Phase II Studies Show Investigational Insomnia Agent Gaboxadol Improved Deep Sleep and Key Sleep Measures in Adults with Primary and Transient Insomnia

Results from phase II studies presented this weekdemonstrated the effects of gaboxadol on sleep architecture (visualpolysomnograph and computerized analysis of the waveform through thestages of sleep) in both primary and transient insomnia. In a model oftransient insomnia the traditional measures of sleep onset andmaintenance were also reported. Gaboxadol, an investigational agentunder clinical development by Merck & Co., Inc. and H. Lundbeck A/Sfor the treatment of insomnia, is a first-in-class selectiveextrasynaptic GABA(A) receptor agonist (SEGA). The data were presentedat the SLEEP 20th Anniversary Meeting of the Associated ProfessionalSleep Societies (APSS).

In the clinical studies presented in primary and transientinsomnia, electroencephalogram (EEG) analyses showed that gaboxadol(5, 10, 15 and 20 mg) caused statistically significant increases indeep sleep, i.e. visually scored slow wave sleep or objectivelymeasured slow wave activity (SWA) as compared to placebo (5 mg:p<0.05; 10, 15 and 20 mg: p<0.001; respectively). Slow waves are thecharacteristic and predominant waveform during sleep stages 3 and 4.Combined sleep stages 3 and 4 are commonly known as slow wave sleep.Slow wave sleep and slow wave activity are associated with deep and/orrestorative sleep. Slow wave activity refers to a low frequency bandon the EEG and is an objectively scored measure of the slow wave sleepprocess. Zolpidem 10 mg was used as an active reference for thesestudies. These studies were not designed to compare the use ofgaboxadol versus zolpidem.

"These data showed that gaboxadol consistently increased theintensity of, and time spent in, deep or restorative sleep," saidStephen Deacon, Ph.D., a lead investigator and head of ClinicalDevelopment, Sleep Disorders, Lundbeck U.K. "The research helps tosupport gaboxadol's potential as a new and different treatment forinsomnia."

Gaboxadol improved deep sleep in primary insomnia patients

Power spectral analysis was performed on data collected from thenon-rapid eye movement (NREM) EEG data collected from two separaterandomized, double-blind, cross-over polysomnograph (PSG) studies tocompare the dose response characteristics of gaboxadol to placeboafter two nights of treatment. Patients aged 18-65 years diagnosedwith primary insomnia (DSM-IV criteria) were included in both of thestudies. Study 1 evaluated 40 patients treated with gaboxadol (10 and20 mg), zolipdem (10 mg) or placebo in a 4-way cross-over trial. Study2 evaluated 26 patients treated with gaboxadol (5 and 15 mg) andplacebo in a 3-way cross-over trial.

Power spectral analyses were performed on the NREM EEG on thesecond night of treatment (using 38 patients for study 1 and 23patients for study 2 per protocol). Night 1 was used as an adaptationnight to avoid first night effects. Gaboxadol (10, 15 and 20 mg)increased spectral power in the lower frequency bands, defined as slowwave activity and theta activity - in a dose dependent manner,compared to placebo (p<0.05). Theta activity is another frequency bandadjacent to SWA seen on the EEG, however, the significance of thetaactivity is unknown.

In Study 1, gaboxadol 10 and 20 mg, when compared to placebo,increased slow wave activity (0.5 - 3.5Hz), by a rate of 33 and 54percent, respectively (p<0.01) and increased theta activity (4 -7.5Hz) by a rate of 24 and 34 percent, respectively (p<0.01 andp<0.001) relative to placebo, but did not have any statisticallysignificant effect on the alpha, beta 1, beta 2, and spindle bands.

In Study 2, gaboxadol 15 mg increased slow wave activity by 21percent relative to placebo (p<0.05), and increased theta activity by20 percent, relative to placebo (p<0.001) with no significant effectat the 5 mg dose.(1) No effect was seen on the other measures (i.e.alpha, beta 1, beta 2, and spindle bands) for either of these twodoses.

Safety was not an endpoint for this analysis. However, in apreviously reported presentation of data for 40 patients included inthis analysis, the most common adverse events included tachycardia,headache, nausea, vomiting and fatigue. Gaboxadol (10 mg and 20 mg)was generally well tolerated. No patients had serious adverse events,and no patients withdrew from the study due to adverse events.

Gaboxadol improved deep sleep and key sleep measures in transientinsomnia

A randomized, double-blind, 5-way cross-over study of 109 healthysubjects aged 18-58 years was conducted to evaluate the efficacy ofgaboxadol (5, 10 and 15 mg) in a model of transient insomnia. Theprimary objective of the study was to evaluate the efficacy ofgaboxadol on traditional self-reported and visually scored PSG sleepmeasures when compared with placebo in a model of transient insomnia.Secondary objectives were to evaluate measures of slow-wave sleep(SWS), residual sedative effects and safety and tolerability. Sleeparchitecture was also assessed using EEG power spectral analysis onNREM sleep.

On self-reported measures, subjects who were administeredgaboxadol 10 and 15 mg reported improvements in the time it took themto fall asleep (sTSO) (34 percent, 29.2 minutes; 27 percent, 32.2minutes; respectively, both p<0.05), compared with placebo (44.4minutes). Gaboxadol 5 mg did not demonstrate a significant effect onsTSO. Patients treated with all three doses of gaboxadol (5, 10 and 15mg) reported significant improvement in the overall time spent asleep(sTST) (5 mg: 6 percent, 414.8 minutes; 10 mg: 7 percent, 419.2minutes; and 15 mg: 9 percent, 426.8 minutes; all p<=0.05) compared toplacebo (391.2 minutes) and the amount of time they spent awakefollowing sleep onset (sWASO) (5 mg: 47 percent, 20.8 minutes; p<0.01;10 mg: 31 percent, 27.2 minutes; p<=0.05; and 15 mg: 37 percent, 24.9minutes; p<0.01) compared to placebo (39.5 minutes). Patients alsoreported that they woke up significantly fewer times throughout thenight (sNAW) (5 mg: 32 percent, 1.9 times; 10 mg: 39 percent, 1.7times; and 15 mg: 36 percent, 1.8 times) compared to placebo (2.8times).

On polysomnographic measures of sleep efficacy, gaboxadol 10 and15 mg significantly decreased the time it took for participants tofall asleep by 31 and 33 percent (latency to persistent sleep, 10 mg:27.0 minutes; p<0.05 and 15 mg: 26.1 minutes; p<0.01), compared withplacebo (39.2 minutes). Gaboxadol at any dose did not significantlyaffect the number of awakenings, relative to placebo. All three dosessignificantly increased total sleep time (5mg: 8 percent, 393.1minutes, 10 mg: 10 percent, 398.2 minutes, and 15 mg: 8 percent, 393.1minutes; all p<0.001) and decreased wakefulness after sleep onset(5mg: 25 percent, 56.8 minutes, p<0.01; 10 mg: 27 percent, 55.0minutes; p<0.01; and 15 mg: 22 percent, 58.8 minutes; p<0.05) comparedto placebo (75.8 minutes).

Gaboxadol significantly increased spectral power in the lowerfrequency bands defined as slow wave and theta activity (5 mg, 6percent, p<=0.05; 10 mg, 19 percent, p<0.001 and, 15 mg, 27 percent,p<0.001 compared to placebo), and increased visually scored slow wavesleep (5 mg, 12 percent, p<=0.05; 10 mg, 37 percent; 15 mg, 42percent; both p<0.001, compared to placebo).

Treatment with gaboxadol was generally well tolerated in the studywith the majority of adverse events mild to moderate in nature. Therewere no serious adverse events and no adverse events that lead towithdrawal from the study. The most common adverse events occurring inpatients treated with gaboxadol were nausea and dizziness. There wereno consistent next day residual effects observed in either thegaboxadol or placebo arms of the study.

About gaboxadol

Gaboxadol is a novel compound currently in phase III developmentfor the treatment of insomnia. Merck and Lundbeck are collaborators inthe clinical development and commercialization of gaboxadol.

About Merck

Merck & Co., Inc. is a global research-driven pharmaceuticalcompany dedicated to putting patients first. Established in 1891,Merck currently discovers, develops, manufactures and markets vaccinesand medicines to address unmet medical needs. The Company devotesextensive efforts to increase access to medicines through far-reachingprograms that not only donate Merck medicines but help deliver them tothe people who need them. Merck also publishes unbiased healthinformation as a not-for-profit service. For more information, visitwww.merck.com.

About Lundbeck

H. Lundbeck A/S is an international pharmaceutical company engagedin the research and development, production, marketing and sale ofdrugs for the treatment of psychiatric and neurological disorders. In2005, the company's revenue was DKK 9.1 billion (approx. EUR 1.2billion). The number of employees is approx. 5,000. For furtherinformation, visit www.lundbeck.com

Merck Forward-Looking Statement

This press release contains "forward-looking statements" as thatterm is defined in the Private Securities Litigation Reform Act of1995. These statements are based on management's current expectationsand involve risks and uncertainties, which may cause results to differmaterially from those set forth in the statements. The forward-lookingstatements may include statements regarding product development,product potential or financial performance. No forward-lookingstatement can be guaranteed, and actual results may differ materiallyfrom those projected. Merck undertakes no obligation to publiclyupdate any forward-looking statement, whether as a result of newinformation, future events, or otherwise. Forward-looking statementsin this press release should be evaluated together with the manyuncertainties that affect Merck's business, particularly thosementioned in the cautionary statements in Item 1 of Merck's Form 10-Kfor the year ended Dec. 31, 2005, and in its periodic reports on Form10-Q and Form 8-K, which the Company incorporates by reference.

(1) These values are calculated based on dose minus placebodivided by placebo