A phase 1b trial of IBC-Ab002, an antibody that briefly blocks the PD-L1 immune checkpoint used in cancer treatment, met its primary goal of safety and tolerability in 40 people with early Alzheimer's disease, according to results published online July 16, 2026 in Nature Medicine. The randomized, double-blind, placebo-controlled study, run at 11 sites in the United Kingdom, the Netherlands and Israel, gave five ascending doses across separate cohorts, with each participant dosed four times at three-month intervals.
The approach differs from how checkpoint inhibitors are used against tumors. Cancer drugs that block PD-1 or PD-L1 are given repeatedly over long periods to keep the immune system switched on, which can carry a risk of autoimmune-type side effects. IBC-Ab002 was engineered for a short half-life and reduced immune-cell binding so that it acts only briefly after each dose, an approach the researchers describe as intermittent rather than sustained checkpoint blockade. The mechanism traces back to laboratory work from Michal Schwartz's group at the Weizmann Institute, which reported in a 2016 Nature Medicine paper that blocking the related PD-1 pathway reduced pathology and improved memory in mouse models of Alzheimer's disease.
Beyond safety, the trial's authors report exploratory findings: in participants who received the highest doses, cerebrospinal fluid levels of neurogranin, total tau and p-tau181, markers associated with neuronal and synaptic injury, trended downward. The results were presented as a late-breaking report at the Alzheimer's Association International Conference in London in 2026. The authors and outside reporting on the paper are explicit that the study was not designed or statistically powered to test whether the drug affects thinking or function, and the biomarker changes are described as trends in a small group, not a demonstrated clinical effect.
The trial's own limits are the fair counterpoint to any headline about a new Alzheimer's treatment. With 40 participants and CSF biomarker readings as the only biological signal reported, the data show that repeated dosing was tolerated and that something measurable moved in blood and spinal fluid at the highest dose; they do not show that the drug slows memory loss or disease progression. The authors and the sponsoring company say the findings support moving to larger, longer, randomized trials designed to test whether those biomarker changes correspond to any difference patients or families would notice.