Steroids

There is a theoretical justification for the use of glucocorticoids in the setting of ntSAH because of their putative effects on cytotoxic edema and inhibition of CSF production.[27] Early animal studies supported the use of corticosteroids to prevent delayed cerebral ischemia (DCI).[24] Human data from clinical trials, however, are scarce and often contradictory. The 2005 Cochrane Collaboration published a systematic review of three randomized controlled trials with a total of 256 patients and concluded that there was no definitive evidence of benefit or harm in administering corticosteroids to patients following ntSAH.[3]

The pendulum may be swinging in the other direction however with the results of a 2010 randomized double-blind, placebo-controlled, pilot trial of high-dose methylprednisolone in 95 patients with aneurysmal SAH. Based on their findings, Gomis et al. concluded that high-dose methylprednisolone administration improved functional outcome scores 1 year after ntSAH, despite a lack of effect on symptomatic vasospasm.[7] These results further support an emerging paradigm challenging the idea that vasospasm is the sole or primary mechanism for neurological decline following ntSAH. The physicians in this survey were essentially split down the middle as to their acceptance of the role of corticosteroids in ntSAH with 57.1% of respondents reporting routine usage. Interestingly, dexamethasone was used exclusively by survey respondents although, as demonstrated by Gomis et al., methylprednisolone may be preferable in this setting.

Thrombolytics and clot clearance

Available studies of thrombolytics in ntSAH differ not only in terms of the type used, but also the dosage, duration, infusion versus bolus therapy, and mode of delivery (e.g., intrathecal via lumbar catheter, intraventricular via EVD, direct infusion into the cistern during craniotomy for clipping, infusion via a catheter placed in the cisterna magna or other cisterns).[4,8,15] Given the heterogeneity of these studies, there is significant controversy among experts as to the efficacy of this approach.

Since the by-products of clot breakdown are felt to contain the culprit “spasmogens”, and clot burden as measured radiographically by the Fisher grade has been correlated with the likelihood of developing vasospasm, early and aggressive clot clearance has been advocated by many in an attempt to eliminate or at least minimize blood breakdown and reduce the incidence of vasospasm.[9,36] Furthermore, as the effects of blood breakdown in the basal cisterns may extend beyond detectable vasospasm, and may worsen outcomes by other as yet unknown mechanisms, there is a strong rationale for early and aggressive clot clearance before any of what may be a number of diverging cascades are initiated leading to neurological decline.[14]

To this end, mechanical agitation (i.e., “head-motion therapy”) as well as complicated catheter irrigation systems (e.g., combination of EVD and lumbar drain or other intrathecal drains) has been investigated as adjunctive measures in promoting CSF circulation and the clearance of cisternal clot.[9,15] In a 2008 study, a combination of lumboventricular lavage with a Ringer solution and low-frequency head-motion therapy continued for 5 days after ntSAH yielded decreased incidence of DCI with superior modified Rankin scores at 3 and 6 months.[9] These findings corroborated the results reported by Kawamoto et al. 4 years earlier on the effectiveness of “the head-shaking method combined with cisternal irrigation with urokinase” for the prevention of vasospasm and improvement of clinical outcomes in ntSAH. Not entirely without risk, the use of intrathecal urokinase in this study was associated with a 1.7% incidence of hemorrhage.[14]

While clot clearance has become a very popular approach in some parts of the world, American neurosurgeons seem to have had a comparatively tepid reaction to the use of thrombolytics in ntSAH. Only approximately 1/5 of the survey respondents said that they ever use intrathecal or intraventricular tPA in patients with ntSAH [Table 1], and over 90% would prefer not using tPA should the practice be standardized in the setting of a future RCT [Table 2]. One can only speculate that this lack of enthusiasm stems from a fear of hemorrhage in patients receiving thrombolytics, a complication that even if uncommon could be potentially devastating. Once the aneurysm has been properly secured by surgical clipping or endovascular means, however, the rate of clinically significant hemorrhage with intrathecal tPA administration is likely low. [21,29] Furthermore, the rate of hemorrhage appears to be dose dependent,[15] and must be carefully weighed against the increased risk of permanent disability from severe vasospasm when clot clearance is not aggressively pursued.

Antiepileptic drugs

Observed, in-hospital seizures may occur in up to 12% of patients with ntSAH.[10] The existing literature has mainly focused on clinically overt focal or generalized tonic-clonic seizure activity despite the fact that as many as 95% of seizures in ntSAH patients are nonconvulsive and can only be detected with EEG.[6] Electrographic seizures detected by continuous EEG have been independently associated with poor outcomes. The incidence of status epilepticus among ntSAH patients with nonconvulsive seizures may be as high as 70%.[2]

There are little data to support the use of prophylactic AEDs in patients with ntSAH. Some advocate for prophylactic AEDs administration for select patients with unsecured aneurysms, increased ICP, or poor clinical grade for whom a seizure might promote cerebral edema and further decline.[28] AED use, however, is not without risk and has been linked to worse outcomes in a recent study of patients with ntSAH by Rosengart et al.; specifically AEDs may alter synaptic growth and connectivity leading to impaired neurological recovery.[28] Certainly, of the available agents, levetiracetam is better tolerated than phenytoin, has fewer interactions with other medications, and is preferred by nearly all of the respondents in the questionnaire.

Magnesium

Magnesium promotes cerebral vasodilation via blockage of voltage-dependent calcium channels, and inhibits the NMDA receptor involved in glutamate-mediated neuronal death.[31,32] A much-anticipated phase III multicenter RCT (IMASH) released in May 2010 found no clinical benefit in ntSAH patients treated with a 10- to 14-day intravenous infusion of magnesium sulfate. Some objections were raised, however, about the magnesium levels in patients in this study and the timing of therapy. The average serum magnesium level in the treatment arm was only 1.67 +/? 0.27 mmol/L.[35] A separate 2010 RCT by Westermaier et al. found a 21% incidence of DCI in patients with serum magnesium levels kept between 2.0 and 2.5 mmol/L compared to 51% in the control group, indicating that higher concentrations may make a difference.[34] Intra-cisternal administration has been advocated to attain the high brain concentrations of magnesium necessary for cerebral vasodilation without generating excessively high serum levels that can result in bradycardia and hypotension.[22]

The average delay in treatment in the IMASH trial was over 30 hours from hemorrhage onset. The IMAGES Study first suggested the possibility that timing of magnesium therapy might play a role in observed efficacy, and offered the recommendation that therapy should be instituted within at least 48 hours of hemorrhage to attain therapeutic levels before the vasospasm period begins.[23] An ongoing trial (FAST-MAG) looks to answer the question whether even earlier administration, within 2 hours, might have an effect on outcomes. In our survey only about 1/5 of the responding physicians used intravenous magnesium routinely in patients with ntSAH[Table 1].

Triple-H therapy

To date, no randomized clinical trials regarding triple-H therapy (i.e., hypertension, hypervolemia, hemodilution) have been accomplished, and the evidence supporting the utility of certain components is less than compelling. While numerous studies have cast doubt on the benefit of hypervolemia and hemodilution,[11,12,18] induced hypertension is still commonly used. The use of hypertensive therapy gained popularity in the 1970s after Kosnik and Hunt described seven secured SAH patients who developed neurological deficits secondary to vasospasm and whose deficits reversed with hypervolemia and phenylephrine-induced hypertension.[16] Because patient outcomes in cases of SAH-induced vasospasm and secured aneurysms have historically been reported as improved with hypertension,[13,16] ethical considerations have prevented a prospective randomized evaluation of hypertensive therapy. Several studies examining the use of prophylactic triple-H therapy have revealed no benefit in ntSAH patients.[25] Despite this, over 70% of the respondents to our questionnaire use mild prophylactic triple-H therapy on all patients with escalation in those who develop symptoms.

Detection of vasospasm

Digital subtraction angiography (DSA) is widely referred to as the “gold standard” for the diagnostic assessment of cerebral vasospasm following ntSAH.[33] This designation, however, may be potentially misleading as only a fraction of those with angiographic vasospasm suffer from symptomatic vasospasm, and there is a questionable relationship between angiographic vasospasm and clinical outcome.[5,25] Investigating the clinical relevance of detected vasospasm, Frontera et al. found that symptomatic vasospasm and DCI as defined by delayed infarct on head CT were both predictive of clinical outcome, while angiographic vasospasm did not seem to be so. Nevertheless, numerous studies have tried to gauge the accuracy of myriad modalities (e.g., TCD, CTA, MRA, CT perfusion, diffusion-weighted MRI, SPECT, continuous EEG, intrathecal markers) by comparing each of these to angiographic vasospasm.

In terms of correlating TCD to angiographic vasospasm, a systematic review from 2001 concluded that TCD was highly specific (99%) for the detection of MCA vasospasm with a sensitivity of 67%.[19] CTA meanwhile has a strong correlation with DSA for detecting vasospasm in proximal arterial segments, especially in the setting of severe vasospasm. Discrepancies between the two tests occur when vasospasm is mild or moderate.[1,26] While both TCD and CTA provide valuable information about the caliber of proximal cerebral vasculature, various perfusion methods including dynamic CT perfusion are increasingly being applied in ntSAH to assess global and relative cerebral blood flow and detect microcirculatory dysfunction.[17] Absolute cerebral blood flow values of less than 25 mL/100 g/min and mean transit times greater than 6.5 seconds using CT perfusion have been associated with a high risk of delayed ischemic deficits.[30] CBF measurements from CT perfusion scanning correlate well with SPECT studies. CT perfusion is widely available and adds only a few minutes to the conventional CT study; its clinical role in predicting DIC following ntSAH will undoubtedly continue to expand.[30]