PDF: Treiman SE
(1998)Objective
To compare four IV treatment regimens for generalized convulsive status epilepticus: lorazepam, phenobarbital, diazepam+phenytoin, and phenytoin alone
Study Summary
• ITT analysis did NOT reach significance (p=0.12); subtle SE poorly responsive to all treatments (7.7-24.2%, p=0.91); no differences in AEs, recurrence, or 30-day outcomes; established IV lorazepam as standard first-line hospital SE treatment
Intervention
Lorazepam 0.1 mg/kg vs Phenobarbital 15 mg/kg vs Diazepam 0.15 mg/kg + Phenytoin 18 mg/kg vs Phenytoin 18 mg/kg alone (all IV)
Inclusion Criteria
Patients with generalized convulsive status epilepticus (overt or subtle)
Study Design
Arms: Lorazepam vs Phenobarbital vs Diazepam + Phenytoin vs Phenytoin Alone
Patients per Arm: ~140 per group (570 total; 384 overt GCSE in primary analysis)
Outcome
• ITT analysis: p=0.12 (NOT significant); Subtle SE: 7.7-24.2% all arms (p=0.91)
• No differences in adverse reactions, recurrence at 12h, or 30-day outcomes among any treatments
Bottom Line
For overt generalized convulsive SE, lorazepam was the most effective first-line agent (64.9% success rate), significantly superior to phenytoin alone (43.6%, p=0.002). The overall 4-way comparison was significant (p=0.02). However, the intention-to-treat analysis did not reach significance (p=0.12), and none of the treatments were effective for subtle SE (7.7-24.2% success). This trial established IV lorazepam as the standard first-line treatment for convulsive SE in hospital settings.
Major Points
- Lorazepam most effective for overt GCSE: 64.9% success vs phenytoin 43.6% (P=0.002), phenobarbital 58.2%, diazepam+phenytoin 55.8%.
- 4-way comparison not significant (P=0.12 overall) — but lorazepam vs phenytoin was significant.
- For subtle SE: no significant difference between agents (success 15-24% across all 4 arms).
- Lorazepam dose: 0.1 mg/kg IV at 2 mg/min (max 4 min).
- Phenytoin: 18 mg/kg IV at 50 mg/min. Phenobarbital: 15 mg/kg IV at 100 mg/min.
- Diazepam+phenytoin: DZP 0.15 mg/kg IV at 5 mg/min, then PHT 18 mg/kg.
- 384 episodes of SE in 570 screened; 518 randomized. Mean age 58. VA hospitals 1990-1995.
- Defined overt GCSE and subtle GCSE as distinct entities — landmark classification.
- Treatment success: seizure cessation within 20 min without recurrence in 60 min.
- Established lorazepam as first-line for GCSE — remains standard of care 25+ years later.
Study Design
- Study Type
- Randomized, double-blind, multicenter trial (VA Cooperative Studies Program)
- Randomization
- Yes
- Blinding
- Double-blind (double-dummy design)
- Sample Size
- 570
- Follow-up
- 12 hours for recurrence; 30 days for outcomes
- Centers
- Multiple VA medical centers
- Countries
- United States
Primary Outcome
Definition: Treatment success: cessation of all motor and EEG seizure activity within 20 minutes, no recurrence for 40 minutes (overt GCSE, per-protocol)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| Phenytoin alone: 43.6% | Lorazepam: 64.9%; Phenobarbital: 58.2%; Diazepam+Phenytoin: 55.8% | - | Overall: p=0.02; Lorazepam vs Phenytoin: p=0.002; ITT overall: p=0.12 (not significant) |
Limitations & Criticisms
- ITT analysis did NOT reach significance (p=0.12); primary per-protocol finding depends on excluding 52 patients with unverified SE
- VA population predominantly male and older, limiting generalizability to women and younger patients
- Very low success rates in subtle SE (7.7-24.2%) with no guidance on alternative treatments
- 20-minute definition for treatment success may disadvantage phenytoin (requires slower infusion rate)
- 5-year enrollment period during which SE treatment may have evolved
- Lacked standardized second-line protocol after treatment failure
- Only 384 patients in primary analysis group, limiting statistical power for 4-way comparison
Citation
Treiman DM et al. N Engl J Med. 1998;339(12):792-798. DOI: 10.1056/NEJM199809173391202