A Different Target for Alzheimer's: Early Immune-Based Data Draw Cautious Interest
By Alex Philippidis 394 views
An antibody aimed at immune aging rather than amyloid or tau showed early biomarker signals in a small trial. Larger studies will decide what those signals mean.

Alzheimer's drug development has a grim track record. A widely cited Cleveland Clinic analysis put the failure rate for clinical candidates at 99.6%, and more recent tallies have barely moved that number. That context is why a small early-stage study from ImmunoBrain deserves attention, even though it's far too early to call it a breakthrough.
The company, founded about 11 years ago, is pursuing an approach that doesn't center on amyloid plaques or tau tangles, the two targets that have dominated the field for decades. Instead, it's aiming at the immune system outside the brain, and at how that system's decline may feed into disease progression. My read: this is a bet on a different question, not just a different molecule.
The idea: aging immunity as a driver
ImmunoBrain says its treatments are meant to restore the body's peripheral immune system so it can support brain protection and repair. The company frames the problem as a loop. Inflammation in the brain disrupts neurodegenerative processes, while the peripheral immune system gradually loses function, and each weakens the other.

That's a sharp departure from the long-held view that Alzheimer's is essentially a disease of toxic proteins, with immune activity in the brain treated as a symptom of damage rather than a contributor to it. Michal Schwartz, PhD, ImmunoBrain's scientific co-founder and chief scientific officer, who is also a professor of neuroimmunology at the Weizmann Institute of Science in Rehovot, Israel, argues the protein story is incomplete. She describes the disease as dormant or asymptomatic for 10 to 20 years before symptoms appear, and says that once symptoms do appear, local inflammation becomes a major driver alongside amyloid. Her phrasing is deliberate: inflammation is "not the primary cause, not the trigger, but definitely a catalyzer."
In this model, microglia, the brain's resident immune cells, normally help prevent damage to neurons. As the peripheral immune system ages, it stops supporting them well, and microglia can shift into a source of inflammation instead. Schwartz's lab has argued for years that the brain depends on the immune system for maintenance and repair, and she puts it bluntly: brain aging reflects immunological aging more than chronological aging.
The candidate and how it's meant to work
ImmunoBrain's lead program is IBC-Ab002, an anti-PD-L1 monoclonal antibody built for chronic neurodegenerative settings. The PD-1/PD-L1 pathway is a checkpoint that normally dampens immune responses, and the checkpoint class is better known from oncology, where it's used to release the brakes on immune attack against tumors. Here, the goal runs the other way. The antibody is designed for transient, intermittent blockade, meant to briefly boost immune activity so that repair-oriented monocytes and regulatory T cells move into the brain, clear senescent microglia and calm inflammation.
The engineering matters too. The company says IBC-Ab002 pairs high-affinity PD-L1 binding with modifications meant to switch off Fc effector function and reduce FcRn binding. Those changes are intended to prevent antibody-dependent cellular cytotoxicity and complement-dependent cytotoxicity, which would otherwise risk damaging the cells the drug is supposed to help. Whether those design choices hold up at scale is exactly what later-stage trials will test.
What the Phase Ib data show, and what they don't
The data come from a Phase Ib first-in-human study (NCT05551741) in people with early Alzheimer's. Forty patients were enrolled across five ascending dose cohorts, from 1 to 30 mg/kg. Each received four doses at three-month intervals and was followed for one year. The primary endpoints were safety and tolerability, which is the usual job of a first-in-human study. Results were presented at the 2026 Alzheimer's Association International Conference in London, held July 12–15, and published in Nature Medicine.
On safety, the company reports a half-life of roughly four days across all dose cohorts, and no brain hemorrhaging or ARIA (amyloid-related imaging abnormalities) was observed. That second point is worth noting, since ARIA has been the main safety concern for the anti-amyloid antibodies already on the market.
The efficacy-adjacent data are biomarkers, and they're exploratory. At 48 weeks, the 30 mg/kg dose was associated with reductions compared with pooled placebo in neurogranin (−23.1%), total tau (−12.4%) and phosphorylated tau181 (−11.0%). Participants treated with IBC-Ab002 showed consistent direction across the three markers. VAMP2, a pre-synaptic marker, followed a similar pattern to neurogranin, and the two correlated strongly at the level of individual patients.
Schwartz is careful about the limits. "The biomarkers show an encouraging result, mainly the neurogranin that is an Alzheimer's characteristic marker and a prediction of cognitive performance," she said. "Yet, we cannot speak about statistical significance because of the small number of patients and the short duration of the follow-up." Forty patients split across five dose groups is a small sample, and biomarker changes don't guarantee that memory or daily function will follow. Those are the questions this study was never designed to answer.
The authors also wrote that because the strategy doesn't target a specific brain protein, it could apply to other chronic neurodegenerative conditions where neuroinflammation plays a role. That's a hypothesis, not a finding, and the paper's data don't test it.
The supporting science, and the skeptics
Schwartz's group has been building the case for the immune mechanism in several directions. A separate paper in Nature Neuroscience, where she was corresponding author, reported in both a mouse model and in Alzheimer's patients that production of monocyte-derived macrophages in the bone marrow, and their recruitment to the brain, are compromised. The authors described "a dysfunction in bone marrow myelopoiesis in AD" and argued that disease progression is driven by maladaptive systemic immune processes.
The proposed culprit is type I interferon. This signaling pathway is a normal part of antiviral defense, but Schwartz argues that when it stays switched on chronically, it starts suppressing the bone marrow's output of repair cells. "If you have an inflammation in the body, the immune system produces type 1 interferon to dampen inflammation. But if this type 1 interferon is chronically produced, it becomes an enemy of the bone marrow," she said. Her group reports seeing the same pattern in mice and humans. That work echoes a 2014 study in Science from Schwartz and colleagues, which linked brain aging in both species to type I interferon. Schwartz also suggests that the damage from viral infections may come less from the virus itself than from an uncontrolled immune response to it.
None of this is settled. Mouse models and a limited set of patient samples are suggestive, not conclusive, and the field has heard compelling mechanistic stories before.
The skeptical case is strong. A 2022 study found a 98% failure rate among candidates in Phase II and III trials, which suggests the high attrition rate hasn't changed much. A paper published in August by Michael A.S. Guth, PhD, JD, examined 567 Phase II and Phase III Alzheimer's trials from 2010 to 2026 and attributed the pattern to "structural and institutional forces that sustained investment in repeatedly ineffective mechanisms of action." Guth argued that results were negative across amyloid, tau, neuroinflammation, metabolic and synaptic approaches alike, which points to problems with disease definition, intervention timing and endpoint validity rather than to any single target. If he's right, a new mechanism alone won't fix the field. Trial design will have to change too.
Where the field stands
The anti-amyloid class has delivered some real, if modest, progress. Biogen and Eisai's Aduhelm was approved in 2021, but weak sales and a failure to win coverage from the Centers for Medicare and Medicaid Services led Biogen to halt development and commercialization in 2024. Leqembi (lecanemab-irmb), approved as a disease-modifying therapy for people with mild cognitive impairment or mild dementia, has gained ground since the FDA approved a subcutaneous maintenance dose last year and a subcutaneous starter dose, marketed as Leqembi Iqlik, in July. In the third quarter, Eisai reported $202 million in global sales, up 67% from $121 million a year earlier. Eli Lilly's Kisunla (donanemab-azbt), a once-monthly infusion approved in 2024 for early symptomatic Alzheimer's with confirmed amyloid pathology, was the first amyloid-targeting therapy with evidence supporting stopping treatment once plaques are cleared.
Not every late-stage program is landing. AriBio presented Phase III POLARIS-AD data for AR1001, an oral PDE5 inhibitor, at the Clinical Trials on Alzheimer's Disease conference in Boston. In a prespecified subgroup, the drug showed a statistically significant difference of −0.44 points versus placebo on the Clinical Dementia Rating–Sum of Boxes at Week 52, which the company described as a 31% slowing of clinical decline. In the full population of 1,535 participants, however, the difference was −0.11 points and not statistically significant. AriBio says it will discuss next steps with the FDA. That's a textbook case of why subgroup results need cautious reading.
Several large anti-amyloid programs are still running. Roche and Genentech expect primary completion in 2028 for two Phase III trials of trontinemab, with a prevention study estimating data in 2032. Eli Lilly expects Phase III data in 2029 for remternetug, its next-generation antibody.
What Collins brings, and what he doesn't settle
ImmunoBrain added Francis Collins, MD, PhD, to its board in October 2025. Collins led the NIH from 2009 to 2021, the longest tenure of any presidentially appointed director, and he retired from the agency in February 2025 after 32 years there. His comments carry weight, but he now has a financial and advisory stake in the company, and readers should weigh his enthusiasm with that in mind.

His critique of the field is pointed. "Most of the efforts I saw happening as NIH director were in terms of developing and testing monoclonal antibodies against amyloid," he said. "While they provided some positive signals, they certainly didn't provide the breakthrough benefit that people had hoped for." He also points to genetics: large genome-wide association studies have identified more than 100 signals associated with Alzheimer's risk, and the dominant pathway involves microglia and innate immunity. "The field seems to be stuck on the amyloid hypothesis, but so far that's not curing people or significantly preventing progression like we hoped."
Collins calls the ImmunoBrain approach "completely compelling and novel" and says the company is building an argument that the immune system is a major player in progression. He's not claiming the case is proven. His own framing is that the next step is a clinical endpoint: "The question is: can we show that this intervention changes the course of the illness?"
What comes next
Schwartz says the team plans a Phase IIb trial with a two-year follow-up and a more rigorous, ambitious endpoint. The company has submitted paperwork to the FDA and is raising funds for the study. If that trial shows clinical benefit, the team hopes it could be considered pivotal, though that would be a major step that hasn't happened yet.
For readers, the practical takeaway is modest. This is early-stage science with a small sample, exploratory biomarkers and no clinical efficacy data yet. It doesn't change how Alzheimer's is evaluated or treated today. If you or a family member are interested in research options, a neurologist or memory specialist can explain which trials are open, what they require and what questions they're designed to answer. Anyone currently on a treatment should keep that plan in place and talk to their own clinician before making changes.
What would make this story more convincing is simple: the same mechanism showing clinical benefit in a larger, longer trial. Until then, the most useful thing to watch is whether the Phase IIb design can tell a real effect apart from the familiar pattern of promising early signals that didn't hold up.