Zogenix: The proof is in the publications
“Men occasionally stumble over the truth, but most of them pick themselves up and hurry off as if nothing ever happened” — Winston Churchill
Zogenix (ZGNX) is a clinical stage biopharmaceutical company developing treatments for orphan neurological diseases. It’s lead (and only) asset is ZX-008, or fenfluramine, for the treatment of Dravet syndrome and Lennox-Gastaut syndrome. Readers may well remember fenfluramine as part of the infamous combination drug fen-phen, which was marketed for weight loss and eventually pulled from the market due to its association with potentially fatal pulmonary hypertension and heart valve defects. These adverse effects were believed to be due to the action of fenfluramine.
Fenfluramine activates serotonergic pathways in the brain by promoting the rapid release of serotonin, inhibiting its reuptake, and by its direct receptor-agonism at 5-HT2 receptors. Fenfluramine may cause pulmonary hypertension through the vasoconstrictor action of serotonin or by altering the depolarization of pulmonary vascular smooth-muscle membrane. The mechanism through which fenfluramine may cause damage to cardiac tissue is through activation of 5-HT2B receptors in the heart. These receptors are abundant in cardiac tissue and are responsible for regulation of cardiac cell growth and proliferation. Thus, aberrant activation of these receptors may lead to excessive valve cell growth. Interestingly, fenfluramine’s mechanism of action in Dravet syndrome may instead be due to its activity as a positive allosteric modulator at the sigma-1 receptor.
There are two key distinctions between the use of fenfluramine as an anti-obesity drug versus ZX-008. First of all, ZGNX is targeting lower systemic exposure of fenfluramine than was required for the anorectic effects of fen-phen. The dose levels being investigated by ZGNX are between 0.2 and 0.8 mg/kg/day (with a maximum daily dose of 30 mg), whereas the typical daily dose of fenfluramine in fen-phen was 60–120 mg. Secondly, Dravet syndrome is a rare and tragically life-altering disease, for which no currently available treatment meaningfully improve patients’ quality of life. As with any drug, the health risks must be weighed against its potential benefits for this patient population.
Dravet syndrome is a form of epilepsy that presents during infancy, and is caused most frequently by an autosomal dominant mutation in the gene coding for the voltage-gated sodium channel NaV1.1. The mutation renders the expressed protein non-functional, and thus affected individuals have an impaired ability to transmit sodium currents across certain neurons. There is evidence that this impairment particularly affects the firing of hippocampal GABAergic inhibitory neurons, which would cause hyperexcitability and contribute to the severe, intractable seizures observed in Dravet syndrome. Later in childhood, the condition leads to ataxia, behavioral disorders, cognitive impairment, and life-threatening status epilepticus. Dravet syndrome affects approximately 20,000 children in the US, most of whom will not survive past the age of 10.
Fenfluramine has been investigated as a treatment for Dravet syndrome over the last two decades through a compassionate use program in Belgium. A retrospective analysis of this long-term, open-label study included data from 12 patients who started treatment between the ages of 16 months and 16 years. Two of these patients discontinued fenfluramine before their last visit — one due to lack of seizure control and the other due to effective seizure control with only the anti-epileptic valproate. Of the remaining 10 patients, 8 had an at least 75% reduction in number of seizures, and 7 were seizure-free for at least one year by their last visit (with a range of 1–19 years seizure-free). In a five-year follow-up report on the 10 patients remaining on fenfluramine, three patients were seizure-free for the entire five year period, and an additional four patients experienced seizure-free intervals of two years during the study, including one of the patients who experienced no observed benefit during the previous study period. Additionally, 90% percent of patients had an average seizure frequency of <1 per month. Notably, no patient showed signs of cardiac vulvopathy or pulmonary hypertension at any point during the entire 27-year treatment period (although two patients had mild but stable valve thickening on the last EKG that was deemed clinically insignificant).
Although the above results bear the dual caveats of small samples size and no placebo comparator, they are clinically meaningful for two reasons. The first is that the subjects’ response to treatment is not what one would expect given the natural history of the disease and the poor efficacy of standard-of-care. The closest apples-to-apples comparison comes from a Japanese study that reported on the long-term outcomes of 31 patients, and found that only 16% were seizure-free for at least 1 year before the last follow-up, compared to 67% in the first Belgian study (on an intent-to-treat basis).
The second revelation stems from when seven of the patients in the first Belgian study were temporarily withdrawn from treatment due to shortage of drug supply. Four of these patients had been seizure-free; three of these four experienced recurrence of seizures upon discontinuing fenfluramine. All three of these patients became seizure-free again after restarting fenfluramine. The onset of efficacy is also noteworthy — in six patients who were seizure-free on fenfluramine, cessation of seizures was achieved within 3 days after start or restart of treatment. In many cases, patients were seizure-free even when exposed to usual triggers (e.g. fevers, video games). In such a severe condition as Dravet, these are not observations one would expect from a mere placebo.
As a follow-on to the Belgian study, Zogenix conducted a prospective, open-label trial in 9 Dravet syndrome patients aged 1–29 years. For the three-month baseline period during which no fenfluramine was administered, the median seizure frequency was 15 per month. After the baseline period, fenfluramine was added to patients’ antiepileptic drug regimens at doses of 0.25–1.0 mg/kg/day (max 20 mg/day). Every patient experienced a reduction in seizure frequency during the treatment period, with a median reduction of 75% (range of 28–100% reduction). The difference from baseline was statistically significant at all time points during treatment. Moreover, seven patients (78%) experienced a ≥50% reduction in seizure frequency. Again, there were no findings of pulmonary hypertension or change in cardiac valve structure or function in any patient over the course of the study.
Following these confirmatory results, a pivotal Phase 3 program was undertaken consisting of two randomized, placebo-controlled studies — 1501 and 1502 — which are identical in design but comprise patients in North America and Europe/Australia, respectively. A total of 119 subjects were randomized into dose groups of 0.2 mg/kg/day ZX-008, 0.8 mg/kg/day ZX-008, or placebo as an adjunct to current therapy. After a six-week baseline period, patients are titrated to their target dose over two weeks, and then held at the fixed dose for 12 weeks of maintenance treatment. The primary outcome of this study is change in seizure frequency between baseline and the combined titration and maintenance periods, for which the trial is 90% powered to see a 40% difference between the high ZX-008 dose and placebo arms. Secondary endpoints include the percentage of patients who achieve ≥50% reduction in convulsive seizures and determination of the longest seizure-free interval per patient, as well as the above analyses for the low dose ZX-008 vs. placebo.
Studies 1501 and 1502 exclude patients who are on stiripentol, which is the only approved drug for Dravet syndrome (although only in Europe) since, at the time the study was designed, the drug-drug interactions with ZX-008 were not known. However, a subsequent PK study showed that 0.5mg/kg/day of ZX-008 with stiripentol led to the same systemic exposure to fenfluramine as 0.8 mg/kg/day of ZX-008 alone. Consequently, ZGNX is conducting a Phase 3 study (1504) that is enrolling up to 70 patients who are on stable doses of stiripentol. The key difference between this study and 1501/1502 is that only a single dose of 0.5 mg/kg/day ZX-008 is being investigated; the efficacy and safety assessments for this study are the same as for 1501/1502.
All patients completing studies 1501, 1502, or 1504 are eligible to be enrolled in study 1503, an open-label extension study to track the long-term safety of fenfluramine in these patients. In the most recent meeting of the data safety monitoring committee in July, the committee recommended continuing the trials without modification, suggesting an absence of alarming safety signals.
There should be little doubt that ZX-008 is highly active in this patient population. As a case in point, let us direct our attention to the results of GW Pharma’s recent Phase 3 study of cannabidiol in Dravet syndrome, which was designed similarly to ZGNX’s Phase 3 studies. In that trial, the median seizure frequency in the placebo group (n=59) was 14.9/month at baseline and 14.1/month during the treatment period (a 13% reduction). This is similar to the seizure frequency observed during the baseline period of Zogenix’s prospective open-label study, and is what we should expect as the placebo response in its ongoing Phase 3. The cannabidiol arm showed a 39% reduction in median seizure frequency between baseline (12.4/month) and treatment period (5.9/month), with 43% of patients experiencing a ≥50% reduction in seizure frequency. The adjusted median difference in seizure frequency between the cannabidiol group and the placebo group was −22.8%, a statistically significant result.
If Zogenix’s Phase 3 demonstrates an effect of ZX-008 on seizure frequency anywhere near the 75% observed in previous studies, not only will the drug be approvable, but will compare favorably to a drug that is the primary driver of GW Pharma’s $2.6B market cap. Given its current $300M market cap, ZGNX remains undervalued given the potential of its lead drug and likelihood of success in Phase 3. The results of studies 1501 and 1502 are expected this quarter, and should be a significant catalyst for the stock. The results of study 1504 are expected in Q1 of 2018, providing another significant near-term catalyst.
The company has $65M in cash and cash equivalents, which should be sufficient to fund operations into the first half of 2018. With two defining data readouts forthcoming, there should be ample opportunity for its share price to appreciate ahead of a future financing. The chief risk for the company is that a negative safety signal emerges from the ongoing trials of ZX-008. However, given that hundreds of patients have been dosed to date, with no evidence of clinically meaningful adverse effects to cardiac or lung health, this risk appears minimal.
