Specific Biologics Raises $35M to Delete the ALS Repeat, Not Add a Gene

Brent Stead's argument is narrower than a sharper blade. He thinks gene editing has been aimed at the wrong target. On 2 November, his company, Specific Biologics, closed an oversubscribed US$35 million Series A to test that claim in people with C9ORF72-associated ALS.

Specific Biologics is a small Canadian company with a long preclinical history. It was founded in 2017 as a spinout from Western University in London, Ontario, and has lived mostly on grants and seed capital since. The Series A is the first money that buys a clinical readout.

Hatteras Venture Partners led the round. Lumira Ventures and adMare BioInnovations, both early backers, returned. Genesys Capital, BDC's Life Sciences Venture Fund, DeepWork Capital and Seido Capital joined the syndicate.

The $35M bet on deleting a repeat

ALS is a progressive, fatal disease that destroys the motor neurons controlling speech, swallowing and breathing. In C9ORF72-associated ALS, the most common known inherited form, a six-letter DNA motif repeats hundreds of times. Past a threshold, the repeat produces toxic RNA and proteins.

Specific Biologics' lead candidate is designed to shorten that stretch at the DNA level. The company says it can remove the pathogenic expansion while leaving the normal sequence intact.

That is a different bet from the one most of the field has made. Antisense oligonucleotides bind the RNA message and reduce its output. They can be tested quickly, and they can be stopped. They also have to be delivered into the spinal fluid, repeatedly, for as long as the patient is treated.

Dualase is meant to be a one-time edit. The payload fits inside a single adeno-associated virus, which the company says can be delivered directly to the central nervous system. A lipid-nanoparticle mRNA version is also described.

One dose against a chronic regimen is a clean commercial story. It is also a harder safety story. A repeated drug can be withdrawn. An edit that reaches the wrong cell cannot.

Why repeat expansions are a different target class

Conventional nucleases cut at one site. Dualase is built around two cuts that generate non-compatible ends, then lean on the cell's own repair machinery. The company argues that this makes removal of a long repeat more precise than a single cut, and less likely to leave a scar the cell misrepairs.

Abnormal repeat expansions are implicated in more than 40 genetic disorders. That number is the commercial spine of the pitch: one editing mechanism, many diseases.

The preclinical evidence is company-reported. In patient-derived ALS cell models, Dualase produced near-complete collapse of the C9ORF72 repeat, according to Specific Biologics. The same effect was reported in disease-relevant animal models, with correction of downstream disease hallmarks and no detectable off-target edits.

None of that has been published in a peer-reviewed paper that an outside lab has replicated. That is ordinary for a Series A-stage platform. It also means the trial is the first independent test of the central claim.

The mechanism is not unique in principle. A granted patent from Tsinghua University describes a dual-gRNA approach for the same C9ORF72 repeat. Academic groups have excised the expansion in patient cells and in animal models. The defensible ground is execution: a compact editor, a delivery route that reaches motor neurons, and a manufacturing process that can be repeated.

What the Phase I/II biomarker trial actually measures

The study is an adaptive Phase I/II. The first cohort is a biomarker proof-of-concept, which is a modest goal by design: the company wants evidence that the edit engages its target in humans, not a functional benefit.

The details that would let an outsider judge the design are not public. Specific Biologics has not disclosed the dose, the route of administration, the biomarkers it will measure, the enrollment target or a start date.

That gap matters for valuation. A biomarker can be a direct measure of the mechanism, such as repeat RNA or the toxic dipeptide proteins the repeat produces. It can also be a downstream marker that moves for reasons unrelated to editing. The difference is the whole investment case.

This financing represents a defining step toward our goal of bringing a potentially transformative medicine to people living with ALS.— Brent Stead, Chief Executive Officer, Specific Biologics

Hatteras partner Ben Scruggs framed the round around precision, specificity and delivery — the three problems every gene editor claims to solve. Lumira's Nikhil Thatte called early clinical proof-of-concept in ALS patients "an important inflection point."

Both statements are investor language for the same thing. The round is priced on a mechanism, not on data in humans.

The 40-disorder platform question

Specific Biologics has described Dualase programs beyond ALS, including earlier grant-funded work on cystic fibrosis nonsense mutations. In March 2026 it received a CAD$1.8 million Genome Canada award, with Western University, to build a machine-learning platform for editor design.

Platform breadth cuts both ways. It widens the eventual market. It also spreads a small team across targets before any one of them has cleared a clinical milestone.

From company and publisher announcements; Qalsody treats a different genetic subtype and is shown for regulatory context only.

The comparison is not a clinical one. Qalsody treats SOD1-ALS, a different genetic subtype, and is included only as a regulatory benchmark. The C9ORF72 programs are the relevant peers, and their record is mixed.

BIIB078, from Biogen and Ionis, was discontinued after Phase 1 for missing efficacy. Wave's WVE-004 is still in early-phase testing. As we wrote in November, uniQure's Huntington's gene therapy reached the FDA on an externally controlled dataset — a reminder that regulators will accept surrogate evidence when the disease is severe enough, and that the bar is not the same as a functional cure.

What has to be true by the readout

Three things have to hold, in order.

First, the edit has to be safe in a human central nervous system. AAV delivery to the brain and spinal cord carries its own risk profile, separate from the editor itself. The company's "no detectable off-target effects" claim comes from cells and animals.

Second, the repeat has to collapse in patients the way it collapsed in patient-derived cells. Target engagement is the bridge between a clean mechanism and a clinical signal.

Third, the biomarker has to predict something a patient can feel. That is the step where most neuro programs lose the thread, and where a first cohort can look clean while the trial later fails.

The cash runway is the quiet constraint. US$35 million funds a first cohort and the platform work around it. It does not fund a registrational program. If the biomarker signal is ambiguous, the next round gets harder.

There is a version of this story where the science works and the company still struggles. A single-dose CNS editor with a first-cohort signal is an attractive partnering asset, not a standalone commercial business. The syndicate is built for that outcome.

What would make this a platform, not a one-off

Will Dualase have human biomarker data before the end of 2028?