More Details on Transcept's Intermezzo Efforts and Pending FDA Action


Yesterday, I wrote a note on Transcept's (TSPT) Intermezzo drug, which gave my thoughts on how the stock might behave if the drug is approved on its Thursday, June 14, PDUFA date, The note did not contain much background information that led to my conclusions. This note goes into more meaningful analysis of why I believe the odds favor approval and what role, if approved, Intermezzo may play in the insomnia market. You may want to reread yesterday's note after this one.

The Insomnia Market and Intermezzo’s Potential Role

Insomnia encompasses several syndromes: (1) difficulty falling asleep, (2) difficulty staying asleep (middle of the night awakening or MOTN), (3) waking too early and (4) poor quality of sleep. The most widely used drugs currently are the GABA A receptor agonists Ambien, Ambien CR and Lunesta, which produce an eight to nine hour drug sedative effect. They are taken at bedtime to produce sleep onset and to then prevent waking too early. Because of their long half-lives, they are not appropriate to take for MOTN and they are not approved for this indication. The FDA is particularly uncomfortable with their next day residual effects.

Intermezzo is the first product to be specifically targeted at and seek an indication for MOTN. It has three highly desirable characteristics: (1) it needs to be used only if MOTN occurs, (2) it quickly returns people to sleep and (3) its low dosage of active drug avoids or minimizes next day residual affect. Intermezzo is a low dose, sub-lingual formulation of the generic drug zolpidem, the active ingredient in Ambien and Ambien CR, the most widely prescribed insomnia agents in the U.S.

Ambien, Ambien CR and Lunesta are not approved for MOTN because they have sedation effects that last for eight to 10 hours. If a patient wakes up during the night and takes one of these drugs, physical performance could be seriously impaired the next day. Another drawback of long acting drugs is that they must be taken prophylactically in anticipation of a middle of the night awakening. Intermezzo is only used when needed. This reduces the exposure to drugs and reduces pills taken, which potentially provides both safety and cost advantages.

Intermezzo’s novel sub-lingual formulation allows about 25% of the dosage to be quickly absorbed through the tissues of the mouth, producing a rapid onset of effect. The remainder of the dose is then absorbed through the gastrointestinal tract and released more gradually like the long acting insomnia agents. It is given as a 3.5 milligram dose for adults which compares to the standard Ambien adult dose of 10 milligrams. There is also a 1.75 milligram dose for the elderly. Although the amount of active drug is lower, its more rapid absorption allows Intermezzo to get patients back to sleep faster. Transcept states that in its sleep lab Phase III study, patients got back to sleep in 14 minutes. The time to sleep onset was slightly longer in outpatient studies.

Size of the U.S. MOTN Market

The current insomnia market in the U.S. is 78 million prescriptions and a sales value of $2.1 billion. The major drugs are Ambien, Ambien CR, Lunesta, Sonata and Rozerem. It is estimated that about 14% of these sales are for MOTN even though this is off label use.

It is hard to estimate the size of the MOTN market since there is no approved drug. Based on research involving 9,000 subjects done over four years by the Stanford Sleep Epidemiology Research Center it is estimated that 35% of Americans or 83 million people suffer from insomnia. Of these the number of patients who have no difficulty falling asleep but experience MOTN is estimated at 22 million people. Of these 22 million, 3 million seek medical help. This is the addressable market for Intermezzo. About 90% of people with MOTN experience three or more awakenings over a six month period. As many as 50% experience MOTN three times per week continually for up to five years.

To estimate the size of the market, let us conservatively assume that there are three million patients each year who have three MOTNs per week for six months. If they were to use Intermezzo, this would require 72 days of therapy or 72 pills. The estimated price of Intermezzo based on comparison to currently marketed drugs could be $5 per day. With these assumptions, the addressable market would be $1 billion. These assumptions are slanted to the conservative side.

Reasons for Complete Response Letter

Transcept completed its Phase III programs and then submitted the Intermezzo NDA. On October 26, 2009, the FDA issued a Complete Response Letter which caused the stock price to be halved. The FDA acknowledged that Intermezzo had shown substantial evidence of efficacy that was in accordance with the proposed label. It also indicated that Intermezzo had not been shown to cause significant next day residual effects.

Unfortunately for Transcept, one of the major issues for the FDA with all insomnia and other hypnotic drugs is next day residual effects. The most widely used class of insomnia products (Ambien, Ambien CR and Lunesta) are GABA A receptor agonists notorious for causing morning after amnesia, next day driving performance impairment and other troublesome side effects that result from their mechanism of action. The FDA is very edgy about this class of drugs to which Intermezzo belongs.

With Intermezzo, the agency was also confronted with a new indication (MOTN) that had not been evaluated for any other insomnia product. This presented the potential for dosing issues. One FDA concern was that a patient might wake up and take a tablet and then awake later, forget having taken the first dose and take a second dose. Another FDA dosing concern was that some patients might ignore or not understand the package insert that specifies that at least four hours should elapse from taking a dose before driving.

Transcept’s Response to the Complete Response Letter

To address the multiple doses in an evening issue, Transcept designed a unit dose pouch. Inside each pouch is a tablet and patients receive a box of 30 pouches with each prescription. They are instructed to keep the box in the medicine cabinet. At night before going to bed, they are told to take a pouch out and put it by their bedside. If they take the product in the middle of the night, then reawaken and find an empty pouch by their bedside, it is a strong reminder that they've already taken the allowed one dose for the evening. The FDA in a written letter informed Transcept that this new packaging appeared to address their concern about inadvertent double dosing in a single night.

The second dosing issue raised in the Complete Response Letter was patient compliance with the requirement to not take the dose if they are going to wake up and drive in less than four hours. Transcept has addressed this with patient educational material that has strong warnings about using Intermezzo in accordance with its label that indicates dosing at least four hours before driving.

To answer the FDA concern about risk of next day residual effects on driving if Intermezzo is used in accordance with its label (dose is taken four or more hours before driving) or off label (less than four hours), Transcept designed and executed a driving study. The intent was to demonstrate that when Intermezzo is taken as directed, it does not present an unacceptable risk for impairing next day driving ability. The study evaluated Intermezzo taken at both four hours and three hours prior to driving. The four hours or greater is of course is the intended usage for the product and the three-hour time point was designed to give the agency an insight into the safety margin if patients misdose the product.

Design and Results of the Driving Study

Transcept began a unique highway driving study in July of 2010 that enrolled 40 healthy subjects in a four way crossover study. Each of the subjects at different time points received: (1) placebo, (2) Intermezzo taken four hours before driving, (3) Intermezzo taken three hours before driving, and 7.5 mg of Lunesta given nine hours before driving. The key endpoint selected for the study was standard deviation of lateral position (SDLP); essentially a measure of how much weaving the subject did during the drive. The SDLP was measured as a subject drove an automobile about 60 miles at a speed of 60 miles per hour in normal traffic conditions on a public highway in Holland. It was performed by a Dutch academic center which is a leader in studying the effects of drugs and alcohol on driving performance.

SDLP as a measure of driving impairment that was devised over 30 years ago that measures variability in lane position while driving. It has been widely used as an experimental assessment of driving performance for subjects who have taken alcohol and other mind altering drugs. However, it has never been relied upon by the FDA to evaluate the capacity of a drug to impair driving performance. Because of this, Transcept had an extensive dialogue with the agency about the type of statistical analysis that should be used and how to interpret the results. Transcept used a new statistical analysis for evaluating SLDP called a symmetry analysis. The FDA informed Transcept that this statistical methodology was a reasonable approach to measure potential next morning driving impairment.

There is no commonly accepted change in SDLP that identifies a driver as being impaired. Based on the body of work done at the Dutch university, the study pre-specified an SDLP of 2.5 cm. as being the primary threshold for impairment. This change is approximately equal to the mean increase in SDLP which occurs with a 0.05% blood alcohol level, which is accepted in Europe and much of the world as the level at which drivers are legally drunk. In the U.S., the alcohol blood concentration for driving under the influence is a higher 0.08% and at this level, the SDLP change is 4.2 cm.

In this primary symmetry analysis, when Intermezzo was given four hours before driving, there was no statistically significant drug effect in comparison to placebo at the pre-specified thresholds. This was the primary endpoint of the trial. When driving began three hours after taking Intermezzo, there was a statistically significant effect as was the case with the zopiclone arm of the study.

A secondary analysis looked at the mean change in SDLP. The Intermezzo four hour arm produced a 0.8 cm change against placebo. The Intermezzo three hour arm produced a 1.5 cm change and the long acting zopiclone arm produced a 2.5 cm change. This compares to a 4.5 cm change caused by a blood alcohol level of 0.08%, the legal limit in the U.S.

The driving study clearly is the centerpiece for the NDA resubmission and the four-hour time point appears to have been met. The three hour time point helps the FDA evaluate the potential effect of misuse of the drug and doesn’t show unacceptable risk. I believe that this should answer the FDA’s concerns and lead to approval. One other factor in my thinking is that the agency is aware that the long acting agents are being used off-label for MOTN and that this poses a considerably greater safety risk than approval of Intermezzo would create.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.