Maritza Ciliberto, 57, has taken an experimental Alzheimer’s drug by infusion every two weeks for more than a decade, and she is now older than her mother, two aunts and an uncle were when an inherited form of the disease began to take their minds.

Her story, and that of Jake Heinrichs, a 52-year-old in Brooklyn whose father and older brother died of the disease, appeared in Business Insider on Nov. 14, in a piece by Hilary Brueck about whether amyloid-clearing drugs can stop Alzheimer’s before it starts. Both are followed by the Dominantly Inherited Alzheimer Network Trials Unit, or DIAN-TU, at Washington University School of Medicine in St. Louis. Neither has symptoms. Neither knows whether the drug is the reason.

A decade of prevention trials in families like theirs has produced less evidence than hope. The first prevention trial in inherited Alzheimer’s, DIAN-TU-001, missed its main goal in 2020. A trial of crenezumab in a Colombian family of about 6,000 people missed its goals in 2022. The A4 study of solanezumab, in older adults with amyloid but no symptoms, failed in 2023. The one encouraging result comes from 22 people who took Roche’s gantenerumab for an average of 8.4 years, much of it in an unblinded extension. Measured against the trial’s main comparison group, even that result fell short of statistical certainty.

That signal, together with evidence from Eli Lilly that patients who start donanemab sooner do better, is part of the case behind two much larger trials now running in people with amyloid and no symptoms: Lilly’s TRAILBLAZER-ALZ 3, with donanemab, and AHEAD 3-45, with Eisai’s lecanemab. What the families have shown is narrower than a 2025 university announcement suggested. The drugs that failed in them did little to plaque, so the idea that clearing plaque early prevents dementia has not been disproved. It has been tested directly in only a few dozen people, without a placebo group.

A stylized brain with orange tumor clusters receives teal liquid through a tube connected to an IV bag.

Families that know their odds

Puerto Rico’s families are a less tidy case than they look. Most early-onset families there carry a founder mutation, PSEN1 G206A, first described in 2001 in families from Puerto Rico and the Dominican Republic and now tied to Alzheimer’s in more than 70 families of Caribbean-Hispanic ancestry, according to Alzforum’s mutation database. In one early group of carriers, symptoms started at an average age of 54.5, but some carriers have stayed free of symptoms into their 90s. The mutation does not always cause disease.

Business Insider did not say which mutation Ms. Ciliberto carries, and the public record doesn’t show it. If it is G206A, the fact that she has outlived her relatives’ age of onset says less than it seems to, because carriers of that variant differ by decades in when, and whether, they fall ill. The University of Puerto Rico in San Juan is a DIAN-TU trial site, and Puerto Rico was among the places where the gantenerumab extension recruited.

The Colombian kindred in Antioquia is the opposite case. Carriers of the PSEN1 E280A mutation almost always develop the disease, typically around age 44, and about 1,200 of the family’s roughly 6,000 members carry it, according to Banner Health. The University of Antioquia’s Neurosciences Group has studied the family for more than 30 years, work begun by the neurologist Francisco Lopera, who died in 2024. A trial there can count on carriers declining on a schedule. No trial in the general population can. CBS’s 60 Minutes sent Lesley Stahl to report on the family.

Preview: The Alzheimer's Laboratory
A 60 Minutes preview of Lesley Stahl’s report on the Colombian family whose inherited mutation causes early-onset Alzheimer’s. Video: 60 Minutes · YouTube

The first verdicts

DIAN-TU-001 opened in 2012. It randomly assigned 144 mutation carriers to gantenerumab, Lilly’s solanezumab or placebo, and on Feb. 10, 2020, Washington University said neither drug had slowed decline on the trial’s composite of memory and thinking tests, according to the Alzheimer’s Association. Gantenerumab did lower amyloid and markers of tau. Roche said in a statement that every gantenerumab participant started on a lower dose and began moving to a fivefold higher target only about halfway through the trial.

Bruce Booth, the biotech investor who writes as LifeSciVC, posted on X when the full data came out that spring. His point was the one that would shadow the field for five years: the drug had reversed the pathology and the patients had not benefited.

Alzheimer’s DIAN-TU trial: 2 beta-amyloid mAbs had no effect on clinical measures including cognition despite Gantenerumab reversing the pathology of amyloid and other biomarkers.

Surprising learning (!@#!): amyloid reductions don’t correlate w/ cognition. We must try harder.

Bruce Booth @LifeSciVC on X · April 2, 2020

In Colombia, the Alzheimer’s Prevention Initiative, led by Banner Alzheimer’s Institute in Phoenix with the University of Antioquia, Genentech and Roche, gave crenezumab or placebo to 252 cognitively healthy relatives for five to eight years. About two-thirds carried the mutation. Non-carriers got placebo, so no one had to learn their status. Topline results on June 15, 2022, showed the drug missed both primary measures. The full paper, published in [The Lancet Neurology](https://www.thelancet.com/journals/laneur/article/PIIS1474-4422(25)00426-0/fulltext) in February, put the yearly decline on the main cognitive composite at 1.10 points on crenezumab and 1.43 on placebo, a difference of 0.33 with a confidence interval running from minus 0.48 to 1.13 (p=0.43).

Crenezumab binds small clumps of amyloid and does not lower plaque. About half the carriers had no amyloid when the trial started, and most participants spent only about two years at the highest dose, Alzforum reported. Eric Siemers, one of the researchers Alzforum asked to comment, estimated about 125 people per arm, against at least 500 in typical phase 3 trials.

a negative result in an underpowered study with a very low dose unfortunately may not be very informative

Eric Siemers, commenting to Alzforum on the Colombian crenezumab trial, alzforum.org

The A4 study tried the idea in people without a known mutation. It enrolled 1,169 cognitively normal adults aged 65 to 85 with amyloid on PET scans. After 240 weeks, solanezumab had not slowed decline (p=0.26) and had not cleared plaque, according to the results in The New England Journal of Medicine in July 2023. The paper’s authors, led by Reisa Sperling of Brigham and Women’s Hospital, concluded that removing plaque may be necessary even at this early stage.

So the three negative trials say less than their headlines did. Two of the drugs barely touched plaque. The third was underdosed for most of its blinded run.

Eight years on gantenerumab

The extension that revived the idea began in June 2020. Seventy-three carriers received gantenerumab at doses up to 1,500 milligrams under the skin every two weeks, at 18 sites in Australia, Canada, Europe, Puerto Rico and the continental United States. Roche dropped the drug in November 2022 after it failed to slow decline in a trial of more than 1,900 people with the common form of the disease, and the extension ended in 2023. Only 13 participants finished three years. Forty-seven stopped because the sponsor ended the study, according to the published report.

A vintage-style hourglass holds orange sand, with a cork plugging its narrow center and a long shadow stretching to the right.

Dr. Bateman’s team published the results in The Lancet Neurology on March 19, 2025. Plaque fell. The clinical effect across the group was unclear: among the 53 participants who had no symptoms at the start and received any gantenerumab, the hazard ratio for progression was 0.83, with a confidence interval of 0.50 to 1.36. The finding that drew attention came from 22 people who were cognitively normal at the start and had taken the drug for an average of 8.4 years. Compared with the trial’s main control group of 74 people, their hazard ratio for worsening on the Clinical Dementia Rating Sum of Boxes was 0.57, with a 95 percent confidence interval of 0.31 to 1.07. Against a larger control group of 86, it was 0.53, with an interval of 0.29 to 0.97, just under the line.

Washington University’s announcement described that as the risk of symptoms falling “from essentially 100% to about 50%.” Even then the plaque had not gone. Mean amyloid fell from 62 to 33 centiloids, still above the positivity threshold of 24, Alzforum reported, and 70 percent of participants remained amyloid-abnormal at the end.

The cost showed up on MRI. Fifty-three percent of participants had amyloid-related imaging abnormalities, or ARIA: 30 percent had brain swelling and 47 percent had microbleeds. Five had symptoms from ARIA, two were hospitalized and recovered, and no one died.

Dr. Bateman, the senior author, said in the announcement that most participants had moved on to lecanemab in hopes that they would never develop symptoms. “I am highly optimistic now,” he said, adding that if late-onset prevention trials had similar results, preventions could soon be available for the general population.

Everyone in this study was destined to develop Alzheimer’s disease and some of them haven’t yet

Dr. Randall J. Bateman, director of DIAN-TU at Washington University School of Medicine, medicine.washu.edu

FOX 32 in Chicago was among the stations that reported the study when it came out.

Scientists announce possible breakthrough in delaying Alzheimer's
A FOX 32 Chicago news report on the Washington University finding that removing amyloid plaque early might delay inherited Alzheimer’s. Video: FOX 32 Chicago · YouTube

Weighing the signal

Robert Howard, a professor at University College London, said in comments collected by the Science Media Centre that the release’s claim of delayed or prevented dementia “is not supported by the data and could give false hope.” The extension had no randomly allocated placebo group, he said, and the authors analyzed many treatment groups, which raised the odds that a difference arose by chance.

no responsible clinical trialist should claim on the basis of these data to have shown a 50% lowering of the risk

Robert Howard, professor at University College London, sciencemediacentre.org

Others on the same page were split. Tara Spires-Jones of the University of Edinburgh called the results “scientifically promising.” Sebastian Walsh of Cambridge said they were “actually ‘null.’” A [letter to The Lancet Neurology](https://www.thelancet.com/journals/laneur/article/PIIS1474-4422(25)00234-0/fulltext?rss=yes) in August 2025 said the trial’s primary findings were null and that the authors had stressed an exploratory analysis of a small, unblinded subgroup. In a [commentary](https://www.thelancet.com/journals/laneur/article/PIIS1474-4422(25)00066-3/fulltext) titled “the wrong drug at the right time?”, Jonathan Schott of University College London’s Dementia Research Centre said the small sample, a stop-start and truncated study and comparisons with convenience control groups made the findings hard to analyze. The journal later corrected his piece to call the 22-person analysis prespecified.

On the narrow question, the record supports Dr. Howard. The blinded phase failed, the extension had no placebo arm, and against the main controls the confidence interval for the headline estimate crosses 1. A 50 percent cut in risk is the middle of a range that includes no effect at all. The correction and the letter don’t contradict each other. An analysis can be planned in advance and still be exploratory, and this one rests on 22 people who chose to keep taking the drug.

The better support for Dr. Bateman’s argument, “Go earlier. Go earlier!” as he put it to Business Insider, comes from outside the families. In Lilly’s TRAILBLAZER-ALZ 2 extension, patients with early symptoms who started donanemab at the outset had a 27 percent lower risk of moving to the next stage of disease than those who began after 76 weeks on placebo (hazard ratio 0.73), the company said in July 2025. That comparison comes from the trial’s randomized delayed-start design. Lilly’s other three-year figures compare patients with an untreated outside cohort, the same kind of comparison the gantenerumab critics objected to.

A hand places a red wedge beneath a leaning domino at the front of a long row of upright dark blue dominoes.

Lilly also said more than 75 percent of treated patients reached amyloid clearance within 76 weeks. In the gantenerumab extension, 70 percent never did. The families’ trials, in other words, have not yet tested a drug that clears plaque the way the marketed ones do.

Amyloid is not destiny

For people without a mutation, the doubt starts before the first dose. In a 2018 model in Alzheimer’s & Dementia, Ron Brookmeyer and Nada Abdalla of the University of California, Los Angeles, estimated that a cognitively normal 65-year-old woman with amyloid and no other markers has a 29.3 percent lifetime risk of Alzheimer’s dementia, and a man 21.9 percent. Their 10-year risks were 2.5 percent and 2.3 percent. A carrier of a fully penetrant mutation faces near certainty. Most amyloid-positive adults do not.

The marketed drugs’ value in people who already have symptoms is itself contested. On April 16, Cochrane published a review of 17 trials with 20,342 participants, led by Francesco Nonino of the IRCCS Institute of Neurological Sciences of Bologna, which found the effects on cognitive decline and dementia severity “absent or trivial” and a likely rise in brain swelling and bleeding. “Unfortunately, the evidence suggests that these drugs make no meaningful difference to patients,” Dr. Nonino said in Cochrane’s announcement. Donanemab and lecanemab each accounted for one of the 17 trials.

A rebuttal led by Megan M. Bernath in Alzheimer’s & Dementia in August said pooling older antibodies with plaque-clearing ones created a “dilution effect” that hid the benefit of the approved two. The UK Dementia Research Institute made the same point, noting that five of the seven drugs reviewed had failed their trials. That objection holds as far as it goes. It does not show that the approved drugs’ benefit is large enough for patients to notice, which is the question Cochrane raised.

Laurie Ryan, a neuropsychologist who heads the clinical interventions and diagnostics branch at the National Institute on Aging, told Business Insider that what counts is the outcome, not the scan.

You care if you get dementia, right? That’s really what we want to stop.

Laurie Ryan, National Institute on Aging, businessinsider.com

Richard Isaacson, a neurologist who has worked on Alzheimer’s prevention for decades and was not involved in the trials, said the drugs could be “a really important additive treatment” in the right person, at the right dose and for the right length of time. On using them preventively now, he said, “I’m not ready to do that.” Dr. Bateman said he knows wealthy clients are pressing doctors for early access and that he does not take part, partly because no test can say which amyloid-positive people will go on to dementia.

The trials now running

Lilly screened 63,124 people in the United States and Japan for TRAILBLAZER-ALZ 3, using a p-tau217 blood test and a telephone memory test, and 2,196 qualified, according to baseline data published in September 2025. Of those, 1,697 began treatment, Alzforum reported. Participants get nine monthly infusions of donanemab or placebo, and the trial runs until 350 people have shown clinical worsening. ClinicalTrials.gov lists estimated completion in November 2027.

On Lilly’s Oct. 29 earnings call, Carole Ho, president of Lilly Neuroscience, described the trial as looking at prevention in cognitively unimpaired people and spoke of “really expanding this market to earlier stages of disease” if it reads out positively. Kisunla, Lilly’s brand name for donanemab, had $193 million in third-quarter sales, up 175 percent.

AHEAD 3-45 closed enrollment in October 2024 after screening 20,721 people. It has two parts: 448 people with intermediate amyloid in a phase 2 trial judged by PET scans, and 1,173 with higher amyloid in a phase 3 trial whose main measure is a cognitive composite at week 216. Seventy-three percent carry at least one APOE4 copy, 80 percent have a family history of dementia and about 90 percent are white, Alzforum reported. Primary completion is estimated for December 2028.

A large umbrella shelters a couple, a potted plant, and a small house from rain in a textured coral and teal illustration.

Back in the families

The gantenerumab veterans now take lecanemab. The DIAN-TU Amyloid Removal Trial began June 10, 2024, planning to give about 65 former extension participants lecanemab every two weeks for at least five years, with time to progression on the Clinical Dementia Rating as the main measure. Its registry entry lists primary completion in June 2030, the year Dr. Bateman named when he said the results could become “incontrovertible proof that we can delay the onset of dementia.” A separate DIAN-TU primary prevention trial of Lilly’s remternetug enrolls carriers as young as 18, up to 25 years before expected onset.

In Colombia, a second trial randomized its first participant, Banner Health said in a release on Oct. 15. About 200 carriers will first get donanemab to clear as much plaque as possible, then be split among continued donanemab, Roche’s experimental oral drug nivagacetor, meant to curb amyloid production, both drugs, or placebo. Forty non-carriers get placebo. The National Institute on Aging is paying about $74.5 million. David Aguillón, director of the Antioquia group, said the trial would show “how effective the different approaches are at maintaining low amyloid levels following plaque removal.”

The kindred’s research record is under scrutiny on another front. On Aug. 13, Science published an investigation by Charles Piller and Jennie Erin Smith questioning later studies of two variants credited with shielding family members from early dementia: APOE3 Christchurch, first described in a 2019 Nature Medicine case report, and a Reelin variant reported in 2023. Forensic analysts the magazine consulted flagged image problems in several papers. Science did not dispute the original 2019 case.

Mr. Heinrichs’s brother David died of Alzheimer’s in 2019, at 49. Mr. Heinrichs is now three years past that age with no sign of decline, and his belief that the drugs are working persuaded him to have a child.

The difference in the outcome of those four years has been everything

Jake Heinrichs, DIAN-TU participant, businessinsider.com

Lilly’s prevention trial is scheduled to finish in November 2027 and AHEAD 3-45’s main readout in December 2028. The 65 former gantenerumab patients now on lecanemab in Washington University’s trial are the nearest thing to a direct test of what the families began, and that trial’s primary completion date is June 2030.