Comparison of outcomes following thrombolytic therapy among patients with prior stroke and diabetes in the Virtual International Stroke Trials Archive (VISTA).

OBJECTIVE

The use of alteplase in patients who have had a prior stroke and concomitant diabetes is not approved in Europe. To examine the influence of diabetes and prior stroke on outcomes, we compared data on thrombolysed patients with nonthrombolysed comparators.

RESEARCH DESIGN AND METHODS

We selected patients with ischemic stroke on whom we had data on age, pretreatment baseline National Institutes of Health Stroke Scale (b-NIHSS), and 90-day outcome measures (functional modified Rankin score [mRS]) and neurological measures [NIHSS]) in the Virtual International Stroke Trials Archive. We compared outcomes between thrombolysed patients and nonthrombolysed comparators in those with and without diabetes, those who have had a prior stroke, or both and report findings using the Cochran-Mantel-Haenszel (CMH) test and proportional odds logistic regression analyses. We report an age-adjusted and b-NIHSS-adjusted CMH P value and odds ratio (OR).

RESULTS

Rankin data were available for 5,817 patients: 1,585 thrombolysed patients and 4,232 nonthrombolysed comparators. A total 1,334 (24.1%) patients had diabetes, 1,898 (33.7%) patients have had a prior stroke, and 491 (8%) patients had both. Diabetes and nondiabetes had equal b-NIHSS (median 13; P = 0.3), but patients who have had a prior stroke had higher b-NIHSS than patients who have not had a prior stroke (median 13 vs. 12; P < 0.0001). Functional outcomes were better for thrombolysed patients versus nonthrombolysed comparators among both nondiabetic (P < 0.0001; OR 1.4 [95% CI 1.3-1.6]) and diabetic (P = 0.1; 1.3 [1.05-1.6 ]) subjects. Similarly, outcomes were better for thrombolysed patients versus nonthrombolysed comparators among who have not had a prior stroke (P < 0.0001; 1.4 [1.2-1.6 ]) and those who have (P = 0.02; 1.3 [1.04-1.6 ]). There was no interaction of diabetes and prior stroke with treatment (P = 0.8). Neurological outcomes were consistent with the mRS.

CONCLUSIONS

Outcomes from thrombolysis are better among patients with diabetes and/or those who have had a prior stroke than in control subjects. Withholding thrombolytic treatment from otherwise-eligible patients may not be justified.