Addyi.
That's the proposed trade name of Flibanserin, the new "female viagra" drug that's making headlines because the FDA has not approved it, but agreed to call a panel to review their decision.
No idea how to pronounce, "Addyi."
There are all kinds of controversies here. The group, Even the Score, is complaining that the FDA is guilty of gender bias. They're funded, in part, by Sprout Pharmaceuticals, the company that makes flibanserin. The FDA claims they just don't think the benefits outweigh the risks, and they seem to have agreed to the review panel because of political pressure. Other women's advocacy groups, like
Our Bodies Ourselves, are opposed to the drug, for the same reasons as the FDA. There are questions about the diagnosis of Hypoactive Sexual Desire Disorder, for which the drug is being developed.
All sorts of difficulties. And who knows the answers? I certainly don't. So let's
look at the drug (FDA report of June 4, 2015).
Flibanserin was originally developed by Boeringer Ingelheim, as an antidepressant. Then Boeringer Ingelheim switched to evaluating it as a treatment for low sexual desire in women. I don't know how that switch came about. In 2012, Sprout Pharmaceuticals acquired flibanserin.
The FDA is now on its third review cycle for flibanserin. In the first review cycle there was a unanimous, 11 to 0 vote not to approve the drug:
The FDA agreed that the NDA could not be approved and issued a Complete Response letter in
2010, citing the following deficiencies:
*Lack of substantial evidence of efficacy because the phase 3 trials did not show a
statistically significant change from baseline for one of the pre-specified co-primary
efficacy endpoints.
*Overly restrictive entry criteria for the phase 3 trials, precluding a full assessment of
efficacy and safety in the target population.
*The need for a DDI study to characterize the effects of a moderate CYP3A4 inducer and
a moderate CYP3A4 inhibitor on flibanserin pharmacokinetics. The letter also asked the
Applicant to submit results from a meta-analysis of phase 1 pharmacokinetic and safety
data in women who concomitantly received flibanserin with an oral contraceptive (a
weak CYP3A4 inhibitor).
*The need to complete the ongoing 12-week trial assessing the concomitant use of
flibanserin with selective serotonin or norepinephrine reuptake inhibitors, with particular
attention to possible exacerbation of depression.
*The need for a study assessing the effects of co-administered flibanserin and alcohol on
tolerability, blood pressure and orthostatic vital signs.
*The need for a study assessing the effects of supra-therapeutic doses of flibanserin on
orthostatic vital signs and risk of syncope in healthy premenopausal women.
*The need for an assessment of the risk of accidental injury with root cause analyses.
*The need for an assessment of the potential for human abuse because of central nervous
system effects.
The
first study in premenopausal women ended in 2008. The primary outcome measures were changes from baseline in frequency of satisfying sexual events (SSE) and changes in desire, both measured by electronic diary entries kept by study subjects. The secondary outcome measure was change in distress associated with reduced sexual desire, assessed by diary entries and by using response to Question 13 in the
Female Sexual Distress Scale-Revised (FSDS-R), which is, "How often do you feel bothered by low sexual desire," rated on a scale of 0-4.
There was no statistically significant difference in the measure of desire. The company blamed this result on "diary fatigue", and proposed using two questions from the desire domain of the
Female Sexual Function Index (FSFI-desire, which uses a 28-day recall period) to
assess changes in desire. I'm pretty sure the two questions are, "How often did you feel sexual desire," and, "How would you rate your level of sexual desire," in the past 4 weeks, on a 1-5 scale.
In other words, they wanted to change the primary outcome measure after the study was complete.
I quote the FDA report (p.11):
The Division advised that failure to meet one of the two co-primary efficacy
endpoints did not constitute an acceptable reason to alter a pre-specified and
agreed-upon endpoint. Furthermore, the Division and the Study Endpoints and
Label Development (SEALD) team had already shared with the Applicant
concerns regarding limitations of FSFI-desire instrument, regarding 1) recall bias
due to the use of a 28-day recall period, and 2) content validity of the two
questions.
In the second review cycle, in 2013, the FDA cited:
*Numerically small treatment differences compared to placebo, which do not clearly
outweigh the safety concerns.
*Concerns with content validity of the FSFI sexual desire domain, used as the co-primary
efficacy endpoint in the new phase 3 trial, and used as a secondary endpoint in the two
prior phase 3 trials.
*Concerns about how healthcare providers would identify appropriate candidates for
flibanserin, taking into account the change in DSM diagnostic criteria.
*A clinically significant interaction with alcohol causing syncope and hypotension.
*Increased exposures to flibanserin with moderate and strong CYP3A4 inhibitors, causing
clinically significant hypotension in some cases.
*Events of central nervous system depression (e.g., somnolence), some of which appear
temporally associated with accidental injury.
The FDA made several requests for additional data and analysis, if the drug was to be reconsidered.
In December, 2013, Sprout submitted a dispute resolution request, and asked that flibanserin be approved without additional data or analyses. This was denied in February, 2014.
And thus began Sprout's campaign to pressure the FDA into reassessing by invoking gender bias.
I apologize for my cynicism, but I suspect a political and social media campaign is cheaper and faster than additional research, especially when there is little expectation that additional research will improve the drug's prospects for approval.
What is flibanserin? It's a 5HT1A agonist and 5HT2A antagonist. It also binds moderately to 5HT2B & C, and dopamine D4. Its mechanism of action in HSDD is unknown.
There are significant interactions with fluconazole (7-fold increase) and ketoconazole (4.6-fold increase), and serious risks of hypotension and syncope, especially with concomitant alcohol use. Flibanserin is not a PRN drug, so interactions with alcohol are unavoidable, unless one completely avoids alcohol, which is unlikely in a general population.
Some alcohol-related data:
The exclusion criteria in the flibanserin studies are important to consider. One was a recent history (6 months) of major depression or history of a suicide attempt, or current suicidal ideation. Another was decreased sexual desire due to medication. So the studies could not assess decreased sexual desire due to depression, medication, or other psychiatric illnesses (another exclusion criterion). The generalizability of the results is therefore questionable.
But the bottom line is, does it work? If it does, then maybe women who are willing to risk the adverse events should be offered the drug. There are women who participated in the trials who feel they were helped by the drug, and are disappointed not to have access to it.
This is the FDA's efficacy conclusion (p.33):
The three pivotal trials conducted in North America of flibanserin 100 mg qhs showed a
statistically significant difference between flibanserin and placebo on the endpoints of SSEs,
FSFI-desire score (but not daily desire measured by an eDiary) and FSDS-R Q13 distress score.
These findings and the magnitude of the treatment effects are consistent across the three trials.
However, the treatment differences are numerically small. The FDA is seeking expert advice
from a multidisciplinary advisory committee panel as to whether these observed effects outweigh
...safety concerns...
To clarify, the first two trials used the eDiary as the primary outcome measure for desire, the condition flibanserin is intended to treat, with no statistically significant results. The third trial switched to FSFI-desire score as the primary outcome measure for desire.
And the overall results(p. 5):
*From a median baseline of about 2-3 SSEs per month, flibanserin resulted in a median
placebo-corrected increase of about 0.5-1.0 SSEs per month.
*From a mean baseline of about 1.8-1.9 on the FSFI desire score, flibanserin resulted in a
placebo-corrected mean increase of 0.3-0.4 (the FSFI desire score range is 1.2-6.0).
*From a mean baseline of 3.2-3.4 on the distress score, flibanserin resulted in a placebocorrected
mean improvement of 0.3-0.4 (on a scale of 0-4).
We'll just have to wait and see.