Occurrence, outcome, and prognostic factors of infantile spasms and Lennox-Gastaut syndrome.

PURPOSE

To analyze the occurrence, outcome, and prognostic factors of infantile spasms (IS) and the Lennox-Gastaut syndrome (LGS) in a defined population.

METHODS

All children treated because of IS and LGS in the Department of Pediatrics, University of Oulu, from January 1, 1976, until December 31, 1993, who came from the primary catchment area of the hospital were included. Detailed information concerning their individual pre-, peri-, and postnatal medical histories and medical and laboratory examinations were compiled.

RESULTS

Thirty-seven children (18 boys) had IS, and 25 (14 boys) had LGS. The occurrence of IS of 0.41/1,000 live births 195% confidence interval (CI), 0.29-0.57/1,000] did not differ significantly from that of LGS, which was 0.28/1,000 live births (95% CI, 0.18-0.41/1,000). Ten (27%) of the 37 patients with IS evolved to LGS, which was 40% of the LGS cases. All the 10 children with both IS and LGS had symptomatic epilepsy, were mentally retarded, and had active epilepsy at the end of approximately 10 years' follow-up. Twenty-six (87%) of the 30 symptomatic IS cases and all the 17 symptomatic LGS cases were due to either congenital or genetic etiologies. The outcome in cryptogenic IS cases was favorable; the risk for a poor neurologic and mental outcome was extremely low; odds ratio, 0.015 (95% Cl, 0.001-0.196), as it was for therapy-resistant epilepsy; odds ratio, 0.013 (95% CI, 0.001-0.166). In LGS patients, cryptogenic etiology did not decrease the risk for a poor outcome.

CONCLUSIONS

Cryptogenic etiology is associated with a very low risk for a poor outcome in IS patients, but not in LGS patients. The outcome of IS children and the relation of IS to LGS are determined by the underlying brain disease, not by the epilepsy itself.