Background

Donanemab, aka N3pG, is a humanized IgG1 monoclonal antibody developed from mouse mE8-IgG2a. This biologic drug recognizes Aβ(p3-42), a pyroglutamate form of Aβ that is aggregated in amyloid plaques. Most Aβ antibodies in therapeutic development bind various soluble or insoluble species but have low affinity to deposited amyloid plaques. The rationale behind donanemab is that targeting deposited plaque itself is necessary to clear existing amyloid burden from the brain, rather than merely prevent deposition of new plaques or growth of existing plaques. Some previous plaque-binding antibodies have been abandoned because they caused microhemorrhages in the brain. The mE8 antibody was reported to clear plaques in mice without causing microhemorrhages (Demattos et al., 2012). 

At the 2014 AAIC conference, Lilly reported that a randomized preclinical study of combination therapy with N3pG and the BACE inhibitor LY2811376 cleared more than 80 percent of amyloid from the brains of PDAPP-transgenic mice, compared to about 50 percent clearance each for the respective monotherapies. Neuropathology studies showed that donanemab removed both cored and diffuse plaques (see Jul 2014 conference news).

In postmortem brain tissue from people with AD or Down's syndrome, donanemab bound to a subset of roughly one-third of amyloid plaques, and strongly reacted with the plaque core. It recognized vascular amyloid similarly to a pan-Aβ antibody. In brain tissue from AD mice, the antibody detected intraneuronal Aβ (Bouter et al., 2022).

Findings

From May 2013 to August 2016, Lilly ran a Phase 1 study in 100 people with mild cognitive impairment due to Alzheimer's disease, or mild Alzheimer's disease, in the United States and Japan. Participants had to have a positive amyloid PET scan. This seven-arm study evaluated five intravenous doses from 0.1 mg/kg to 10 mg/kg, infused monthly up to four times, and a single subcutaneous injection against placebo for safety, pharmacokinetics, and pharmacodynamics. At the 2016 AAIC conference in Toronto, Lilly presented results from 49 patients, average age 74, who had completed the trial. Thirty-seven volunteers received a single initial dose, 12 placebo. After adverse-event monitoring, they received up to four additional monthly infusions of the same dose; people in the 0.1mg/kg cohort subsequently received 0.3mg/kg. The highest dose was reported to reduce plaque load in the brain, though that group had but six participants. Overall, their mean florbetapir SUVR fell from 1.65 at baseline by 0.26 over seven months, corresponding to a 40 percent reduction. No cases of ARIA-E were seen in this small trial, but there were two asymptomatic cases of ARIA-H. The antibody was reported to be strongly immunogenic. In the multiple-dose phase, six of the 37 patients had an infusion reaction with chills, flushing, dizziness, rash, and fever, and anti-drug antibodies in plasma (Aug 2016 conference news; Irizarry et al. 2016). Full trial results are published (Lowe et al., 2021). Only the 10 mg/kg dose reduced amyloid deposits, and the drug had a shorter than expected half-life of 10 days. Most people developed anti-drug antibodies within three months of dosing.

In December 2015, Lilly started a second Phase 1 study in 150 people with prodromal to moderate AD, again in the United States and Japan. This trial had three dosing regimens, the first consisting of a single dose of 10, 20, or 40 mg/kg, the second of 10 mg/kg every other week for 24 weeks, and the third of 10 or 20 mg/kg every month for 16 months. Participants were randomized 3:1 to drug or placebo. The study measured primarily donanemab's effect on brain amyloid load with florbetapir PET; secondary outcomes were blood pharmacokinetics of donanemab and auto-antibodies directed against this biologic drug. At the 2018 AAIC, Lilly reported that a six-month course of 20 mg/kg shrank amyloid load by an average of 70 centiloids, with three of the six people who had reached that timepoint falling below the threshold for amyloid positivity. About one in four participants developed ARIA-E, mostly asymptomatic. Nearly all patients made anti-drug antibodies, and one had an infusion reaction (Aug 2018 conference news). 

This trial ended in August 2019, having enrolled 61 people. According to a presentation at the 2019 CTAD conference, monthly doses of 10 or 20 mg/kg for 16 months reduced amyloid by an average of 90 to 100 centiloids. All five people on the higher dose fell below the amyloid-positivity threshold. As observed at the earlier timepoint, one-quarter of patients developed ARIA-E; two cases were symptomatic and resolved after dosing stopped (Dec 2019 conference news). Trial results were peer-reviewed and published (Lowe et al., 2021).

In December 2017, Lilly began TRAILBLAZER-ALZ, a combination trial of two investigational drugs targeting different points in the amyloid cascade. This Phase 2 study was to evaluate safety, tolerability, and efficacy of an 18-month course of donanemab alone and in combination with Lilly's BACE inhibitor LY3202626. It aimed to enroll 375 participants whose memories had been worsening for at least six months, who met a cutoff on the CogState Brief Battery, and who had positive flortaucipir PET scans between 1.15 and 1.46 SUVR. Within this range, people have detectable cognitive decline over one to two years, but are not yet at advanced disease stages. Because the antibody requires an infusion and the BACE inhibitor comes as a capsule, the three treatment arms in this blinded, placebo-controlled trial were as follows: One group received intravenous donanemab plus placebo administered orally, the second group received intravenous donanemab plus LY3202626 orally, the third group received both intravenous and oral placebo. Hence the trial was to evaluate the antibody alone and in combination with the BACE inhibitor, but not the BACE inhibitor alone. Donanemab was dosed at 700 mg monthly for the first three months, then 1,400 mg for up to 18 months. The primary outcome was change on the Integrated Alzheimer's Disease Rating Scale (iADRS), a combined cognitive/functional measure for early stage AD developed by Lilly (Wessels et al., 2015). Secondary measures included the ADAS-Cog13, CDR-Sum of Boxes, MMSE, ADCS-iADL, as well as amyloid and tau PET and volumetric MRI. To ease the logistical burden of trial participation, this study offered rides to and from study sites via a partnership between the Global Alzheimer’s Platform Foundation and the ridesharing company Lyft. The study started up at 63 sites in North America and was set to run until December 2020.

In October 2018, Lilly discontinued the BACE inhibitor arm of this trial; evaluation of donanemab was completed in December 2020, with a final enrollment of 272 participants.

In January 2021, Lilly announced by press release that TRAILBLAZER-ALZ had met its primary endpoint, with donanemab slowing decline on the iADRS by 32 percent compared to placebo at 18 months (Jan 2021 news). The company claimed improvement on all secondary endpoints of cognition and function, although not all were statistically significant. Treatment resulted in an average reduction in amyloid plaque by 84 centiloids, from 108 at baseline. Safety was similar to earlier trials. ARIA-E developed in 27 percent of treated patients, with 6 percent becoming symptomatic. According to data presented at AD/PD and simultaneously published, two-thirds of treated participants were amyloid-negative by the end of the trial. Donanemab slowed the rate of accumulation of tau neurofibrillary tangles in the frontal cortex and other regions (Mar 2021 conference news; Mintun et al., 2021). All secondary clinical measures trended in favor of treatment, but only the ADAS-Cog13 slowed significantly, by 39 percent. In addition to ARIA-E, treated participants also had more ARIA-H, superficial siderosis due to small brain bleeds, and nausea. The groups did not differ in serious adverse events or death. Ninety percent of patients developed anti-drug antibodies. Like other anti-amyloid treatments, donanemab treatment accelerated brain shrinkage assessed by MRI. Loss of brain volume is sometimes attributed to amyloid removal; however, the extent of this effect and timing relative to donanemab administration may point to other causes, possibly inflammation (Ayton, 2021). In a “time saved” analysis, treatment resulted in an estimated 5 months delay in progression on the iADRS and CDR-SB (Dickson et al., 2023).

At the July 2021 AAIC, the company showed additional data from TRAILBLAZER-ALZ. Donanemab led to rapid amyloid reduction in the first six months of treatment, which was sustained to the end of the trial. Participants who cleared amyloid below a brain-wide threshold, and were switched to placebo, did not reaccumulate amyloid after one year. People with amyloid clearance also showed a reduction in tau burden in the temporal, parietal, and frontal lobes, and a significant decline in the biomarker plasma ptau217. At the end of the trial, ptau217 concentrations fell by 24 percent in the treated group, while rising 6 percent in placebo (see Aug 2021 conference news). A post hoc analysis of the data was subsequently published, showing that people with high baseline amyloid had greater plaque reduction in the first six months, but less chance of reaching complete clearance compared to those with lower brain amyloid. Slowing of tau accumulation was more pronounced in people with complete amyloid clearance, and in brain regions affected later in disease. The extent of amyloid reduction correlated with change on the iADRS clinical endpoint only in ApoE4 carriers (Shcherbinin et al., 2022). In October, Lilly published additional fluid biomarker results from this trial, showing that plasma GFAP had fallen, but plasma NfL continued to rise (Pontecorvo et al., 2022).

More data on side effects was presented at the April 2023 AD/PD conference. In Phase 2, 92 percent of participants developed neutralizing antibodies to donanemab. Those with high titers of anti-drug antibodies showed lower blood concentrations of donanemab between doses, less plaque clearance, and about one-quarter less efficacy. People with higher anti-drug antibodies were also more likely to have infusion reactions (Apr 2023 conference news). A population pharmacokinetic analysis from a Phase 1 and this Phase 2 trial reported that donanemab's elimination half-life was 11.8 days and that antidrug antibodies reduced donanemab exposure (Gueorguieva et al., 2023).

In November 2020, an open-label extension began to enroll 100 participants for a one-month validation of at-home assessments, followed by 18 months of donanemab. The primary endpoints are correlation between video teleconference and in-clinic scores on ADAS-Cog, ADCS-ADL, MMSE, CDR-SB in the first month, and safety over the entire dosing period. Secondary outcomes are changes in cognitive and functional measures, amyloid PET, and MRI from baseline to 18 months, as well as pharmacokinetics, and development of anti-donanemab antibodies. The extension will run through early 2024.

In October 2020, Lilly began recruiting for TRAILBLAZER-ALZ 2, initially as a Phase 2 safety and efficacy trial in 500 people with early Alzheimer’s. Inclusion criteria are similar to TRAILBLAZER-ALZ, i.e., participants must have had a progressive and gradual memory decline for at least six months, MMSE scores between 20 and 28, and meet criteria on amyloid and tau PET scans. Participants will receive donanemab or placebo, with the primary outcome being change in CDR-Sum of Boxes after 18 months. Secondary measures include the MMSE, ADAS-Cog13, iADRS, and ADCS-iADL, amyloid and tau PET, and volumetric MRI, plus pharmacokinetics and measures of anti-donanemab antibodies. The trial was set to run through early 2024 at 87 sites in the United States, Canada, Japan, The Netherlands, and Poland.

Subsequently, Lilly enlarged TRAILBLAZER-ALZ 2 to become a Phase 3 registration study with 1,800 participants. The ongoing study had already enrolled some people with tau-PET above 1.46 SUVR, but primary efficacy will be determined in 1,000 people whose tau burden is below this cutoff. The primary outcome is iADRS, and effectiveness will be judged using a disease-progression model, rather than solely on change at the final time point. As of October 15, 2021, 1,625 participants had been randomized (Nov 2021 conference news). Results were expected in the fall of 2023.

In June 2021, the FDA granted donanemab Breakthrough Therapy designation, to speed development and review. Lilly submitted a licensing application in October 2021, under the same accelerated approval pathway used for aducanumab, with rolling submission of trial data (press release).

In August 2021, Lilly and the Banner Alzheimer’s Institute began a placebo-controlled Phase 3 prevention trial. Called TRAILBLAZER-ALZ3, it aimed to enroll 3,300 cognitively normal people, age 55 to 80 and at high risk for AD based on elevated plasma ptau217. The primary outcome is time to clinical progression measured by the Clinical Dementia Rating–Global Score. After nine monthly infusions of donanemab or placebo, participants will be monitored every six months until 434 become cognitively impaired, defined as a score above zero on the CDR for two consecutive evaluations. Secondary outcomes include a battery of cognitive tests, steady-state donanemab concentration in plasma, and the incidence of anti-donanemab antibodies. This is a decentralized trial, where assessments and monitoring are done remotely using video calls and tablet computers; drug infusions, blood draws, and MRIs will be done at local facilities. Participants will report to study centers only for amyloid or tau-PET scans. In mid-2022, the minimum age was raised to 65, and in 2023, enrollment was adjusted downward to 2,600. The trial is running at more than 200 sites in the United States, Japan, and Puerto Rico until November 2027 (Jul 2021 news; Nov 2021 conference news).

In November 2021, Lilly began TRAILBLAZER-ALZ 4, a Phase 3, open-label, head-to-head comparison of plaque clearance by donanemab and Biogen’s Aduhelm. Two hundred participants with a positive amyloid PET scan and mild dementia were to be randomized to either aducanumab treatment per label or intravenous donanemab every four weeks for 18 months. The primary outcome is the number who reach complete plaque clearance after six months on each treatment, judged by florbetapir PET. Seventeen secondary outcomes all relate to amyloid PET measures out to 18 months. The trial, at 31 locations in the United States, reached primary completion with 140 participants in September 2022. According to results presented at the 2022 CTAD conference, donanemab removed four times more plaque than aducanumab in the first six months of treatment (Dec 2022 conference news). Thirty-eight percent of people on donanemab fell below the amyloid positivity threshold by six months, compared to 2 percent for aducanumab. Plasma p-tau217 fell 25 percent on donanemab, but not at all on aducanumab. ARIA-E was similar in both groups, occurring in 22 percent of participants. The trial was completed in September 2023, and results are posted at clinicaltrials.gov.

In August 2022, Lilly registered TRAILBLAZER-ALZ 5, another Phase 3 safety and efficacy trial in participants with early symptomatic Alzheimer’s. The trial began in October 2022 to enroll 1,500 participants with the same entry criteria as TRAILBLAZER-ALZ 2. They will receive monthly infusions of donanemab or placebo, against a primary outcome of change on the iADRS at 18 months. Planned for 148 sites in China, Korea, Taiwan, and Europe, the study will run until mid-2027. The company concurrently ran a Phase 1 safety trial involving 36 healthy Chinese people in Beijing, which was completed in January 2023.

From October 2022 to July 2023, Lilly conducted a Phase 1 study assessing single-dose blood exposure and degradation in 42 healthy volunteers.

In January, 2023, the FDA rejected donanemab’s accelerated approval application, citing insufficient safety data. Fewer than 100 people had stayed on drug for one year in the Phase 2 trial (news).

In February 2023, a Phase 3 trial, TRAILBLAZER-ALZ6, began testing the effect of different donanemab dosing regimens on the frequency and severity of ARIA-E in people with early AD. The 800 participations are to be divided among four dose schedules, including placebo when required to preserve blinding, against a primary outcome of percent of participants with ARIA-E over six months. The study also measures amyloid plaque clearance, ARIA-H, pharmacokinetics, and anti-drug antibodies for up to two years. It is ongoing at 47 sites in the U.S. and U.K., with completion planned for May 2025.

On May 4, 2023, Lilly announced positive top-line results for Trailblazer-ALZ2 (news). In this Phase 3 study, treatment significantly slowed decline on the primary outcome of iADRS by 40 percent, and improved all secondary clinical endpoints. Incidence of ARIA was similar to Phase 2, with ARIA-E and ARIA-H affecting one quarter and one third of those on drug, respectively. Serious ARIA occurred in 1.6 percent of participants, and caused two deaths, with a third possibly related. The company said it planned to apply for FDA approval in the next months. Full results were presented at the Alzheimer’s Association International Conference in Amsterdam (July 2023 conference news), and published the same day (Sims et al. 2023). Significant slowing of cognitive decline was apparent after three months on treatment for the iADRS, and six months for the secondary outcome CDR-SB. This amounted to a 4.5 months delay in progression on the iADRS, and 7.5 months on the CDR-SB by the end of the study. People who became amyloid negative continued to show slower decline after stopping the antibody. Benefits were greatest for those with mild cognitive impairment, who had 60 percent slowing on iADRS. Treatment reduced blood p-Tau and GFAP biomarkers, but did not change tangle load assessed by tau PET. Twenty-five percent of participants developed ARIA, six percent were symptomatic, and two percent had serious symptoms. Three people on donanemab died from brain bleeds during the trial. Two were ApoE4 carriers; none were taking anticoagulants. Data presented at the 2023 CTAD conference showed that people with lower tau at baseline and younger in age benefited more from treatment (Nov 2023 conference news). A post-hoc subgroup analysis of trial data, presented at the same conference, suggested that tangle accumulation slowed in people who had rapid removal of amyloid (Nov 2023 conference news).

According to investor documents on the company web site, Lilly applied to the FDA for traditional approval of donanemab in the second quarter of 2023, and expects regulatory action early in 2024.

For all trials of this antibody, see clinicaltrials.gov.

Clinical Trial Timeline