Two pipelines are running against early Alzheimer’s disease at the same time, and they are not the same medicine. One is already in clinics: Eli Lilly’s donanemab-azbt, sold as Kisunla, an anti-amyloid antibody with a U.S. label, an EU authorization, and a boxed warning for brain swelling and bleeds. The other is still in animals and leftover oncology inventory: IDO1 inhibitors built for melanoma and breast cancer that restarted glucose use in astrocytes and rescued memory in mouse models. The AEGIS Alliance is separating those tracks so a 科学 paper does not get sold as a prescription.
The U.S. Food and Drug Administration gave Kisunla traditional approval on July 2, 2024, for adults with mild cognitive impairment or mild dementia and confirmed amyloid. TRAILBLAZER-ALZ 2 showed about 35 percent slowing on the primary clinical scale in the less-advanced group and about 22 percent overall, with a 37 percent lower risk of moving to the next clinical stage at 18 months. Infusions run every four weeks. Treatment can stop once plaques are cleared. The boxed warning is ARIA — amyloid-related imaging abnormalities, meaning swelling (ARIA-E) or small bleeds (ARIA-H). Trial rates ran about 24 percent ARIA-E and 31 percent ARIA-H, with three deaths tied to bleeds. APOE4 testing and scheduled MRI scans are part of the label, not optional paperwork.
On July 9, 2025, the FDA cleared a slower titration drawn from TRAILBLAZER-ALZ 6. Instead of three opening doses at 700 milligrams, the new schedule uses 350, then 700, then 1,050 milligrams before the 1,400-milligram maintenance dose. That change cut ARIA-E by about 41 percent at 24 weeks and 35 percent at 52 weeks with similar plaque removal. Japan approved the drug in September 2024. The United Kingdom approved it in October 2024 without NHS funding. China approved it in December 2024. The European Commission issued a marketing authorization on September 24, 2025, limited to people with zero or one ApoE4 copy and confirmed amyloid. In England, NICE paused its appraisal on July 30, 2026 so Lilly and the National Health Service could talk price. Lilly said Kisunla led new U.S. amyloid-targeting prescriptions in the second half of 2025.
The cancer crossover is a different molecule and a different claim. On August 22, 2024, Science published work from Stanford’s Katrin Andreasson, Penn State’s Melanie McReynolds, Paras Minhas, and collaborators showing that IDO1 — indoleamine-2,3-dioxygenase 1 — sits on the pathway that starves hippocampal astrocytes of usable glucose. Inhibitors already sitting in oncology pipelines restarted that metabolism and rescued synaptic plasticity and memory in amyloid and tau mouse models. “We’re showing that there is high potential for IDO1 inhibitors, which are already within the repertoire of drugs being developed for cancer treatments, to target and treat Alzheimer’s,” McReynolds said. That sentence is a trial rationale. It is not an FDA indication. Nobody should walk into an oncology clinic and ask for an Alzheimer’s refill.
The metabolic story is the new angle that the antibody fight does not cover. Alzheimer’s brains have run hot on amyloid talk for twenty years. The Stanford and Penn State team asked a quieter question: what if the cells that feed neurons cannot process glucose because a cancer enzyme is hijacking the kynurenine path? IDO1 inhibitors were designed to pull that brake off in tumors. In the mouse work they pulled it off in hippocampus. Human-cell models showed the same glucose recovery in astrocytes. That is why the paper used the phrase “high potential.” Potential is a grant word. It is not a pharmacy word.
The metabolic angle matters because amyloid antibodies only work early, and only in people who still have plaques to clear. Kisunla does not rebuild a destroyed cortex. It slows the slide in a defined window. Families who arrive after that window hear a no. Families inside the window hear a bargain: monthly infusions, MRI surveillance, a real risk of ARIA that climbs in ApoE4 homozygotes, and a modest delay measured in months rather than a cure. Lilly’s three-year extension numbers, compared against an external ADNI cohort rather than a continuing placebo arm, have been used to talk about time saved. They are not a second randomized trial.
GLP-1 receptor agonists looked, for a year, like a third track. Diabetes and obesity drugs such as semaglutide had observational hints and a biological story about inflammation and insulin in the brain. Novo Nordisk put that story into Phase 3. EVOKE and EVOKE+ enrolled about 3,800 adults aged 55 to 85 with mild cognitive impairment or mild dementia due to Alzheimer’s. In November 2025 the company said the trials missed their primary endpoints. Peer-reviewed and CTAD presentations in 2026 confirmed the miss: oral semaglutide did not significantly slow clinical progression even though some biomarkers moved. That result is useful because it kills a shortcut. A shot that shrinks a waistline is not, on current evidence, a disease-modifying Alzheimer’s drug.
Auburn University’s troriluzole work in mice, posted online August 30, 2024, is another preclinical file. The compound lowered glutamate and improved learning in one model. It belongs on the same shelf as IDO1: interesting, not approved, not a substitute for a workup that includes amyloid confirmation. Blood-based biomarkers are the piece that actually changed clinic flow. Plasma p-tau and related assays are moving diagnosis earlier, which is the only setting where donanemab and lecanemab (Leqembi) have a labeled job.
Cost and access are the unglamorous half of the story. Infusion chairs, MRI slots, genetic tests, and payer fights decide who gets the slowing and who gets a brochure. NICE’s pause in July 2026 is a price argument dressed as process. The EU limit to non-carriers and heterozygotes is a safety argument dressed as a label. U.S. clinics that followed the original fast titration learned why the FDA later blessed the slower ramp. ARIA is not a footnote. It is the reason some eligible patients will never start.
The AEGIS Alliance has kept other medical files in the same 卫生新闻. stack, including a fungal anti-cancer compound synthesized after fifty years, reporting on gum disease and heart risk, and the sickle-cell gene-therapy file. Those stories share a rule this page is using: a mouse result and a marketed vial do not get the same headline. Kisunla is a marketed vial with a boxed warning. IDO1 inhibitors are a published mechanism. Semaglutide had a large Alzheimer’s bet and lost it.
Readers hunting a single “breakthrough” sentence will not find one that survives contact with the labels. Early Alzheimer’s now has two approved anti-amyloid antibodies in the United States, a safer titration for one of them, a European license with a genotype fence, an English cost standoff, a cancer-drug idea that still needs human trials, and a GLP-1 idea that already failed the trial it needed. That is a field moving. It is not a cure.
This article is not medical advice. Talk to a clinician about diagnosis, genotype testing, imaging, and whether any approved antibody is appropriate.
