CNS is probably the toughest segment in the drug industry with a frustratingly low success rate and a diminishing pipeline. This is best exemplified by the disproportionally low number of CNS programs that received Breakthrough Therapy Designation (BTD). Of almost 70 designations, only three (J&J’s esketamine, Neurocrine’s NBI-98854 and Acadia’s pimavanserin) were for CNS indications and all three are hardly as exciting as other BTD drugs (anti-PD-1, CARs, Harvoni etc.). Not surprisingly, there are very few “hot” CNS-related stocks.
Sage (SAGE) and Marinus (MRNS) are two exceptions, with two of the few exciting CNS projects out there. Sage’s SAGE-547 and Marinus’ ganaxolone, which have the same mode of action (MOA), are developed for the treatment of epileptic seizures. Unlike other CNS projects in late-stage development, SAGE-547 and ganaxolone are significantly de-risked based on a clear clinical proof of concept and understanding of MOA.
Both drugs activate GABA-A receptors, a family of receptors in the brain that reduce neuronal activity. GABA-A activation is a well validated mechanism used by many tranquilizers and anti-anxiety drugs such as Ambien and Valium. SAGE-547 and ganaxolone are distinguished from available GABA-A drugs based on to their ability to activate a wide spectrum of GABA receptors, including those found outside of the synapse (extra-synaptic). This may allow the two agents to overcome resistance to existing drugs and maintain a durable anti-seizure effect.
Despite the similarity between SAGE-547 and ganaxolone, each company is pursuing a different development path. SAGE-547 is an injectable formulation of a naturally occurring substance (allopregnanolone) for the treatment of super-refractory status epilepticus (SRSE), a severe form of uncontrolled seizures. Ganaxolone is an orally bioavailable derivative of allopregnanolone for the treatment of epilepsy.
The most notable difference between Sage and Marinus is valuation. Sage’s $1.33B market cap is 10-fold higher than that of Marinus ($129M). This gap can be ascribed to the different data packages for each drug, as SAGE-547 has a clear advantage in terms of unequivocal efficacy and time to market. Marinus, on the other hand, has positive (albeit not spectacular) results from a randomized phase II in epilepsy and an established safety profile.
Sage – impressive activity in SRSE
SAGE-547’s activity in SRSE is nothing short of remarkable. To date, the company reported clinical experience in 30 patients whose seizures could not be managed with available treatments. These patients are typically put into a medically induced coma with general anesthesia followed by weaning off anesthesia in an attempt to stop seizures.
Of 26 evaluable patients who received SAGE-547 on top of standard of care, 19 (73%) were successfully weaned off their anesthetic agents with successful resolution of seizures. Although the trial had no control arm, this figure compares favorably to historic success rates of 25-30%. Importantly, the effect persisted after discontinuation of SAGE-547. Based on these results, Sage expects to initiate a randomized phase III trial evaluating SAGE-547 as adjunctive therapy to standard of care in the coming months.
Marinus – Awaiting P3 validation in epilepsy
Marinus’ ganaxolone was originally designed as a chronic anti-seizure treatment. The drug was created as a close derivative of allopregnanolone using a minor chemical modification (see figure below). This modification was introduced in order to solve two issues that prevent allopregnanolone from becoming a chronic drug: It is not orally bioavailable and it can be converted to a steroid in the body. Importantly, the minimal modification enabled ganaxolone to retain significant GABA-A activity.
Source : Marinus Pharmaceuticals
P2 data in epilepsy point to a differentiated profile – Ganaxolone’s phase II was conducted in epilepsy patients whose disease was not adequately controlled with standard of care. The drug led to a statistically significant mean reduction in seizure frequency of 19.6%. The difference in median percent reduction in seizure frequency was 15.7%, which did not reach statistical significance. A responder analysis (proportion of patients with 50% decrease in seizures) also demonstrated numerical differences that did not reach statistical significance.
Source : Marinus Pharmaceuticals
So far, safety profile appears favorable with a clear differentiation over most approved agents (most common side effects were fatigue and sleepiness). A benign safety profile is crucial in epilepsy as patients take multiple drugs with complementary MOAs. The lack of any signs of reproductive toxicity (predominantly from animal experiments) is particularly relevant for women in childbearing ages.
Effect size appears clinically meaningful – Ganaxolone’s effect size across the different endpoints is at the low range of what has been seen historically with anti-epileptic drugs. Still, a median 15.7% reduction in seizure frequency (if corroborated in phase III) is clinically meaningful and should allow ganaxolone to capture some market share given its favorable safety profile. Vimpat (lacosamide), one of the recently approved anti-epilepsy drugs with annual sales of $500M, demonstrated a similar effect (15-17%) in its two pivotal trials. (see figure below).
Sources: Halász, Epilepsia 2009 and Chung, Epilepsia 2010
Data from ongoing P3 expected in Q1 2016 – Marinus is enrolling a phase III trial with a similar trial design with the goal of replicating the P2 data (Results expected in Q1 2016). If results are positive, Marinus plans to conduct another phase III (Regulators require 2 trials for approval) which is expected to read out in 2018.
Exploring additional indications with ganaxolone – P3 readout next year is by far the most important event for Marinus, but ganaxolone may have value in additional indications, including those pursued by Sage. Marinus is evaluating ganaxolone in two genetic orphan indications characterized by dysregulation of GABA-A signaling.
Fragile X Syndrome is the leading genetic cause of autism and intellectual disability. Decreased GABA-A signaling has been observed in mouse models of the disease and a recently published paper reported a treatment effect with ganaxolone in animal models. Data from a randomized double- blind phase II are expected to be released in the coming months.
PCDH19 female pediatric epilepsy is a recently identified subset of pediatric epilepsy characterized by cluster seizures. The disease, caused by a mutation in the PCDH19 gene, is characterized by low levels of allopregnanolone forming the rationale for evaluating ganaxolone. A 10-patient study was initiated last month with potential preliminary data in Q4 2015. As a single-arm study, the trial will evaluate changes in seizure frequency on ganaxolone vs. baseline seizure frequency.
IV ganaxolone is Marinus’ wild card
Marinus is working on an injectable (IV) formulation of ganaxolone, which is expected to enter the clinic in 2015. Although no concrete plans were disclosed, it is natural to expect Marinus to pursue SRSE or similar conditions characterized by severe uncontrolled seizures in hospitalized patients.
Based on the striking resemblance between ganaxolone and SAGE-547, and the latter’s strong proof of concept, likelihood of success is relatively high, assuming ganaxolone and SAGE-547 are truly interchangeable in their GABA-A modulation. Ganaxolone clearly has significant GABA-A activity but there is always the risk for a slight chemical modification to alter a drug’s biological activity.
Recent comparison data from Sage should be noted but are not a show stopper – Sage is understandably trying to differentiate between SAGE-547 and ganaxolone. Last year, the company presented a poster comparing the two drugs as well as a next-generation drug (SAGE-217). In some models, ganaxolone had comparable (in some cases even superior activity) to that of SAGE-547, but in some settings ganaxolone appeared inferior to SAGE-547 (and both drugs were inferior to SAGE-217). The primary concern stemming for this poster is that ganaxolone has a lower effect on extra-synaptic GABA-A, which is believed to be a crucial property for activity in SRSE. (see figure below)
Source : Robichaud, EILAT XII 2014
On the other hand, some of the differences may be technical artifacts (due to formulation, protocol etc.) and variability, which is always an issue in preclinical experiments, let alone in neurology. Therefore, Sage’s poster creates some uncertainty but given the striking similarity and the large body of scientific evidence for ganaxolone, the drug still has significant potential in SRSE.
IV ganaxolone could narrow the SAGE/MRNS valuation gap – SRSE alone represents a $~2B global opportunity (assuming 70,000 cases per year and a $30,000 treatment cost), so even a 15-20% market share represents a significant commercial opportunity. Using a 50% likelihood of technical success (i.e. that ganaxolone and SAGE-547 are equally effective) can support a $170 valuation excluding all other chronic (epilepsy, fragile X, PCDH19) and acute care settings for ganaxolone. Therefore, SRSE and similar indications could become meaningful value drivers for Marinus even before actual clinical results (H2 2016).
I intend to hold both Sage and Marinus. Sage is significantly de-risked based on its early clinical data in SRSE and a high likelihood of success but this is already priced in. Potential upside may come from expansion to other indications as well as next-generation GABA-A modulators that are expected to start entering the clinic in Q4 2015 (targeting earlier stage status epilepticus). Marinus, on the other hand, is a high risk bet with a lower likelihood of success but the company has a lot of catalysts in the coming year, starting from early data in orphan indications by year-end 2015 followed by epilepsy P3 readout in Q1 2016. IV ganaxolone could add significant optionality value when clinical testing begins and concrete development plans are disclosed.
Portfolio holdings – April 12, 2015