Response to Letter by Vergouwen et al
Thank you for your comments. The first point the Vergouwen and colleagues make is prior studies have shown an association between angiographic vasospasm, delayed neurological deterioration (DND) and poor outcome after subarachnoid hemorrhage but that there is no proof that this relationship is causal. We already concluded this in a recent review that they cite.1 Of the references in their letter, only Fisher et al2 show an association between angiographic vasospasm and DND. The other articless compare “symptomatic vasospasm” with DND. This is a circular comparison because patients develop DND and then undergo transcranial Doppler or rescue therapy. The velocities are sometimes elevated so they are diagnosed with DND, but this is not an accurate test for angiographic vasospasm. If they improve after rescue therapy they are diagnosed with DND, but this also is not necessarily angiographic vasospasm. In the end, as we wrote, we agree with the summary sentence from our article that the authors quote and with the conclusion from the article of Vergouwen et al,3 which includes the hypothesis that there are multiple interacting pathways that lead to DND.
On the other hand, absence of evidence for causation does not mean there is no causation. The exact contribution of angiographic vasospasm to DND cannot be determined until such time as a treatment, completely devoid of any other pharmacological actions or side effects of any kind, cures angiographic vasospasm and the effect on DND and outcome is observed. The CONSCIOUS-1 study did decrease angiographic vasospasm, but this does not provide the answer because there is more than one explanation for the dissociation, the lack of importance of vasospasm being the one proposed by the authors of this letter. This is certainly possible, but one cannot scientifically exclude the other possibilities that already have been discussed and include microcirculatory spasm, thromboemboli, delayed neuronal apoptosis and cortical spreading depressions.3,4 Sample size issues, drug toxicity and insensitivity of outcome measures also need to be considered.
That angiographic vasospasm is not a cause of DND, however, seems far-fetched. We find it difficult to not use angioplasty of vasospastic arteries in, for example, patients with hemiparesis and severe middle cerebral artery spasm. If you do use it, then you have to believe angiographic vasospasm is of some importance. All of the patients that died in the placebo group in CONSCIOUS-1 died with severe diffuse angiographic vasospasm and large, multiple, territorial cerebral infarctions with brain swelling and increased intracranial pressure. This is an association again, but if vasospasm was not the cause of death then one has to conclude that microcirculatory spasm, thromboemboli, delayed neuronal apoptosis and/or cortical spreading depressions caused the angiographic vasospasm secondarily or that the 2 processes always coexist. It is frequently cited that patients have DND but no vasospasm.3 In CONSCIOUS-1, according to the investigator opinion and including all patients, there was a good correlation between severity of vasospasm and DND. Only 8 of 272 (3%) patients with no or mild angiographic vasospasm had DND compared to 68 of 110 (62%) of those with moderate or severe vasospasm (P<0.0001, Pearson χ2 test). Furthermore, because there is some evidence, albeit anecdotal, that angioplasty improves neurological condition and prevents territorial infarctions, the reverse could be true.5
Second, the intention-to-treat analysis produces the same results as the results that are presented. Only 4 patients that received no study drug were not included. Including them has no impact on the results. The authors comment again on the dissociation between angiographic vasospasm and DND. Reasons for this are already discussed above.
The question regarding the end points chosen for the CONSCIOUS-2 study misses the point. Regulatory authorities require studies of clazosentan to answer 2 questions. First, by reducing vasospasm, does clazosentan, as it did using blinded central assessment in CONSCIOUS-1, significantly reduce morbidity/mortality related to vasospasm?6 This is the primary end point in CONSCIOUS-2. Second, does this morbidity/mortality reduction correspond to an improvement in 3 month outcome as measured by extended Glasgow outcome score, which is the secondary end point in CONSCIOUS-2? A positive study would be significance on the first and at least a trend on the second. One has to consider, as will the regulatory authorities, whether to approve a new treatment for subarachnoid hemorrhage when we are unable to demonstrate a clear process through which the efficacy is achieved. We think that the design of CONSCIOUS-2 will achieve this.
Dr Macdonald is a consultant for Actelion Pharmaceuticals.
Nolan CP, Macdonald RL. Can angiographic vasospasm be used as a surrogate marker in evaluating therapeutic interventions for cerebral vasospasm? Neurosurg Focus. 2006; 21(3): E1.
Macdonald RL, Kassell NF, Mayer S, Ruefenacht D, Schmiedek P, Weidauer S, Frey A, Roux S, Pasqualin A. Clazosentan to overcome neurological ischemia and infarction occurring after subarachnoid hemorrhage (CONSCIOUS-1). randomized, double-blind, placebo-controlled phase 2 dose-finding trial. Stroke. 2008; 39: 3015–3021.