With regard to agitation in Alzheimer’s disease (AD), doctors “are in a very difficult position right now,” said Axsome Therapeutics Inc. CEO Herriot Tabuteau. That’s because the products currently used off-label are antipsychotics – all of which carry an FDA black box warning “specifically against their use in elderly patients with dementia, including AD,” since they double the risk of stroke and of mortality.
Enter New York-based Axsome, which rolled out gratifying phase II/III data from the ADVANCE-1 (Addressing Dementia via Agitation-Centered Evaluation) study testing oral, non-competitive N-methyl-D-aspartate (NMDA) receptor antagonist AXS-05 (dextromethorphan/bupropion modulated delivery tablet). The strong results raised eyebrows, goosed the stock and stimulated further Wall Street buzz about the candidate’s odds for approval.
AXS-05 met the primary endpoint, knocking down agitation quickly and substantially as compared to placebo. In ADVANCE-1, 366 AD patients were randomized to treatment with FDA fast-tracked AXS-05, dose-escalated to 45 mg/105 mg twice daily, bupropion (dose-escalated to 105 mg twice daily), or matching placebo, for five weeks. Agitation, which afflicts 70% of AD patients, is associated with accelerated cognitive decline, earlier nursing home placement and higher death risk, the company noted. Officials will meet with the FDA regarding a second phase III trial to support a regulatory filing.
“When you have a patient who is agitated and who cannot be controlled and who is at risk to themselves and to family members potentially, something must be done,” Tabuteau said during a conference call with investors. “The clinical need here is great, and in all the antipsychotics used, the randomized data show that they're not effective. It is possible, obviously, to control anybody's agitation by significantly sedating them, but in patients who are elderly there is a significant risk of inducing falls.” Sedation also impairs cognition, already a hallmark of the disease.
Specifically, AXS-05 turned up a statistically significant mean reduction in the Cohen Mansfield Agitation Inventory (CMAI) total score compared to placebo at the fifth week, with mean reductions from baseline of 15.4 points for AXS-05 and 11.5 points for placebo (p=0.010). The results represent a mean percentage reduction from baseline of 48% for AXS-05 vs. 38% for placebo. The CMAI is a 29-item caregiver-rated scale that assesses the frequency of agitation-related behaviors in patients with dementia, including pacing and restlessness, verbal aggression such as screaming and shouting, and physical aggression – grabbing, pushing and hitting.
AXS-05 was also superior to bupropion on the CMAI total score (p<0.001), and it rapidly improved agitation symptoms. Betterment on the CMAI total score with AXS-05 was numerically superior to placebo starting at the second week and rang the statistical-significance bell at the third week (p=0.007), only one week after full dosing. A statistically significant greater proportion of patients achieved a clinical response on the CMAI (defined as a 30% or greater improvement from baseline) with AXS-05 as compared to placebo (73% vs. 57%, p=0.005). The results were consistent with clinicians’ global assessments of change, measured using the modified Alzheimer’s Disease Cooperative Study-Clinical Global Impression of Change for Agitation. On that measure, too, AXS-05 demonstrated statistically significant greater improvement in agitation as compared to placebo (p=0.036).
On the call, Guggenheim analyst Yatin Suneja called the ADVANCE-1 results “a very nice surprise.” Wainwright’s Raghuram Selvaraju also liked the outcome, pointing out in a report that “the safety profile continues to look clean – a particularly key consideration in the AD agitation indication.” The news proved particularly welcome after AXS-05’s recent primary endpoint miss in the phase III study called Stride-1 in treatment-resistant depression (TRD). AXS-05 showed only a numerical separation vs. bupropion on the Montgomery-Asberg Depression Rating Scale (MADRS) change at the sixth week and was not statistically significant (2.2-point difference, p=0.117), “which was a disappointment and, in our view, was probably due to inappropriate powering,” SVB Leerink’s Marc Goodman said in a report. The drug hit key secondary endpoints to show rapid and statistically significant improvements in depressive symptoms on MADRS vs. the active comparator at the first week, the second week and overall (key secondary endpoints, p=0.02, 0.035 and 0.031). Axsome said the data support continued development in TRD, and the company plans to launch another phase III experiment in the third quarter of this year.
More upside to come: analyst
Meanwhile, the prospects in AD seem enticing. Suneja conceded that “some investors may be concerned about the low placebo score in this study,” but the findings are in line with fifth-week placebo responses from a phase III study with brexpiprazole (Rexulti), from Tokyo-based Otsuka Pharmaceutical Co. Ltd. The 12-week effort with Rexulti showed placebo responses of between 16.5 and 17.8 points, which for Suneja confirmed his theory that a shorter study of five weeks is better, “as placebo rates in neuropsychiatry trials tend to increase over time.” Rexulti was approved in the summer of 2015 for major depressive disorder and schizophrenia.
Cross-trial comparisons are tricky, but research with atypical antipsychotic brexpiprazole “provides a useful benchmark for success,” in Suneja’s view. On top of that, a phase II study using dextromethorphan plus quinidine (DM+Q) as Nuedexta, from Avanir Pharmaceuticals Inc., of Aliso Viejo, Calif., makes for a “compelling” proof of concept with regard to AXS-05's mechanism of action, he said. Nuedexta was cleared by the FDA in November 2010 for pseudobulbar affect.
Otsuka and H. Lundbeck A/S, of Valby, Denmark, ran two phase III Rexulti trials in AD agitation. Study 1 assessed a fixed-dose regimen of the drug and Study 2 deployed a flexible dosing schedule. The study duration for both trials was 12 weeks and the primary endpoint was change in CMAI total score at week 12. While in Study 1, 2 mg/day of brexpiprazole yielded a statistically significant improvement in CMAI from baseline to week 12, the 1-mg/day dose didn’t. In Study 2, the flexible dosing ranged from 0.5-2 mg/day and fell short of statistically significant superiority over placebo. But a post-hoc analysis found that patients who titrated to the highest dose (2 mg/day) did show a statistically significant improvement in total CMAI score as well as Clinical Global Impression scale – Severity of Illness. Otsuka launched a third effort in 2018, evaluating higher doses of the drug (2-3 mg), and could report data this year. With Avanir, Otsuka has AVP-786 (deuterated DM+Q), which turned up mixed results from two phase III studies (one in 410 patients, one in 522), showing a possible benefit; a third study is ongoing, based on a new analysis of the first two.
AXS-05 acts not only against NMDA but is also a sigma-1 receptor agonist, nicotinic receptor antagonist, selective serotonin reuptake inhibitor and norepinephrine reuptake inhibitor. Nudexta and AVP-786 act similarly. Whether or not the speculative picture put together by Suneja holds true, SVB Leerink’s Goodman said he thinks Axsome stands to gain more credit for its contender. The stock hike on the latest clinical news “is at the low end of the range, and probably reflects some ‘sell on the news’ by some investors, given this is the last of several near-term events.” He predicted that, “as more investors do more work in the area, they will get even more positive on Axsome.” Goodman raised his price target to $115 on Axsome, which bears a 52-week high of $109.94 and a low of $13.64. Shares (NASDAQ:AXSM) closed at $90.13, down $4.90.