General Anesthesia Again Implicated in Poorer Outcomes With Endovascular Stroke Therapy

In patients receiving endovascular therapy for acute ischemic stroke, use of general anesthesia is associated with worse neurological outcomes and increased mortality, according to an analysis of the Interventional Management of Stroke (IMS) III trial, published online July 2, 2015, ahead of print in Stroke. 

The study is the latest of several—including an analysis of MR CLEAN presented at this year’s International Stroke Conference and the NASA registry study—to suggest that anesthesia type plays a role in endovascular stroke therapy outcomes.

“The IMS III dataset adds some of the strongest data yet on the negative association between general anesthesia and outcomes,” Alex Abou-Chebl, MD, of Baptist Health Louisville (Louisville, KY), said in an email with TCTMD.

Unlike MR CLEAN, IMS III was a negative study. Published in the New England Journal of Medicine in February 2013, it was halted after interim analysis showed no difference in benefit between acute ischemic stroke patients randomly assigned to receive t-PA with or without endovascular therapy.

For the new analysis, Dr. Abou-Chebl and colleagues examined outcomes for patients in the endovascular therapy arm of IMS III treated with local (n = 269) or general anesthesia (n = 147). The local and general anesthesia groups were similar for baseline characteristics except for lower National Institutes of Health Stroke Scale (NIHSS) scores (median 16 vs 18, respectively; P < .0001) and a slightly lower rate of internal carotid artery occlusion (P = .06) in the local group.

Negative Effects of General Anesthesia Emerge

Successful reperfusion (defined as Thrombolysis in Cerebral Infarction [TICI] score  ≥ 2) was similar between the general and local groups (76.4% vs 72.8%; P = .48). 

However, patients who received general anesthesia were less likely to achieve a good clinical outcome (modified Rankin Scale score of ≤ 2 at 90 days) and had higher rates of in-hospital death, differences that remained after adjustment for stroke severity  (NIHSS ≤ 19 vs NIHSS ≥ 20), age, and time from onset to groin puncture. The apparent disadvantage of general anesthesia in terms of more subarachnoid hemorrhage and symptomatic intracerebral hemorrhage was erased after adjustment for use of mechanical embolectomy (table 1).

The drawbacks to general anesthesia were most apparent in the subset of general anesthesia patients with medical indications for intubation compared with those who received local anesthesia. Medically indicated patients had lower probability of a good outcome (adjusted RR 0.49; 95% CI 0.30-0.81) and higher mortality (adjusted RR 3.93; 95% CI 2.18-7.10) than the local anesthesia group, whereas each of these endpoints were equivalent between patients receiving routine general anesthesia and local anesthesia. Those with medical indications for general anesthesia also had higher risk of in-hospital mortality than those who received general anesthesia as part of routine practice (adjusted RR 2.16; 95% CI 1.09-4.29).

‘Difficult to Justify’ General Anesthesia

“In the context of MR CLEAN and the NASA registry study, IMS III emphasizes that the outcomes associated with general anesthesia appear to be device independent,” Dr. Abou-Chebl said in his email.

He added that the lack of increase in intracranial hemorrhage with local anesthesia in IMS III is an important finding. “[O]ne of the reasons that general anesthesia is used is because of a [what we now know to be an] unfounded fear that awake patients will move around, resulting in wire- or device-induced perforation of 1 of the cerebral vessels, thus causing intracerebral hemorrhage,” Dr. Abou-Chebl explained.  

One weakness of all 3 analyses, though, is that “a direct causal relationship has not been proven because the reasons for using general anesthesia are not standardized in the studies and vary widely across centers and operators,” Dr. Abou-Chebl said.

“However, given that there are no published data (that I am aware of) showing that general anesthesia is of benefit in the setting of endovascular stroke therapy and that the retrospective and prospective data have all shown a consistent negative association between general anesthesia and neurological outcomes and mortality,” he continued, “it is difficult to justify or support general anesthesia except in patients who have a clear indication (eg, respiratory arrest, severe hypoxemia, etc).”

Dr. Abou-Chebl said a randomized trial is needed to confirm the association of anesthesia type with outcomes since “retrospective datasets sometimes prove misleading.”

Multiple Mechanisms Suggested

Also unclear are the mechanisms by which general anesthesia may contribute to poorer outcomes.

One theory implicates hemodynamic perturbations—especially hypotension—when anesthesia is begun, the investigators say. Another possibility is that general anesthesia may mask neurological deterioration or headache during the procedure. Additionally, the MR CLEAN analysis found that general anesthesia is associated with delayed initiation of treatment; in the IMS III substudy and in 2 retrospective studies, however, no such association was seen, Dr. Abou-Chebl said.

That seems counterintuitive, since it is known that induction of anesthesia takes time, he acknowledged. The lack of delay “could be due to the fact that the centers in these trials may have protocols in place to avoid delays or they may have access to more personnel because of the presence of the anesthesia team, [which] could result in faster patient preparation. Another possibility is that centers that do not use [general] anesthesia move slower because they feel that they have more time.”

Marker Rather Than Predictor?

William A. Gray, MD, of Columbia University Medical Center (New York, NY), told TCTMD in a telephone interview that while the latest analysis is well done, “it is not clear from this retrospective study that general anesthesia itself causes some decline in neurologic outcomes during endovascular treatment of acute stroke.”

Additionally, Dr. Gray said, since IMS III used older recanalization devices and since the outcomes in general were not positive, “you may be dealing with a cohort of patients across the board who would not be as predictive for any outcome that you wish to look at with regard to endovascular stroke therapy as patients in any of the 5 recent positive trials.”

Without randomized data, he said, a logical conclusion could be that patients who had general anesthesia may have been fundamentally different from those who underwent local anesthesia. “So, general anesthesia may be more of a marker rather than a predictor of worse outcomes in these patients,” Dr. Gray said.

While there are possible mechanisms by which general anesthesia may contribute to worse outcomes, he acknowledged, these cannot be used to conclusively explain the results.

The study findings are “hypothesis generating and probably need to be tested in a randomized trial, especially since we are doing more and more endovascular stroke interventions,” he concluded.

Source: 
Abou-Chebl A, Yeatts SD, Yan B, et al. Impact of general anesthesia on safety and outcomes in the endovascular arm of Interventional Management of Stroke (IMS) III trial. Stroke. 2015;Epub ahead of print. 

Disclosures:

• IMS III was supported by the National Institute of Neurological Disorders and Stroke. 
• Drs. Abou-Chebl and Gray report no relevant conflicts of interest. 

Related Stories:

• Meta-analysis Supports Endovascular Therapy for Acute Stroke
• MR CLEAN Subanalyses Show Better Acute Stroke Outcomes With Local Anesthesia, Shorter Reperfusion Times
• Stroke Registry: Better Outcomes with Local Rather than General Anesthesia in Mechanical Embolectomy