ASA: No Benefit Found for Post-Stroke Tight Glucose Control

SAN FRANCISCO, Feb. 12 -- Tight glucose control during hospitalization for a stroke may not improve survival, British researchers said here.

Glucose, potassium and insulin infusions for stroke patients with mild to moderately elevated glucose levels yielded no mortality advantage compared with placebo, said Christopher S. Gray, M.D., of Newcastle University in Sunderland, England, at the American Stroke Association meeting.

"What our results mean in practice for the average stroke physician is that for the average patient there is no need to try to intensively lower blood glucose levels," he said.

Furthermore, there may be a suggestion of harm because 90-day mortality was actually slightly higher among patients who received the infusions (odds ratio 1.14, 95% confidence interval 0.086 to 1.51) and patients who died had a higher mean reduction in glucose than the overall group, Dr. Gray said.

"Physicians need to exert a degree of caution that intensive lowering of glucose lowering may indeed be harmful and we need await further studies before we advocate intensive lowering of blood glucose levels," he added.

These findings counter those of the large intensive care unit studies showing a definite mortality benefit from tight glucose control, said Philip Gorelick, M.D., M.P.H., of the University of Illinois at Chicago, who moderated a press conference where the study was discussed.

He said many hospitals are adopting new practice guidelines to initiate insulin infusion in the ICU to maintain serum glucoses within a narrow range, and a single study is unlikely to change this practice for stroke patients.

"I don't think one study should necessarily stop the show," Dr. Gorelick said. "We are going to need further validation."

However, Dr. Gray said that while there is good evidence for a mortality benefit in certain subgroups of patients, such as myocardial infarction and acute illness, there is no evidence to support acute management of high blood sugar following stroke.

Their Glucose Insulin in Stroke Trial (GIST-UK) trial is the first to look at the issue, he said.

The study included 933 patients enrolled within 24 hours of a stroke who had plasma glucose in the range of 6.0 to 17 mmol/l. Patients with diabetes mellitus, coma, coronary heart failure or a Rankin score above three were excluded.

Participants were randomized to receive saline solution or continuous glucose, potassium, insulin infusions aiming to reduce plasma glucose to 4 to 7 mmol/L. Patients were monitored every two hours with glucose adjusted if needed every eight hours. Plasma glucose at admission was similar between groups.

The researchers found that both treatment and placebo groups had improvement in glucose levels. The treatment group had an overall mean 0.57 mmol/l reduction in plasma glucose over 24 hours while glucose levels also fell spontaneously with simple saline hydration.

"In the majority of patients, treatment with a simple saline infusion will correct mild to moderate hyperglycemia," Dr. Gray said.

The primary outcome measure of death at 90 days was not significantly different between groups (30.0% versus 27.3% placebo, P=0.37). Kaplan-Meier survival curves were similar at 90 days as well.

There was also no difference in the secondary outcome of disability. The odds ratio findings were:

• 0.84 for Barthel Index score less than nine (95% CI 0.59 to 1.20), and
• 0.96 for modified Rankin score greater than three excluding death (95% CI 0.70 to 1.32).

Interestingly, there was a significant 9 mm Hg reduction in systolic blood pressure in the treatment group though it unclear what this would mean, Dr. Gray said.

He noted that the study was powered to show a 6% reduction in mortality at 90 days with 83% power with the planned sample size of 2,355. The trial was stopped before accrual reached that level because of slow recruitment that would have made the findings outdated by the time the trial was completed, he said.