PDF: DIAS
(2005)Objective
To find a safe and effective dose of intravenous desmoteplase for acute ischemic stroke, administered in an extended 3-to-9-hour time window to patients selected by MRI perfusion/diffusion mismatch.
Study Summary
• A second part tested lower, weight-adjusted doses, found a good safety profile, particularly for the 125 μg/kg dose.
• The 125 μg/kg dose of desmoteplase significantly improved reperfusion (71.4% vs. 19.2% for placebo) and favorable 90-day clinical outcome (60.0% vs. 22.2% for placebo) with zero sICH in that group.
Intervention
Intravenous desmoteplase at various fixed (Part 1: 25, 37.5, 50 mg) and weight-adjusted (Part 2: 62.5, 90, 125 μg/kg) doses versus placebo, given as a bolus 3 to 9 hours after stroke onset in patients with ≥20% perfusion/diffusion mismatch on MRI. Follow-up was for 90 days. 104 randomized (47 Part 1, 57 Part 2); 102 analyzed (27 placebo, 75 desmoteplase).
Study Design
Arms: Array
Outcome
• The symptomatic intracranial hemorrhage (sICH) rate in Part 2 of the trial was low (2.2% across all desmoteplase doses) compared to the high rate in Part 1 (26.7%). No sICH occurred in the 125 μg/kg group or in any placebo group.
• Overall 90-day mortality was 1/27 (3.7%) with placebo and 9/75 (12.0%) with desmoteplase; 7 of the 9 desmoteplase deaths occurred in Part 1 (high-dose).
Bottom Line
Intravenous desmoteplase, when given 3 to 9 hours after stroke onset to patients selected by MRI perfusion/diffusion mismatch, is associated with higher rates of reperfusion and better clinical outcomes compared to placebo. A weight-adjusted dose of 125 μg/kg demonstrated significant efficacy with a low rate of symptomatic intracranial hemorrhage.
Major Points
- DIAS was a Phase II, randomized, double-blind, placebo-controlled, dose-finding trial that used MRI perfusion/diffusion mismatch (≥20%) to select patients for treatment in an extended time window of 3 to 9 hours.
- The trial was conducted in two parts. Part 1 tested high, fixed doses (25-50 mg) and was stopped early due to an unacceptably high rate of symptomatic intracranial hemorrhage (sICH) of 26.7%.
- Part 2 tested lower, weight-adjusted doses (62.5, 90, and 125 μg/kg) and found a much lower overall sICH rate of 2.2% in the desmoteplase groups.
- The 125 μg/kg dose of desmoteplase resulted in a significantly higher reperfusion rate at 4-8 hours (71.4%) compared to placebo (19.2%; P=0.0012).
- The 125 μg/kg dose also led to a significantly higher rate of favorable clinical outcome at 90 days (60.0%) compared to pooled placebo (22.2%; P=0.0090).
- Early reperfusion was strongly correlated with favorable clinical outcome at 90 days (P=0.0028).
Study Design
- Study Type
- Phase II, placebo-controlled, double-blind, randomized, dose-finding trial
- Randomization
- Yes
- Blinding
- Double-blind
- Sample Size
- 102
- Follow-up
- 90 days
- Centers
- 44
- Countries
- Germany, Switzerland, Spain, Austria, France, Singapore, Australia, Belgium, Finland, Norway, United Kingdom
Primary Outcome
Definition: The study had co-primary safety (symptomatic ICH) and efficacy (reperfusion at 4-8 h and 90-day clinical outcome) endpoints. The most effective and safe dose from Part 2 is highlighted here versus pooled placebo.
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| Pooled Placebo (n=27): sICH 0/27 (0.0%); reperfusion 5/26 (19.2%); favorable 90-day outcome 6/27 (22.2%) | Desmoteplase 125 μg/kg (best dose, n=15): sICH 0/15 (0.0%); reperfusion 10/14 (71.4%); favorable 90-day outcome 9/15 (60.0%) | - (sICH — pooled placebo 0.0% (95% CI 0.0-12.8), 125 μg/kg 0.0% (95% CI 0.0-21.8); reperfusion — pooled placebo 19.2% (95% CI 6.6-39.4), 125 μg/kg 71.4% (95% CI 41.9-91.6); favorable 90-day outcome — pooled placebo 22.2% (95% CI 8.6-42.3), 125 μg/kg 60.0% (95% CI 32.3-83.7)) | Reperfusion P=0.0012; favorable clinical outcome P=0.0090 (125 μg/kg vs pooled placebo) |
Limitations & Criticisms
- The trial was redesigned after Part 1 was halted due to high rates of symptomatic intracranial hemorrhage, resulting in two distinct parts with different dosing schemes.
- The sample size in each dose group was small, particularly in the dose-escalation design of Part 2, which limits statistical power.
- Baseline DWI lesion volumes were not well-balanced across the treatment groups in Part 2, which may have confounded the results for the higher-dose desmoteplase groups.
- The dose-escalation design of Part 2, while blinded, could have introduced bias in MRI interpretation because the likely dose group was known based on the study phase.
- Overall 90-day mortality was numerically higher with desmoteplase (12.0%) than placebo (3.7%), driven largely by the high-dose Part 1 deaths.
Citation
Stroke. 2005;36:66-73.