[No wonder Goldberg emphasized safety on the conference call. The safety analysis was based on all of the patients that took their oral insulin drug (3 x 35 patients.) The efficacy claim was based on 4 of 10 patients in the highest dose arm having a drop of 1% in HbA1c levels, which was statistically significant relative to placebo. In the earlier conference call Goldberg said that HbA1c levels are the only thing that matter, but that they should have picked patients with higher levels in order to get more meaningful data in clinical trials.
Most of the patients in the high dose group had minor decreases in HbA1c.
Yet this PR says that "Patients with HbA1c levels of 7.0% or greater are classified as inadequately controlled diabetics." which is presumably why they were included in the study.
From today's CC, Goldberg saying that they've taken a lot of heat for how they "tortured the data" to reach statistical significance.]
Emisphere Reports Additional Clinical Data From Phase 2 Oral Insulin Trial Wednesday November 8, 8:00 am ET
Company to Present Results Today During Presentation at Rodman and Renshaw Healthcare Conference
* Patients Across Three Dose Groups of Oral Insulin with Stable Hemoglobin A1c (HbA1c) Levels at Baseline Experienced Mean Decrease in HbA1c * Highest Dose of Oral Insulin Shown To Be Statistically Significant Versus Placebo in Decrease of HbA1c in Patients with Baseline of 8.0% and Above * Webcast of Presentation Today at 11:40 a.m. EST on Company Website
TARRYTOWN, N.Y., Nov. 8 /PRNewswire-FirstCall/ -- Emisphere Technologies, Inc. ("Emisphere", Nasdaq: EMIS) announced today additional results from its 90-day, Phase 2 study of its oral insulin product utilizing the Company's eligen® oral delivery technology.
Among all the patients, those on the high dose of oral insulin (10mg QID) had the most profound effect on HbA1c reduction:
* In patients with HbA1c levels at baseline between 7% and 8.9%, a decrease of 0.1% was seen in patients on oral insulin, versus an increase of 0.05% on placebo. * In patients with HbA1c levels of 7.5% to 8.9%, a decrease of 0.22% was seen in patients on oral insulin, versus an increase of 0.075% on placebo. * In patients with HbA1c levels of 8.0% to 8.9%, a statistically significant decrease of 0.74% was observed in patients on oral insulin, while no change (0.00%) was observed in patients on placebo (n=17, p=0.03)
In order for an efficacy determination to be made one requires a meaningful baseline measure. For efficacy evaluation purposes all patients whose screening to baseline HbA1c changed no more than 0.3% absolute were included in the efficacy database. For the safety evaluation all patients were included.
Also seen in these patients was a reduction of other glycemic parameters, such as fasting blood glucose, fructosamine and proinsulin, supporting that it was Emisphere's high-dose oral insulin product that was responsible for the drop in patients' HbA1c levels.
When further evaluating patients in the high dose oral insulin group, 2 of the 10 patients experienced a significant drop in HbA1c to 7% or lower while none of the placebo patients achieved this level of control. These patients both had HbA1c levels at baseline of 8.1%, and at day 90 had HbA1c levels of 7.0% and 6.0%, respectively. Patients with HbA1c levels of 7.0% or greater are classified as inadequately controlled diabetics.
In the high dose group, 4 out of the 10 patients experienced a decrease greater than or equal to 1% in HbA1c, while no patients in the placebo group experienced this magnitude of a drop, and 6 out of the 10 patients in the high dose group achieved a decrease greater than or equal to 0.7% in HbA1c vs. none in the placebo group.
Please see data slides related to today's announcement on the investor relations portion of Emisphere's website at www.emisphere.com/ir.asp.
"An in-depth analysis of our Phase 2 clinical data shows the potential of Emisphere's oral insulin product to be beneficial and practical in a diverse population of Type II diabetics," said Michael M. Goldberg, M.D., Chairman and CEO of Emisphere Technologies. "The study met the Company's expectations that we could demonstrate the oral administration of insulin using our technology for 90 days, eliminate the key safety concerns associated with the administration of insulin and patient monitoring, and achieve statistical significance with a particular dose in a small Phase 2 trial. We believe these results will allow us to effectively pursue a partner for the drug."
Enrollment Criteria
Patients enrolled in the trial were first observed for three weeks, and the variability of their HbA1c levels was monitored over this time. They were all on a fixed dose of a single oral agent, metformin, to treat their diabetes. The inclusion criteria for patients was a baseline HbA1c level of 7.0% - 9.5%.
During the initial monitoring stage, the three-week period between screening and baseline where there was no change in their treatment regimen and which was pre- randomization (n=91), patients' HbA1c changed by as much as 1.2%. Evaluated in the efficacy endpoint in this clinical study were data from patients whose HbA1c levels did not change by more than 0.3% during the three week period.[I take it this means they picked the patients with the best control over their disease, or perhaps the ones who changed their weight, diet, and exercisr patterns the least.]
The four arms of the study consisted of three doses of oral insulin (10mg four times daily, 10mg twice a day and 5mg four times daily) and placebo. There were 35 patients enrolled in each oral insulin arm of the trial, and 36 in placebo.
Emisphere's oral insulin product demonstrated an encouraging safety profile. During the study, there were no statistically significant differences in hypoglycemic events, serious adverse events, or insulin antibody formation for our oral insulin product compared to placebo. The patients on high dose oral insulin saw no difference in weight and experienced no nighttime hypoglycemic events.
Webcast Information
Michael M. Goldberg, M.D. will present these results today during Emisphere's presentation at the Rodman and Renshaw Healthcare Conference at 11:40 a.m. EST, and will also discuss these results when the Company holds its third quarter earnings results conference call on Thursday, November 9, 2006 at 10:00 a.m. EST. Investors may access the presentations within the investor relations portion of the Company's website at http://www.emisphere.com/ir.asp.
Market Data 
Markets 
AI
