was significantly lower, and both mean transit time and cerebral perfusion asymmetry were increased on perfusion computed tomography.4 Functional status of cerebral metabolism: hyperthermia and seizures can increase both the cerebral metabolic activity and the risk of DCI; furthermore, metabolic crises associated with SAH may increase the cerebral vulnerability to DCI.5 Site of ischemia or infarction in the brain: SAH patients may present large cerebral infarctions even remaining asymptomatic or mildly symptomatic (infarction can occur in clinically silent brain areas). Intracranial pressure: intracranial hypertension can further decrease CBF in patients with cerebral vasospasm,6 predisposing to DCI. Cardiocirculatory capacity to optimize cerebral blood flow: moderate-to-severe left ventricular dysfunction is associated with an increased risk of cerebral infarction from vasospasm; in addition, cardiac troponin elevation and depressed postoperative cardiovascular hemodynamic performance are also risk factors for DCI after SAH.7 Microthrombosis in the cerebral vascular bed after SAH: clinical studies show that SAH is associated with an activation of the coagulation cascade, impaired fibrinolytic activity, and inflammatory and endothelium-related processes, which can lead to the formation of microthrombi and DCI. In some cases, it is evident that vasospasm seems to be an epiphenomenon rather than a predominant mechanism for DCI.8 For the cited reasons, it is expected that a sole factor, CBF velocity, would correlate poorly with the risk of DCI, and it is likely that if various factors are analyzed in parallel, the model for predicting symptomatic vasospasm could be improved. In fact, Gonzalez et al9 recently proposed the vasospasm probability index (VPI), a combination of TCD blood flow velocities, CBF measurements, and clinical risk factors to predict DCI after aneurysmal SAH. They showed that the use of TCD velocities, Lindegaard ratio, and spasm index separately are of limited value for the detection of clinical vasospasm, whereas if Fisher grade, Hunt and Hess grade, and spasm index (ratio between TCD blood flow velocities and CBF measurements) are taken into account, the VPI had an accuracy of 92.9%, a sensitivity of 85%, and a specificity of 84%. According to this model, arteries with the probability of clinical vasospasm of 22% or more are confirmed to have clinical vasospasm. On the other hand, evaluating the cerebrovascular reactivity, Carrera et al2 reported that the sensitivity and specificity of impaired cerebrovascular reactivity for subsequent DCI are 91% and 49%, respectively. It is possible that if the testing of cerebrovascular reactivity were considered for the determination of VPI, the results would be improved further. Unfortunately, models that consider a number of predictive factors are time-consuming, expensive, invasive (need administration of carbon dioxide and radioactive substances) and, therefore, are limited for clinical practice. Edson Bor-Seng-Shu Marcelo de-Lima-Oliveira Manoel Jacobsen Teixeira Sao Paulo, Brazil Ronney B. Panerai Leicester, United Kingdom