Visual P300 latency predicts treatment response to modafinil in patients with narcolepsy

Objective: To evaluate the hypothesis that visual P300 latency (VL) predict s treatment response to modafinil (a new wake-promoting agent) in patients with narcolepsy. Methods: Design: Comparison of responders and non-responders in a double-bl ind randomized placebo-controlled trial. Setting: Private practice referral sleep disorders center. Patients: Twenty one patients with narcolepsy (age s 17-65 years). Interventions: Auditory and visual P300 testing using 31 ev enly spaced scalp electrodes, and baseline polysomnograms and objective and subjective tests of daytime sleepiness, followed by modafinil treatment fo r 9 weeks. Polysomnograms and tests of sleepiness were then repeated. Main outcome measure: The Maintenance of Wakefulness Test (MWT). Response define d as a final MWT > 7.3min (normative sample mean - 3 SD), plus an increase > 1SD based on normative sample (3.6 min) over baseline MWT. Results: Non-responders had longer age-adjusted 31-electrode mean VL (448.4 ms vs. 410.8 ms, P = 0.024), and larger auditory P300 amplitude, with no t opographical P300 differences. Non-responders and responders did not differ on any other baseline clinical variable. Using a cut-off of 0.5 SE from no rmal regression constant, shorter age-adjusted VL predicted modafinil respo nse, with specificity of 0.71 and sensitivity of 0.86. Conclusions: VL predicts treatment response to modafinil in patients with n arcolepsy. (C) 1999 Elsevier Science Ireland Ltd. All rights reserved.