← Back
NeuroTrials.ai
Neurology Clinical Trial Database

VA Cooperative Study

Comparison of Carbamazepine, Phenobarbital, Phenytoin, and Primidone in Partial and Secondarily Generalized Tonic-Clonic Seizures

Share Share Copied! Compare

Year of Publication: 1985

Authors: Mattson RH, Cramer JA, Collins JF, ..., McCutchen CB

Journal: New England Journal of Medicine

Citation: Mattson RH et al. N Engl J Med. 1985;313(3):145-151. DOI: 10.1056/NEJM198507183130303

Link: https://www.nejm.org/doi/10.1056/NEJM198507183130303?

PDF: https://www.nejm.org/doi/pdf/10.1056/NEJM198507183130303

Clinical Question

Which of the four major first-generation antiepileptic drugs -- carbamazepine, phenobarbital, phenytoin, or primidone -- provides the best balance of efficacy and tolerability for partial and secondarily generalized tonic-clonic seizures?

Bottom Line

Carbamazepine had the best overall composite score combining seizure control and tolerability (p<0.002). All four drugs were similarly effective for secondarily generalized tonic-clonic seizures, but carbamazepine was significantly superior for complex partial seizures. Primidone had the worst tolerability with the highest early dropout rate, while carbamazepine was best tolerated with fewer effects on cognition and motor function. This landmark VA study established carbamazepine as the first-line drug for focal epilepsy.

Major Points

  • First large RCT comparing 4 AEDs for new-onset seizures: CBZ, PHT, PB, PRM.
  • For partial seizures: CBZ and PHT had fewest treatment failures; CBZ best tolerated overall.
  • For generalized tonic-clonic: all 4 drugs equally effective; PB and PRM caused more side effects.
  • 622 adults randomized across VA hospitals. Double-blind, 1975-1983.
  • CBZ dose: up to 1800 mg/day; PHT up to 400 mg/day; PB up to 240 mg/day; PRM up to 2000 mg/day.
  • Retention rate (drug continuation) highest for CBZ (65%) and PHT (62%), lowest for PRM (51%) and PB (56%).
  • Established CBZ as first-line for partial epilepsy — standard for 20+ years until SANAD.
  • Key finding: tolerability drives long-term treatment success more than seizure control alone.
  • Published Epilepsia 1985 (Mattson et al.). Multi-part analysis: partial and GTC separate publications.
  • All arms had similar seizure freedom rates (~50% at 1 year) — difference was in side effect burden.

Design

Study Type: Prospective, multicenter, randomized, double-blind, parallel-group trial (VA Cooperative Study No. 118)

Randomization: 1

Blinding: Double-blind (identical-appearing capsules prepared by VA pharmacy)

Enrollment Period: Approximately 1975-1980

Follow-up Duration: 2 years

Centers: 10

Countries: United States

Sample Size: 622

Analysis: Intention-to-treat; stratified by seizure type

Inclusion Criteria

  • Adults (age 18 and older) with partial seizures (simple partial, complex partial) and/or secondarily generalized tonic-clonic seizures
  • Previously untreated or undertreated (inadequate prior monotherapy)
  • At least 2 seizures in the 6 months preceding enrollment
  • Able to attend regular clinic visits at VA centers

Exclusion Criteria

  • Drug or alcohol abuse (strict exclusion)
  • Progressive neurological disease
  • Primary generalized epilepsy (absence, myoclonic)
  • Seizures due to correctable metabolic or systemic causes
  • History of poor compliance
  • Significant hepatic or renal impairment

Baseline Characteristics

All Groups (combined):

  • N: 622 (~155 per group)
  • Sex - Male: ~87-88% (VA population)
  • Mean age: ~40-42 years (range 18-82)
  • Seizure type: 100% focal-onset
  • Mean epilepsy duration: ~5-6 years

Arms

FieldCarbamazepine (CBZ)Phenobarbital (PHB)Phenytoin (PHT)Primidone (PRM)
InterventionCarbamazepine monotherapy, target plasma levels 4-12 mcg/mLPhenobarbital monotherapy, target plasma levels 15-40 mcg/mLPhenytoin monotherapy, target plasma levels 10-20 mcg/mLPrimidone monotherapy, target plasma levels 5-15 mcg/mL (with derived phenobarbital levels)
Duration2 years or until treatment failure2 years or until treatment failure2 years or until treatment failure2 years or until treatment failure

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Overall composite score combining seizure control and toxicityPrimaryLowest composite: Primidone; Intermediate: PhenobarbitalHighest composite: Carbamazepine and Phenytoin<0.002 (overall comparison)
Secondary
Secondary
Secondary
Secondary
Primidone acute toxicityAdverseN/AHighest: nausea, vomiting, dizziness, severe sedation, decreased libido/impotence
PhenobarbitalAdverseN/ASedation, cognitive slowing, depression, decreased libido; lowest motor disturbances
PhenytoinAdverseN/AGingival hyperplasia, hirsutism, facial coarsening, rash, cerebellar effects
CarbamazepineAdverseN/ABest tolerated overall; some GI effects, hyponatremia, rash; rare blood dyscrasias

Subgroup Analysis

By seizure type: drugs equivalent for SGTC but CBZ superior for CPS (key finding); by prior treatment status

Criticisms

  • Overwhelmingly male population (87-88%): severe limitation on generalizability to women
  • Older mean age (~40 years): does not represent younger adults or children with epilepsy
  • No valproate arm (addressed in follow-up VA Cooperative Study 264, Mattson 1992)
  • Composite outcome combining efficacy and tolerability is difficult to interpret; hard to separate effects
  • High early dropout with primidone may have been due to overly rapid dose titration rather than inherent drug intolerance
  • 2-year follow-up may not capture long-term differences in efficacy or toxicity
  • Only first-generation drugs studied; less relevant to modern prescribing with newer AEDs

Funding

U.S. Department of Veterans Affairs (VA Cooperative Studies Program) -- non-industry

Based on: VA Cooperative Study (New England Journal of Medicine, 1985)

Authors: Mattson RH, Cramer JA, Collins JF, ..., McCutchen CB

Citation: Mattson RH et al. N Engl J Med. 1985;313(3):145-151. DOI: 10.1056/NEJM198507183130303

Content summarized and formatted by NeuroTrials.ai.