Treatment of Poor-Grade Subarachnoid Hemorrhage Trial

Discussion

Endovascular treatment is relatively less invasive than clipping and has been demonstrated to improve outcome in patients with good-grade SAH.¹ As a result, a number of centers are now also treating patients with poor-grade SAH with early endovascular coiling. Several single-center, retrospective, nonrandomized studies on early endovascular treatment of patients with poor-grade SAH have been published,^{5⇓⇓⇓⇓⇓⇓–12} as summarized in Table 4. While these studies provide important data in understanding outcomes in this group of patients, there are a number of difficulties in interpreting their results. First, sizes of the studies are mostly rather modest and vary widely, from 11⁶ to 111 subjects.¹⁰ Second, the inclusion criteria are also extremely variable. For example, while 1 study considers WFNS grade IV and V to be poor-grade,¹² another study has only included WFNS grade V,⁶ and some have included patients on the basis of Hunt and Hess grades 4 and 5.^5,7,8,10 Significant heterogeneity also exists in outcome measures and follow-up. A small majority of studies have used the Glasgow Outcome Scale over the modified Rankin Scale. The follow-up also varied between 6 months¹² to up to 2 years.¹⁰ There are also differences in the definition of good and poor outcomes.^{5⇓⇓⇓⇓⇓⇓–12}

With regard to neurosurgical studies, the International Cooperative Study² reported a good recovery rate in patients with poor-grade SAH ranging from 10% to 33% in the various subgroups based on the timing of surgery. Some smaller scale single-center studies suggested that early surgical clipping may lead to better results, but mostly in selected groups.^3,13,14 A study that included patients with poor-grade SAH who did not receive aneurysm treatment (ie, neither clipping nor coiling) reported a mortality rate of 71%.¹⁵

Specifically lacking from the endovascular studies are data analogous to the evidence base for early clipping, which compares coiling with conservative treatment; and the argument largely depends on the extrapolated assumption that coiling is a benign intervention. While the studies mentioned inform clinical practice in poor-grade SAH to a limited extent, significant uncertainties still exist, and robust evidence in the form of a randomized controlled trial is lacking.

The present study demonstrates that recruitment to such a randomized controlled trial is feasible. The recruitment rate among patients eligible for the study was encouraging (57%). The recruitment rate among the total number of patients screened for the study was much more modest at 16%. Screening and recruitment in the current study were largely performed during normal working hours, thus excluding a number of patients outside these hours. As shown in Table 1, the most frequent reasons for ineligibility (36%) were the following: 1) expiry of the time window for randomization, and 2) improvement of neurologic status before the patient could be randomized. If patient screening and recruitment are supported by the research staff 7 days a week, a higher recruitment rate would be realistic.

The study set out to assess the safety of conducting a randomized controlled trial in this group of patients. Six of 8 patients in the study underwent endovascular aneurysm treatment, while the other 2 died before aneurysm treatment. None of the patients undergoing endovascular treatment had any treatment-related complication; this finding suggests that a trial such as the present one is safe to conduct. As expected, a proportionately high number of disease-related adverse events did occur (Table 2). With regard to the feasibility of performing early coiling within the stipulated timeframe, the average time from randomization to treatment in patients in the early treatment arm was 9 hours 30 minutes, which suggests that the trial requirement of a maximum of 24 hours between randomization and treatment is achievable. No patient in this study had to be excluded because of noncompliance to this requirement.

Hydrocephalus is often thought to be the cause of incorrect WFNS grading and, therefore, incorrect classification as poor-grade SAH. To identify “true” poor-grade SAH, some centers advocate sedation-reversal followed by formal neurologic assessment.³ However sedation-reversal increases the risk of aneurysm rebleed, and postreversal neurologic assessment introduces delay into the process of randomization and treatment, making it difficult to study the benefits of early treatment. In this study, the decision regarding sedation-reversal has been left to the intensive therapy unit team. If a patient's postreversal neurologic status improved rapidly to the extent that he or she no longer had a poor-grade SAH, then the patient was not randomized. If a patient was randomized, then he or she remained in the allocated trial arm regardless of the speed of recovery. Hydrocephalus was used as a stratification criterion to ensure that an equal proportion of patients with hydrocephalus were included in the 2 arms and that the outcomes were not skewed by this factor.

The overall mortality rate in this study was 75%, which is higher than that in many reported series. This is most probably due to the small size of this study and possible bias toward randomizing more patients with grade V than with grade IV. Most patients in this study had admission WFNS grades V; and 5 of 6 patients who died had an admission WFNS grade of V.

Although the numbers are small, our experience suggests that the benefit of early endovascular treatment in a poor-grade SAH population should not be assumed unless there are robust multicenter randomized controlled trial data to support it. In this study, the good outcome rate in the treatment after recovery arm (33%) was better than that in the early coiling arm (20%).