The MADRS was utilized as the primary efficacy assessment of a participants level of depression. The MADRS consisted of 10 items, all rated on a 0 to 6 scale with 0 being the "best" rating and 6 being the "worst" rating. The MADRS Total Score is the sum of ratings for all 10 items; therefore, possible total scores range from 0 to 60. The MADRS total score were to be unevaluable if less than 8 of the 10 items were recorded. If 8 or 9 of the 10 items were recorded, the MADRS total score was the mean of the recorded items multiplied by 10 and then rounded of to the first decimal place.

The efficacy sample was the Full Analysis Set (FAS) comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The SDS was a self-rated instrument used to measure the effect of the participants symptoms on work/school, social life, and family/home responsibilities. For each of the three items, scores ranged from 0 through 10. The number most representative of how much each area was disrupted by symptoms was marked along the line from 0= not at all, to 10= extremely. Scores of 5 and above are associated with significant functional impairment. The SDS total score ranges from 0 to 30, with higher values indicating greater disruption in the participant's work/social/family life. For the work/school item, no response was to be entered if the participant did not work or go to school for reasons unrelated to the disorder and a response therefore not being applicable. The Mean SDS score were calculated over the three item scores. All three item scores were needed to be available with the exception of the work/school item score when this item was not applicable.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The MADRS was utilized as the primary efficacy assessment of a participants level of depression. The MADRS consisted of 10 items, all rated on a 0 to 6 scale with 0 being the "best" rating and 6 being the "worst" rating. The MADRS Total Score is the sum of ratings for all 10 items; therefore, possible total scores range from 0 to 60. The MADRS total score were to be unevaluable if less than 8 of the 10 items were recorded. If 8 or 9 of the 10 items were recorded, the MADRS total score was the mean of the recorded items multiplied by 10 and then rounded of to the first decimal place.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The severity of illness for each participant was rated using the CGI-S. To perform this assessment, the investigator had to answer the following question: "Considering your total clinical experience with this particular population, how mentally ill is the participant at this time?" Response choices included: 0 = not assessed; 1 = normal, not at all ill; 2 = borderline mentally ill; 3 = mildly ill; 4 = moderately ill; 5 = markedly ill; 6 = severely ill; and 7 = among the most extremely ill participants.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The IDS-SR was a 30-item self-report measure, that was used to assess core diagnostic depressive symptoms as well as atypical and melancholic symptom features of major depressive disorder (MDD). For individual items, the scores range from 0 to 3. The IDS-SR are scored by summing responses to 28 of the 30 items to obtain a total score ranging from 0 to 84, higher values indicate greater disruption in the depressive symptoms.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The HAM-D17 was utilized as a secondary assessment of a participants level of depression. The HAM-D (17-Item) consisted of 17 items. Eight items were rated on a 0 to 2 scale (items 4, 5, 6, 12, 13, 14, 16 and 17), while nine items (items 1, 2, 3, 7, 8, 9, 10, 11, and 15) were rated on a 0 to 4 scale (twice the weight of the other items). For all of these items, 0 was the "best" rating and the highest score (2 or 4) was the "worst" rating. The possible total scores were from 0 to 52.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

The efficacy of study medication was rated for each participant using the CGI-I. The study physician would rate the participants total improvement whether or not it is due entirely to drug treatment. Response choices included: 0 = not assessed, 1 = very much improved, 2 = much improved, 3 = minimally improved, 4 = no change, 5 = minimally worse, 6 = much worse, and 7 = very much worse.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

A MADRS response was defined as >=50 percent reduction in MADRS Total Score from end of Phase A (Week 8 visit). The MADRS consisted of 10 items, all rated on a 0 to 6 scale with 0 being the "best" rating and 6 being the "worst" rating. The MADRS Total Score is the sum of ratings for all 10 items; therefore, possible total scores range from 0 to 60. The MADRS total score were to be unevaluable if less than 8 of the 10 items were recorded. If 8 or 9 of the 10 items were recorded, the MADRS total score was the mean of the recorded items multiplied by 10 and then rounded of to the first decimal place.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

A MADRS remission was defined as MADRS Total Score =< 10 and >= 50 percent reduction in MADRS Total Score from end of Phase A (Week 8 visit). The MADRS consisted of 10 items, all rated on a 0 to 6 scale with 0 being the "best" rating and 6 being the "worst" rating. The MADRS Total Score is the sum of ratings for all 10 items; therefore, possible total scores range from 0 to 60. The MADRS total score were to be unevaluable if less than 8 of the 10 items were recorded. If 8 or 9 of the 10 items were recorded, the MADRS total score was the mean of the recorded items multiplied by 10 and then rounded of to the first decimal place.

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.

CGI-I Response was defined as a CGI-I score of 1 (very much improved) or 2 (much improved).

The efficacy sample was the FAS comprised of participants who received 1 dose of double-blind study medication and had both end of Week 8 visit value and 1 post-randomization efficacy assessment for MADRS total score in double-blind Phase B. The LOCF method was used to impute missing data.