Javelin Pharmaceuticals, Inc. (NYSE Amex: JAV - News) today announced that a reexamination conducted by a third party of pain score measurements from its Phase III study of Ereska™ (intranasal ketamine) showed that top line results for its primary endpoint were statistically significant. Previously, Javelin had reported that the top line results for its primary endpoint were not statistically significant. The previously disclosed top line results, based upon data captured by an external vendor, had inconsistencies whose presence was verified by a third party biostatistics company engaged by Javelin that thoroughly reexamined the trial’s conduct and the initial primary analysis. The third party’s correction of these inaccuracies yielded today’s results. As previously disclosed, global patient satisfaction scores for Ereska were also statistically superior to placebo. All other secondary endpoints, while numerically superior to placebo, did not achieve statistical significance. Ereska, a nonopioid analgesic, was generally well tolerated in the trial. The Company plans to meet with the FDA to review the trial’s results and discuss Ereska’s registration program. This randomized, multicenter, double-blind, 1:1 placebo-controlled study assessed the safety and analgesic efficacy of repeated doses of Ereska over 6 hours in 259 patients with acute moderate to severe pain following orthopedic surgery. The predefined primary outcome measure was the summary of pain intensity differences over a 6 hour period after initial drug dosing (SPID-6). The results for SPID-6 (means plus or minus standard errors), as determined by the third party biostatistics firm, were 78.5 ± 12.4 for the Ereska group and 47.3 ± 12.3 for the placebo group, yielding a P-value of 0.046. (The standard for statistical significance in pivotal clinical trials is a P-value of 0.05 or less.) The initially announced results released in August 2009, which the Company stated at that time would be thoroughly reexamined and could change, showed SPID-6 values of 78.2 ± 12 for Ereska and 47.9 ± 12.3 for placebo, yielding a P-value of 0.053.