Look at the companies furthest along the pharmaceutical track in psychedelic drug development right now, and a pattern emerges that has less to do with clinical data than with strategic posture. Eli Lilly bought its way into the space at scale. So did AbbVie, for a different compound entirely. Compass Pathways built a defensible position from scratch rather than buying one. And Resilient Pharmaceuticals, the company with the longest clinical history of any psychedelic developer, reportedly went back to the FDA without new data to answer the exact deficiencies that sank its first attempt. None of this is a story about which molecule works best. It is a story about several different theories of how to win a maturing drug category, playing out simultaneously, inside the same few months.

Two acquisitions, not one, and that changes the read

Lilly’s agreement to acquire AtaiBeckley, announced July 16, values the company’s equity at approximately $2.8 billion upfront, with contingent value rights that could bring total consideration to roughly $3.8 billion. The deal gives Lilly BPL-003, an intranasal formulation of mebufotenin, in late-stage development for treatment-resistant depression, alongside VLS-01, a DMT buccal film, and EMP-01, a single-isomer R-MDMA candidate that reported preliminary Phase 2a social anxiety results in February. That third asset matters more than it first appears: Lilly did not simply acquire a psilocybin-adjacent bet. It acquired a company running programs across three distinct psychedelic chemical classes at once.

AbbVie’s acquisition of Gilgamesh’s bretisilocin program, announced roughly a year earlier for up to $1.2 billion, is the deal this desk had not previously examined closely enough, and it changes how the Lilly transaction should be read. Bretisilocin is a short-acting psychedelic in Phase 2 for major depressive disorder, and AbbVie structured the deal to acquire only that program, letting Gilgamesh spin off its remaining pipeline, an oral NMDA receptor antagonist, an ibogaine analog, and an M1/M4 agonist program, into a new entity that retains AbbVie’s existing option-to-license relationship. Two different large pharmaceutical companies, on two different timelines, both chose the same basic structure: acquire a specific late-stage or near-late-stage asset rather than build the underlying chemistry in-house. That is no longer one company’s unusual bet. It is a repeated pattern.

Compass: building the fortress instead of buying one

Compass has taken the opposite approach from the start. Rather than relying on acquired or licensed IP, it built a layered patent estate around COMP360’s specific crystalline form, Polymorph A, a strategy borrowed directly from conventional small-molecule pharma. That estate has faced sustained, direct challenge. Freedom to Operate, a nonprofit founded by psychedelic philanthropist Carey Turnbull specifically to contest what it considers overbroad patent claims in this industry, has pressed prior-art arguments against it for years, contending the underlying crystalline forms were already established decades earlier; by published accounts of the dispute, that pressure led Compass to withdraw an original 27-claim application and a resubmitted 10-claim version. A Post-Grant Review petition Freedom to Operate filed was denied by the Patent Trial and Appeal Board in 2022, a result Compass has treated as validating and the challenger has treated as narrow rather than final.

Compass is now in a rolling FDA submission for COMP360 in treatment-resistant depression, and separately announced a partnership with the Department of Veterans Affairs in August on PIVOT, a multi-site randomized trial evaluating COMP360 in veterans with treatment-resistant depression, with or without concurrent PTSD, for which Compass is donating the study drug directly. That VA relationship, alongside two completed Phase 3 studies and a Commissioner’s National Priority Voucher, gives Compass a credentialing profile few competitors in this sector can currently match, even as its core patent position remains contested.

What one attorney’s read adds to this picture

Sabune Winkler, a life sciences and regulatory attorney who is Principal and Founder of Saphan Pont Law Partners and Senior Counsel at Outside GC, with a background spanning clinical trials, regulatory affairs, and research ethics, offered this desk a specific framing worth engaging directly. Her view is that defensible intellectual property has become a more decisive factor in psychedelic acquisitions than clinical data alone, a shift she traces to secondary-patent tactics, filing follow-on patents on incremental tweaks specifically to extend exclusivity rather than because the tweak reflects a meaningful improvement, several of which courts have struck down as anticompetitive rather than innovation-protecting. Her reading of Compass is that its strong clinical program sits directly against active IP challenges, a tension a prospective acquirer would have to weigh explicitly.

On whether acquisition becomes the dominant path to market industry-wide, her answer is a qualified no: acquisition, in her framing, will remain the loudest headline rather than the whole story, since companies with de-risked data and defensible patents retain the option to license or stay independent, while earlier-stage assets needing capital will keep defaulting to acquisition as their most realistic exit. As evidence that a CNS drug does not need a large-pharma acquirer to reach patients and scale, she pointed to two patterns. Johnson & Johnson’s Spravato, now approved as a standalone monotherapy for treatment-resistant depression and, on analyst projections, tracking toward roughly $2.3 billion in annual sales, was built inside a large pharmaceutical company rather than acquired. And Axsome’s Auvelity shows a CNS drug reaching commercial scale at a still-independent company, without a large-pharma parent at all. She also cited Intra-Cellular’s Caplyta as a second independent example, though it no longer fits: Johnson & Johnson completed its acquisition of Intra-Cellular Therapies in April 2025. The broader counterpoint, that reaching patients and scale does not require being bought by a larger pharmaceutical company, still holds against a narrative built entirely around M&A.

Resilient: the same evidentiary problem, met with an argument instead of new data

Resilient, formerly Lykos and before that MAPS Public Benefit Corporation, holds the longest and largest clinical dataset of any psychedelic developer, and reportedly resubmitted its MDMA-assisted therapy NDA for PTSD on August 9, roughly two years after the FDA’s original Complete Response Letter. The 2024 CRL identified three specific deficiencies: the company’s failure to collect what regulators termed “positive” adverse events, insufficient durability data, and high rates of both prior MDMA use among trial participants and prescreening failures, concerns that compounded a functional unblinding problem this desk has tracked since. The resubmission proceeds without a new Phase 3 trial. Proponents, per reporting this desk independently confirmed, are arguing the original trial’s core efficacy and adverse event findings remained accurate despite the unblinding issue, rather than generating new data to answer the CRL’s specific concerns directly.

That is a materially different strategy than the one Panorama, Definium’s second GAD trial, is running for a different compound and indication: Definium built new trial architecture, a low-dose active comparator specifically designed to make functional unblinding harder to achieve, to generate fresh evidence the FDA’s own concerns don’t apply to its result. Resilient is resubmitting on existing data and arguing the same underlying category of concern, applied to a different compound, didn’t meaningfully bias its outcome in the first place. Both companies are answering the same open methodological question the FDA’s own July guidance now codifies. They are answering it in opposite ways, and which approach the agency finds more persuasive, for either program, is unresolved. MAPS, which funded and initiated the original clinical program but holds no current operational role in Resilient’s strategy, has publicly read the FDA’s new guidance favorably, describing it as consistent with methodology the organization discussed with the agency for over two decades, though MAPS has not independently confirmed the resubmission report itself.

What comes next across the sector, on a shared timeline

Compass and Usona are both expected to submit psilocybin NDAs later this year, and because both hold Commissioner’s National Priority Vouchers, their FDA reviews could run as short as one to two months once accepted, a compressed timeline relative to standard review. Definium is expected to submit its own NDA, for DT120 in generalized anxiety disorder, in 2027. Set against Resilient’s resubmission, this means the field may see multiple, methodologically distinct psychedelic drug submissions reach FDA decision points within a relatively short window of each other, giving the agency, and this desk, an unusually direct basis for comparing how different evidentiary strategies fare.

What the pattern says about market structure

Four companies, four strategic postures, and none of them primarily a bet on molecule quality alone. Lilly and AbbVie are both betting that acquiring a de-risked, late-stage or near-late-stage asset is worth the premium over building the underlying chemistry from scratch, and that scale can absorb legal and regulatory risk a smaller company couldn’t. Compass is betting that a self-built, self-defended patent estate is worth the years and the litigation it invites, because it removes the acquisition-target risk the other two accepted. Resilient’s approach, whether or not it succeeds, is a live test of whether an evidentiary gap can be closed through argument and reinterpretation rather than new data, a different and higher-risk bet than either of the others.

For investors and sponsors watching this sector, the practical lesson is that “psychedelic company” remains a poor single category for assessing risk. A company acquiring its way into the space inherits different risks than one building its own IP from scratch, and both differ sharply from a company betting that its existing dataset can survive a second regulatory review without material new evidence. Consolidation in this sector is happening along at least four axes at once, inside a small enough set of companies that they get discussed as if they were running the same race.

The caveats

The Lilly-AtaiBeckley deal has not closed and remains subject to shareholder and regulatory approval. Resilient’s resubmission has been reported by Psychedelic Alpha via sources who spoke on condition of anonymity and separately addressed by MAPS without independent confirmation; Resilient itself has not publicly discussed the filing, and this desk has not verified it through a primary company or FDA source. Whether the FDA even accepts the resubmission for review, whether a new advisory committee is convened, and specifically what the agency signals about functional unblinding as an analytically resolvable question rather than a disqualifying one, are all open, and the second and third questions bear directly on every other psychedelic sponsor’s own trial design choices going forward. Winkler disclosed to this desk that she previously advised MAPS and Lykos Therapeutics, Resilient’s direct predecessor, in support of their clinical development program and FDA submission for MDMA-assisted therapy, and that she holds no equity in any company discussed here; that self-disclosed history is relevant context for her comments.

The frame

The headline version of this story is that big pharma has arrived in psychedelics. The more useful version is that at least four companies are running four different experiments in how to own a position in a drug class this contested and this early, buying scale, building a moat, and in Resilient’s case, betting that an argument can do the work new data would otherwise need to do. Nobody has proven their approach correct yet, and the next several months, Compass and Usona’s expected NDA submissions, whatever the FDA decides about accepting Resilient’s resubmission, and how the agency’s own new guidance gets applied in practice, will be the first test of more than one of these strategies at once. Watching all four together, rather than covering each company’s news in isolation, remains the way to understand where this market is heading.