You may not have noticed, but there was news last week Monday morning on a small biotech company that sheds some much needed light on the risks associated with effectively handing over control of the delivery of medical care in this country to the Federal government.
Early morning December 7th, Somaxon Pharmaceuticals (Nasdaq: SOMX) announced that, for the second time this year, the FDA has declined to approve the company’s treatment for insomnia, Silenor.
The FDA’s decision wasn’t based upon any risk of harm to the public. Silenor is an extremely low-dose form of doxepin. Doxepin belongs to a class of anti-histamines. If you’ve ever taken Benadryl, Zyrtec or Claritin then you’ve taken an anti-histamine. In terms of both the potential for overdosing and the susceptibility to addiction, anti-histamines as a class are among the safest drugs on the market.
Anti-histamines were discovered by the Swiss Nobel-prize winner Daniele Bovet in 1937, with the first treatment for allergies produced in 1942. Since then, there have been numerous anti-histamines brought to market, each improving and expanding the scope of the market. Anti-histamines inhibit the action of histamine in the body by blocking the H1 or H2 receptors in our body. Allergic reactions such as itching and hay fever may occur when histamines in our body stimulate these two receptors. By blocking these receptors you can prevent many allergic reactions; the market for such medication is in the billions.
Along the way, scientists discovered other novel usages of anti-histamines. In the late 1960s, scientists at what is now Boehringer-Mannheim synthesized doxepin, an anti-histamine that also blocked the uptake of serotonin, which meant it could act as an anti-depressant. After over 5 years of testing in Germany and Switzerland, doxepin was approved to treat depression and anxiety disorder. It has been widely prescribed and used in the United States for over 30 years.
With the experience of treating millions of patients over the years with anti-histamines, scientists began to notice certain side effects to the drugs. One of the most common side effects was drowsiness. Certain anti-histamines showed a stronger affinity for drowsiness than others, and eventually doxepin began to be prescribed to insomniacs and others who had difficulty sleeping.
While doctors are allowed to prescribe medication for indications other than those the drug was originally approved for, it’s not ideal. For one thing, since the original approval of the drug was based upon dosing levels ideal for its intended usage, the prescribing doctor has nothing but his or her own experience, and that of his peers, to rely on when it comes to deciding how much of a drug to prescribe to treat a different condition.
In the case of doxepin, it was found to be therapeutically effective for the treatment of depression at dosages of 25 to 100 milligrams, with some patients needing upwards of 300 mg. Although doctors knew that doxepin was both effective at treating insomnia as well as extremely safe at very high dosage levels, no one knew exactly how much doxepin one needed to consume in order to receive its sleep-improving properties. This was exactly the question Somaxon Pharmaceuticals spent the better half of this decade pursuing.
Somaxon completed 6 phase 2 & 3 trials totaling over 1200 patients. The results were startling. As opposed to treating depression, where dosage levels can reach 300 mg, Doxepin showed statistically significant improvement at just 3 mg. The differences in the trials between the 3mg and 6mg dosages were small, indicating that for most patients just 3 mg should do the trick. What was especially interesting was that, using the standard measurement of WASO, or wake after sleep onset, doxepin performed slightly better than Ambien, a currently marketed sleep drug commanding nearly 50% of the market. What this means is that, while Ambien works better at helping a patient fall asleep, doxepin does a better job of keeping a patient asleep for a longer time period. Ambien has a well-known addictive side-effect, among other negative traits. Anti-histamines in general and doxepin in particular have no such addictive properties, are extremely well-tolerated, and have decades of safety and other incidental usage data. At first blush, this seemed to be good news for those suffering from insomnia - a condition that afflicts millions of Americans. Approximately 14 million people are diagnosed and treated with sleep medication in general, and it’s estimated that twice that figure remain untreated. Somewhere around 2/3rds of patients who are treated say they are unhappy with their current medication.
All this data from the 6 trials completed by Somaxon should have been welcome news to those who suffer from insomnia. Unfortunately the scientists at FDA did not see things in the same light. After reviewing the data submitted by Somaxon for over a year, in February the FDA declined to approve doxepin. Some of the things FDA wanted, such as dropping a low 1mg dosage in favor of the 3 & 6 milligram dosages, seemed reasonable. But what really shook up Somaxon was what FDA had to say about the effectiveness of Somaxon as a sleep maintenance drug.
In its’ letter to Somaxon, the FDA recognized the statistically significant data, measured both objectively (using polysomnography) and subjectively (using patient’s own written recollection) in both adults and the elderly. However, the FDA questioned the “robustness” of the subjective data insofar as it is sustained by a patient taking the medication over time. In other words, they questioned whether, based upon patient recollection, the drug continued to work as well over time as it does after initial usage.
This question should have seemed rather easy to answer. The statistical relevance of data from a drug trial is discerned using well-established statistical tools. In order to rule out that a drug trial’s results are not random, a ‘p value’ is calculated that takes into account both the divergence of data between the drug’s effect and that of an identical trial using a placebo. The more patients enrolled the easier, statistically speaking, to prove statistical significance as the odds that hundreds of patient received the same benefit from a new drug by chance are extremely remote. Under industry-accepted guidelines, the results must show a greater than 95% chance the results were due to a drug’s efficacy as opposed to sheer luck, in order to claim a trial was successful. The ‘p value’ for many of the endpoints in Somaxon’s trials exceeded 99%. Numerous statisticians reviewed Somaxon’s data from its trials, and all deemed the efficacy data generated, both objectively and subjectively, statistically significant. So what did the FDA mean by “lack of robustness”?
In meeting with FDA officials, Somaxon learnt that they were concerned about the subjective data from one single night in the trial. The data on this particular night did not show efficacy, even though the same patients were judged objectively to have improved their sleep time. Such occurrences within a trial are not uncommon, especially when you are measuring subjective impressions of patients, which can by definition vary wildly at times from the reality of the patient’s condition. Since this night came towards the end of the trial period in adults, the FDA was concerned about the durability of the drug’s efficacy.
Somaxon went back and re-analyzed the data, again employing numerous Phd’s conversant in medical statistics. The results were clear : statistically speaking, the data from the trials was proven to be significant and the drug shown to be effective in maintaining sleep. Somaxon re-filed its application for doxepin with the FDA in June. On Monday, Somaxon announced that the FDA once again denied its application to approve doxepin.
Once again, the FDA informed Somaxon it could not approve doxepin because the data was not “robust” enough. It also added, for the first time in the many years of meeting with Somexon regarding doxepin, that it would require a Risk Evaluation and Mitigation Strategy (REMS), which would include a medication guide to be distributed along with the drug. Shareholders, who were convinced the odds of approval were high given where the stock closed out last week, reacted with predictable horror on Monday, sending the price of Somaxon shares down 60%.
Disclosure: Long TSPT, SOMX