You've probably heard about Fanapt (iloperidone) and Saphris (asenapine), two antipsychotics that have recently won FDA approval for the acute treatment of schizophrenia, and, in the case of Saphris, for the treatment of manic and mixed episodes of bipolar disorder. Are these more “me too” atypicals or do they represent novel treatment options? Let’s take a look at the data.

Fanapt
Fanapt (iloperidone) is a mixed D2/5HT2A antagonist, like most of the second-generation antipsychotics. Fanapt has relatively high alpha-1 antagonism, which may explain its side effect of orthostatic hypotension, and has low anti-histaminergic (H1) effects, which may explain its relatively low weight gain and sedation. Its recommended therapeutic dose is 12-24 mg/day.

This new antipsychotic has a checkered history, and has been sold off to various companies, perhaps due to sub-par efficacy results. Fanapt was originally developed by Hoechst-Roussel Pharmaceuticals (now merged with Sanofi-Aventis) in 1990, and in 1997, Titan Pharmaceuticals purchased exclusive worldwide rights to the drug. A year later, in 1998, they sublicensed it to Novartis for further development. Novartis performed three initial clinical trials that suggested that Fanapt was less effective than Risperdal (risperidone) and Haldol (haloperidol) (Potkin SG et al., J Clin Psychopharmacology 2008;28(Suppl 1):S4-S11), and unloaded it to Vanda Pharmaceuticals (its current owner) in 2004.

Let's take a closer look at those initial disappointing Novartis trials, which enrolled, in total, 1,943 subjects with schizophrenia or schizoaffective disorder. All three studies had similar designs: they were six week double-blind trials comparing Fanapt with placebo and active drugs.

In Study one, patients were randomly assigned to one of five possible treatments: Fanapt 8 mg/day, Fanapt 10 mg/day, Fanapt 12 mg/day, Haldol 15 mg/day, or placebo. Neither Fanapt 8 mg/day nor 10 mg/day separated from placebo on the primary outcome variable, which was the change in total PANSS (Positive and Negative Syndrome Scale) score from baseline. While Fanapt 12 mg/day did outperform placebo, Haldol’s performance was numerically more robust (13.9 pt improvement with Haldol vs. 9.9 pt improvement with Fanapt).

In Studies two & three, higher doses of Fanapt (up to 20 mg/day) were compared to Risperdal (up to 8 mg/day) and placebo. In these trials, Risperdal outperformed placebo at both the 4 mg/day and 8 mg/day doses, whereas Fanapt’s superiority over placebo was less consistent (it beat placebo in three of four doses tested.)

Fanapt’s lackluster efficacy was not helped by its problematic side effect profile. Common side effects of Fanapt in these initial trials were dizziness, headache, dry mouth, nausea, and insomnia. Of more concern were Fanapt’s effects on the heart. A pooled analysis of the three trials showed that Fanapt had caused statistically significant dose-dependent increases in the QT interval at all doses (up to 9.1 msec at 20-24 mg/day), whereas Haldol’s effect was milder (5 msec at 15 mg/day), and Risperdal showed almost no effect (0.6 msec at 4-8mg/day). Fanapt caused a mild average weight gain of 1.5-2.1 kg, similar to that of Risperdal (1.5 kg), whereas patients on Haldol and placebo both lost weight (-0.1 kg and -0.3 kg, respectively). On the positive side, Fanapt caused fewer EPS side effects (particularly akathisia) than either Haldol or Risperdal (Weiden PJ, J Clin Psychopharmacology 2008;28:(Suppl 1):S12-S19).

On the whole, these initial Fanapt studies were not particularly impressive. When Vanda acquired the drug from Novartis, it attempted to breathe life into the compound by conducting a new randomized trial comparing Fanapt 12 mg BID with both Geodon 80 mg BID and placebo (Cutler AJ et al., J Clin Psychopharmacology 2008;28:(Suppl 1):S20-28). Both active drugs outperformed placebo by four weeks. Fanapt caused less akathisia than Geodon but more weight gain (Fanapt 2.8 kg, Geodon 1.1 kg, and placebo 0.5 kg) and a greater increase in blood glucose (Fanapt 7.9 mg/dL increase, Geodon 4.7, placebo 3.2). Regarding the crucial issue of cardiotoxic side effects, Fanapt and Geodon are comparable, with mean QTc interval elongation of 7.2 msec for Fanapt and 6.1 msec for Geodon.

If you choose to prescribe Fanapt, you must titrate the dose up gradually to prevent orthostatic hypotension and dizziness. Start at 1 mg twice daily, then increase by no more than 2 mg per day over the next week to reach the target dose range of 12-24 mg/day. Regarding drug interactions, Fanapt is metabolized by CYP2D6 & CYP3A4, and inhibitors of 3A4 (e.g., ketoconazole) or of 2D6 (e.g., fluoxetine, paroxetine) can double Fanapt's serum levels. Because Fanapt is not metabolized by CYP1A2, smoking status should not affect its levels (as opposed to Clozaril, Haldol, and Zyprexa) (Citrome L, Int J Clinical Practice 2009;63:1237-1248)

Fanapt Pros: Lower risk of EPS than Haldol, less akathisia than Geodon. Fanapt Cons: twice a day dosing, need for seven-day titration, QT prolongation (at least as much as Geodon), dizziness, more weight gain and greater increases in blood sugar than Geodon.

Saphris
Saphris (asenapine) is manufactured by Schering-Plough and is unique in that it has to be dissolved under the tongue, with no eating or drinking for ten minutes afterwards, in order to be adequately absorbed into the bloodstream. It binds to more receptors than most other antipsychotics, strongly antagonizing D1, D2, D3, D4, 5-HT1A, 5-HT2A, 5-HT2C, alpha1, and H1, although whether all this binding poses an advantage or disadvantage is unknown. Unlike Fanapt, Saphris is approved both for the treatment of schizophrenia and manic episodes of bipolar disorder.

Reviewing the controlled data on Saphris is complicated by the fact that Schering-Plough has apparently come down with a nasty case of “publication bias”—the all-too-common industry practice of selectively publishing positive results (see, for example, Turner E et al., N Engl J Med 2008;358:252-60). Schering-Plough got into trouble recently for having delayed the release of negative data regarding Ezetimide, a medication for hyperlipidemia (see Greenland P et al., JAMA 2008;299(8):953-955.) Our early review indicates that the company now appears to be pulling similar shenanigans regarding Saphris as well.

Saphris for schizophrenia. A single study has been published thus far regarding Saphris for schizophrenia. In this six week study, 174 patients were randomized to one of three groups: sublingual Saphris 5 mg BID, oral risperidone 3 mg BID, or placebo. Both Saphris and risperidone separated equivalently from placebo (Potkin SG et al., J Clinical Psychiatry 2007;68:1492-1500). However, a methodologic aspect of this trial makes it unlikely to be relevant to standard clinical practice—namely, patients who had a history of failing other antipsychotics were excluded from this study. Saphris’ unusual sublingual route of administration would make it unlikely to be most clinician’s first choice for most patients. Common sense dictates that most of us would pull Saphris out of our tool bags after having rotated through several standard oral atypicals, such as Risperdal or Abilify. Will Saphris work for such treatment resistant patients? There is no data to suggest so.

Of more concern is the fact that data obtained from hundreds of other patients enrolled in Saphris trials are still missing from the literature. At medical meetings and in press releases, Schering- Plough has reported preliminary results for over 1500 patients in short term trials and 1200 patients in long term trials, but thus far results for only 174 patients have seen the light of day. In at least two of the unpublished trials, comprising 1600 patients, Saphris did more poorly than Zyprexa (see http://bit.ly/U5eZ3 and http://bit.ly/3GocoG), while in a small trial of 146 patients, Saphris apparently beat Zyprexa on negative symptoms only (http://bit.ly/4FGqZW). Bottom line: the truth is out there somewhere. Whether and when we will get a chance to see it is unclear.

Saphris for acute mania. Thus far, a single paper has been published, which reported on a three-week trial randomly assigning 488 patients with acute mania to Saphris, Zyprexa, and placebo. Both Saphris and Zyprexa outperformed placebo (McIntyre R et al., Bipolar Disorders 2009; 11:673-686). Treatment-related adverse events were reported as 60.8% for Saphris, 52.9% for Zyprexa, and 36.2% for placebo. At the APA meeting there were two successful trials presented, including close to 1000 patients (McIntyre R., APA 2008; Oral abstract #42 http://bit.ly/U5eZ3). When contacted, Schering-Plough told us that the publication uses data from only one of the two trials presented at the APA—it remains unclear why they chose not to publish both.

Regarding side effects, Saphris causes weight gain, though in one trial it caused less "clinically significant" weight gain than Zyprexa (15% on Saphris vs. 36% on olanzapine). Other common side effects of Saphris include sleepiness, akathisia, oral numbness, and dizziness.

Saphris Pros: Available as sublingual tabs—useful for those who cannot swallow pills. Saphris Cons: Available only as sublingual tabs—leading to mouth numbness and chalky sensation in some patients. Sedation, dizziness, weight gain, and akathisia common. The majority of efficacy data is still unavailable.

TCPR Verdict:

Overall, the data supporting the use of either Fanapt or Saphris seem underwhelming, despite their FDA approval. Fanapt appears to be less efficacious than other antipsychotics, and it causes QTc prolongation and metabolic side effects. Saphris is harder to evaluate, because so little data have been published. It appears to be less effective than Zyprexa, and it requires sublingual administration, an advantage for the tiny percentage of patients who are pill-phobic, but an annoyance for most others. At this point, we would not recommend either of these drugs as first-line agents.