The placebo arm of APPROACH improved by almost 12 points on the MADRS depression scale over six weeks. HLP003 still opened another 7.6-point gap on top of that.

That is the core of the first pivotal result for Helus Pharma’s deuterated form of psilocin, reported on November 16. APPROACH enrolled 222 adults with major depressive disorder who had an inadequate response while taking a stable antidepressant. They stayed on that antidepressant and received either two 16 mg doses of HLP003 or two doses of inactive placebo, about three weeks apart. HLP003 met the primary endpoint at Week 6 and the key secondary endpoint at Week 12.

The program now has a positive placebo-controlled result in a trial more than six times the size of its Phase 2. Three questions come next: whether a second pivotal trial reproduces it, how often patients would need to return for treatment, and what the safety record contains.

What APPROACH showed

HLP003 separated from placebo from Day 2 through Week 12
MADRS improvement from baseline, HLP003 versus placebo, at Day 2, Week 6 and Week 12Bar chart of least-squares mean MADRS reduction from baseline. Day 2: HLP003 16.1 points, placebo 11.0. Week 6: HLP003 19.5, placebo 11.9. Week 12: HLP003 16.9, placebo 9.9.Day 2, HLP00316.1 pointsDay 2, placebo11 pointsWeek 6, HLP00319.5 pointsWeek 6, placebo11.9 pointsWeek 12, HLP00316.9 pointsWeek 12, placebo9.9 points019.5 pointsMADRS improvement from baseline, HLP003 versus placebo, at Day 2, Week 6 and Week 12Bar chart of least-squares mean MADRS reduction from baseline. Day 2: HLP003 16.1 points, placebo 11.0. Week 6: HLP003 19.5, placebo 11.9. Week 12: HLP003 16.9, placebo 9.9.Day 2, HLP00316.1 pointsDay 2, placebo11 pointsWeek 6, HLP00319.5 pointsWeek 6, placebo11.9 pointsWeek 12, HLP00316.9 pointsWeek 12, placebo9.9 points019.5 points

Least-squares mean reduction from baseline in MADRS total score (a longer bar is more improvement). HLP003 16 mg was administered on Days 1 and 22; both groups remained on a stable background antidepressant.

Source: Helus Pharma, APPROACH Phase 3 topline presentation, Nov. 16, 2026.

On the 0-to-60 Montgomery-Åsberg Depression Rating Scale (MADRS), from baseline scores of 32.7 (HLP003) and 33.8 (placebo), the placebo-adjusted difference was 5.0 points at Day 2, the day after the first dose (p=0.0013); 7.6 points at Week 6, the primary endpoint (p<0.0001; effect size 0.653); and 7.0 points at Week 12, the key secondary endpoint (p<0.0001; effect size 0.581).

Response means a MADRS reduction of at least 50 percent, and remission a score of 10 or less. Every comparison below had a p-value under 0.0001:

RateWeek 6 (HLP003 vs placebo)Week 12 (HLP003 vs placebo)
Response56.3% vs 28.2%45.5% vs 19.1%
Remission50.9% vs 18.2%40.2% vs 15.5%

The early numbers are the most striking: a day after the first dose, 39.3 percent of HLP003 patients met the remission threshold, against 15.5 percent on placebo. Helus said on the call that it has not yet analyzed whether the same patients stayed in remission through Week 12.

Participants had moderate-to-severe depression (MADRS of at least 24) and had improved by less than 50 percent on their current antidepressant. The trial’s registry record excludes patients who had failed two or more classes of antidepressant. This is an add-on treatment for patients whose antidepressant has not worked well enough. It is not a trial in treatment-resistant depression, the population Compass Pathways studied for COMP360.

How much of Phase 2 carried over

Phase 2, controlled portionAPPROACH
Patients analyzed34 (24 HLP003, 10 placebo)222 (112 HLP003, 110 placebo)
Dose12 mg or 16 mg16 mg
Doses before the endpointOneTwo, about three weeks apart
Placebo-controlled timepointsWeek 3 onlyDay 2 through Week 12
Background antidepressantAllowed, not requiredRequired, at a stable dose
Placebo-adjusted difference−14.08 (12 mg); about −13 (16 mg); −13.75 pooled−7.6 at Week 6; −7.0 at Week 12

Phase 2 figures are from Cybin’s releases of October 31 and November 30, 2023; APPROACH figures are from Helus Pharma’s release of November 16, 2026. In Phase 2 the placebo group received HLP003 at its second session.

On the call, Chief Executive Officer Michael Halstead said Helus saw a very similar drug effect in Phase 3 to the one in Phase 2. The two numbers do not measure the same thing: Phase 2 compared drug with placebo once, three weeks after a single dose; APPROACH compares them after two doses, at Week 6, in a differently selected population, with the placebo arm kept in place. That makes Phase 2 useful context, not a baseline against which APPROACH’s 7.6 points can be read as a decline.

Phase 2’s Week 3 result was reported mainly as a difference between arms; a March 2024 Cybin release put the improvement from baseline at about 17 points without the placebo arm’s own change or the cohorts covered. Without Phase 2 arm-level changes for its controlled comparison, the disclosed data cannot show whether the two trials’ differences rest on similar drug and placebo responses. What APPROACH shows is that a substantial treatment difference held in a controlled trial more than six times larger.

The safety record

Most common adverse events and treatment-related serious adverse events in APPROACH

EventHLP003 16 mg (n=112)Placebo (n=110)
Treatment-related serious adverse events3: depression (2), suicide attempt (1)Not given
Perceptual disturbance42.9%11.8%
Nausea40.2%12.7%
Headache34.8%23.6%
Visual hallucination33.0%6.4%
Blood pressure increased11.6%3.6%

Selected from the adverse events Helus reported in at least 10 percent of participants. Source: Helus Pharma, APPROACH Phase 3 topline presentation, Nov. 16, 2026.

Helus’s presentation says most adverse events were mild to moderate and most occurred and resolved on the dosing day. The topline release had given only a count of three treatment-related serious adverse events with HLP003; the presentation identifies them as two reported as depression and one suicide attempt.

On the webcast, management said the suicide attempt occurred weeks after the second dose, in a participant who had not responded to treatment, and that its broader review of the study data, including Columbia-Suicide Severity Rating Scale results, did not indicate a suicidality signal. That is Helus’s assessment; the presentation nonetheless lists the event among the three treatment-related serious adverse events. The presentation gives no further detail on the two depression events.

A placebo response Helus says may reflect its blinding

HLP003 is a psychedelic and the comparator was inactive. That combination raises a known interpretive problem: participants may recognize their assignment. Helus’s mitigations were:

  • remote raters with no information about the dose or the session;
  • a firewall keeping session events from the study team;
  • recruitment of participants largely new to psychedelics.

Asked about the placebo response on the call, Chief Medical Officer Amir Inamdar offered an unusual reading: its size may itself reflect those measures, together with how expectations were managed at informed consent. In other words, placebo patients may not have known they were on placebo. Helus has published no treatment-guess, expectancy or intensity data that would test that, and the APPROACH registry record lists no blinding assessment. The measures limit what raters knew; they do not show what patients knew.

The second pivotal trial will not settle it on its own. EMBRACE is randomizing about 330 patients to 16 mg, 8 mg or placebo on the same two-dose schedule, with a Week 6 primary endpoint. When the company, then called Cybin, announced the design in 2024, it called the 8 mg dose “sub therapeutic.” A difference between two doses that both produce noticeable effects cannot separate pharmacology from expectation unless the trial measures expectation directly.

Two doses, then what

APPROACH establishes that a two-dose regimen kept a placebo-adjusted effect through Week 12, about nine weeks after the second dose. Both arms gave back some ground in that time, and the difference narrowed by 0.6 points (Behavioral Wire arithmetic from the reported 7.6 and 7.0).

It does not establish what the second dose adds. Inamdar said the second dose was intended to secure durability, and that the trial did not measure its incremental benefit as a separate endpoint.

Nor does it say how often patients would need treatment after Week 12. That falls to EXTEND, an open-label study following patients for about a year. Patients there can receive up to three further treatments, triggered by protocol criteria Helus described as symptom severity plus a clinician’s assessment. EXTEND can describe retreatment intervals, relapse and long-term safety. It cannot measure efficacy against placebo.

Session length is also not yet known. Helus said it collected monitoring data that are not part of the topline, and that acute effects in Phase 2 lasted about four to six hours. For a clinic-delivered treatment, retreatment interval and visit length together define what clinicians and payers would be weighing.

The path to an application

Asked whether it had discussed filing on one trial, Halstead said the FDA can approve on a single trial and that Helus will present the data to the agency. He also said Helus is planning for the standard program of two pivotal trials plus the extension, with EMBRACE results expected in the fourth quarter of 2027, and he repeated its guidance for a potential new drug application in 2028. Whether the Day 2 result could support a rapid-onset claim on a label, Inamdar said, is for the FDA to decide after reviewing the data. HLP003 holds Breakthrough Therapy designation. Helus’s filings say HLP003 would be treated as a Schedule I substance, so DEA rescheduling would have to follow FDA approval before it could be marketed.

Commercially, Helus is pitching HLP003 as an add-on rather than a replacement. Halstead argued that clinicians prefer building on a partially effective antidepressant to washing a patient off it, and set HLP003 against the atypical antipsychotics approved as add-ons in depression. Pricing, he said, is too early to discuss.

The readout also starts a financial clock. About 9.0 million warrants exercisable at $8.14, issued with Cybin’s October 2025 offering, expire 30 days after the topline press release. Exercised in full, they would raise roughly $73 million (a Behavioral Wire calculation from the company’s September 30 warrant count less an October exercise). Whether holders exercise depends on the share price.

APPROACH answered the question Phase 2 could not: whether HLP003 separates from placebo, with the comparison held for 12 weeks, in a pivotal-size population. EMBRACE now has to reproduce that. EXTEND has to answer what APPROACH was not built to: how often patients would need to come back.