We studied 348 (231 subjects from the CLOTBUST databank and 117 subjects from the CLOTBUST trial phase II) with documented intracranial artery occlusion treated with intravenous recombinant
tissue-plasminogen activator. Serum
glucose was determined at baseline before intravenous recombinant
tissue-plasminogen activator administration.
Hyperglycemia was defined as a
glucose level ≥140 mg/dl (7·7 mmol/l). Transcranial Doppler findings were interpreted using the thrombolysis in
brain ischemia flow grading system as persistent
arterial occlusion, re-occlusion or complete recanalization. Poor clinical outcome was defined by modified Rankin score > 2 at three-months.
RESULTS: At baseline, 138 patients (37·4%) were hyperglycemic and 210 patients (56·9%) normoglycemic. Baseline characteristics based on glucose ≥ 140 (7·7 mmol/l) or less 140: age (70·0 ± 12·4 vs. 67·3 ± 14·1, P = 0·065), baseline National Institutes of Health
Stroke Scale (17·0 ± 5·5 vs. 15·8 ± 5·5, P = 0·054), time to recombinant
tissue-plasminogen activator (141·4 ± 69·1 vs. 145·3 ± 48·4 mins, P = 0·56), and history of
diabetes mellitus [60/138 (43·5%) vs. 22/210 (10·5%), P < 0·001]). Patients with
hyperglycemia have a higher rate of persisting occlusion [72/138 (52·2%) vs. 66/210 (31·4%)] and less rate of complete recanalization [34/138 (24·6%) vs. 82/210 (39%), P < 0·001]. Median time to recanalization in patients with severe
hyperglycemia (glucose ≥ 200) (11 mmol/l) and
glucose <200 was 163 ± 79 and 131 ± 90 mins, respectively (P = 0·045). Sixteen patients (11·6%) in the hyperglycemic group and 12 (5·7%) in the normoglycemic group had symptomatic
intracerebral hemorrhage (P = 0·049). Seventy-eight patients (69%) in the
hyperglycemia group and 72 patients (41·6%) in the normoglycemic group had poor clinical outcome (three-month modified Rankin score > 2) (P ≤ 0·001). After adjusting for
stroke risk factors, patients with
hyperglycemia were more likely to have poor clinical outcome (three-month modified Rankin score > 2) (adjusted odds ratio = 2·22, 95% confidence interval: 1·2-4·11, P = 0·011) and low complete recanalization rate (adjusted odds ratio: 0·5, confidence interval: 0·3-0·92, P = 0·025) with trend of increase risk of symptomatic
intracerebral hemorrhage (adjusted odds ratio: 2·07, confidence interval:0·8-5·1, P = 0·114). There was no association between baseline
glucose as a continuous variable and poor clinical outcome, but there was with the complete recanalization's rate.
CONCLUSION: