Background

Crenezumab represents a passive immunotherapy approach in which patients are treated with monoclonal antibodies that specifically recognize Aβ peptides. Crenezumab recognizes multiple forms of aggregated Aβ, including oligomeric and fibrillar species with high affinity, and monomeric Aβ with low affinity. This humanized antibody uses an IgG4 backbone. It was engineered to clear excess Aβ while exerting reduced subsequent effector function on microglia; the rationale was to stimulate amyloid phagocytosis while limiting release of inflammatory cytokines as a way to avoid side effects such as vasogenic edema (see Adolfsson et al., 2012). In terms of its binding specificity, crenezumab is similar to solanezumab (May 2015 news).

Crenezumab was confirmed to bind Aβ oligomers in vivo. In the PS2APP mouse model of amyloidosis, the antibody localized to brain regions rich in Aβ oligomers, including the periphery of plaques, and hippocampal mossy fibers. It did not bind to the dense core region of plaques, or vascular amyloid (Meilandt et al., 2019).

Findings

Two Phase 1 safety trials in healthy volunteers and people with Alzheimer's disease produced no evidence of vasogenic edema and a low incidence of asymptomatic cerebral microhemorrhage (Guthrie et al., 2020). This allowed Phase 2 to use higher doses and achieve higher brain exposure than was possible with previous immunotherapy approaches.

ABBY was a Phase 2 trial evaluating up to 15 mg/kg per month of subcutaneous crenezumab injections, conducted at more than 90 sites in North America and Europe in 450 people with mild to moderate AD. ABBY, and a 91-patient biomarker study called BLAZE, were completed in spring 2014; both continued into an open-label extension trial of 361 patients that ran until May 2016. ABBY missed its primary endpoints of change on ADAS-cog and CDR-SOB. Further analysis suggested a possible efficacy signal in mild AD, similar to solanezumab's EXPEDITION 1 and 2 results (Jul 2014 conference news; Cummings et al., 2018). BLAZE reported no separation between treatment and placebo groups on the primary endpoint of PET amyloid imaging, but did report a separation on the secondary endpoint of CSF Aβ (Dec 2014 conference news; Salloway et al., 2018; Yoshida et al., 2020). Among 104 participants of both ABBY and BLAZE, crenezumab reduced CSF Aβ oligomers in treated participants (Yang et al., 2019). An ultrasensitive immunoassay for oligomers detected a median 43 percent to 48 percent decline after treatment, compared to 13 percent in the placebo group. More than 80 percent of treated people had lower oligomeric Aβ, whereas in the placebo group, CSF oligomeric Aβ declined in half and increased in the other half.

Crenezumab was also tested in a prevention paradigm. In a landmark, adaptive, five-year study that started in 2013, crenezumab was the first immunotherapy to be evaluated as part of the Alzheimer Prevention Initiative (API, see May 2012 conference news). Bimonthly, subcutaneous injections of crenezumab were compared to placebo for its ability to stave off cognitive decline and affect Alzheimer's biomarkers in presymptomatic carriers of the autosomal-dominant PSEN1 E280A mutation. Most trial participants live in and near Medellin, Colombia; some U.S. participants with similarly predisposing gene mutations also were recruited. The participants in this trial did not meet criteria for mild cognitive impairment at the time of enrollment. The trial used a composite consisting of five separate cognitive tests as its primary outcome (Feb 2013 webinar). It also used an extensive list of secondary outcomes, including safety, time to progression to MCI, as well as clinical outcomes and fluid and imaging biomarkers (see Tariot et al., 2018).This Phase 2 trial was expected to recruit 300 participants, but ultimately randomized 252 (Aug 2019 conference news; Rios-Romanets et al., 2020). This trial enrolled 150 participants into a tau PET substudy to evaluate the effect of crenezumab treatment on the change in tau burden over time in carriers of this mutation.

Following results of ABBY and of solanezumab's EXPEDITION trial, discussion centered on increasing the dose in subsequent crenezumab trials. In February 2015, Genentech started a Phase 1b study in 72 people with mild to moderate AD to compare three doses of intravenous crenezumab to placebo. Doses were not disclosed, but a three-month, double-blind course of monthly infusions was followed by a 12-month option of open-label dosing.

In July 2015, crenezumab was advanced into Phase 3, initially with a single study in prodromal AD. In January 2016, a study of 750 people with MCI or prodromal AD with biomarker evidence of Aβ pathology started enrolling. Called CREAD, this trial used change on the CDR-SB as its primary outcome and a range of cognitive and functional measures as secondary outcomes. CREAD used 233 study locations globally and was expected to run until 2020 (see also Blaettler, 2016).

At the 2016 CTAD conference, crenezumab's sponsors announced results of the 72-patient Phase 1 trial, as well as of exposure-response modeling done in a disease-progression simulation model. Company scientists claimed that both datasets predicted a stronger treatment benefit from the higher dose of 60 mg/kg of crenezumab infused once a month for the CREAD Phase 3 study, also confirming that this was the dose being evaluated in this trial.  

On February 28, 2017, AC Immune announced that Genentech had decided to start a second Phase 3 trial of 750 participants with prodromal to mild AD, to be called CREAD2 (press release).

In January 2019, Roche terminated both Phase 3 CREAD trials (Jan 2019 news). Initial results for CREAD1, presented in March at the AD/PD conference in Lisbon, Portugal, were wholly negative (May 2019 news). The data available by that time covered 13 percent of the 813 enrollees who had completed the treatment period of the trial as of January 2019; 14 percent had dropped out. Treatment and placebo groups were identical on the primary and secondary outcomes. Subgroup analyses by prodromal versus mild disease stage at baseline, by MMSE greater or less than 24, or by ApoE status yielded no treatment signal. The high dose used in this trial produced numerically more side effects than placebo, but the differences were small and raised no new safety concerns. According to more data presented at CTAD in 2019, treatment did not significantly affect amyloid PET, brain volume, or CSF tau measures. Other CSF markers—Aβ, neurogranin, NfL, α-synuclein, YKL-40, and GFAP—moved in the desired direction with treatment, indicating target engagement (Dec 2019 conference news). Full results for both trials were published after peer review (Ostrowitzki et al., 2022). Together, the trials enrolled 1,619 participants, with 173 completing two years of treatment in CREAD1. In CREAD2, no one reached two years of treatment before the trials were terminated. As previously reported, crenezumab was well-tolerated, but changed neither rate of decline on the CDR-SB nor biomarkers.

From 2014 to 2019, Genentech ran two Phase 1 studies evaluating high-volume, high-flow subcutaneous infusion of crenezumab. The placebo-controlled studies enrolled 140 healthy people and compared different doses, volumes, and flow rates, with or without the permeation enhancer hyaluronidase. This enzyme transiently disrupts the extracellular matrix of the skin to accommodate larger infusion volumes and enhance bioavailability. According to published results, subcutaneous infusions of 6,800 mg antibody in up to 40 ml, at flow rates up to 4 ml/minute, were well-tolerated, with or without recombinant human hyaluronidase. Participants reported low pain scores during the abdominal infusions, and no serious or dose-limiting adverse events. Transient mild skin redness (erythema) was the most common infusion site reaction and occurred in most people, with larger areas involved at higher infusion volumes. Swelling at the infusion site was also common, and lessened when hyaluronidase was included. Pharmacokinetics were dose-proportional, and 6,800 mg infused in 40 ml over 10 minutes gave the same blood levels as the 60 mg/kg intravenous infusion used in the CREAD trial (Dolton et al., 2021).

On June 15 2022, Roche released topline results of the API Colombian prevention trial (June 2022 news). The trial was negative on its original primary, the API ADAD composite, and the coprimary of the Free and Cued Selective Reminding Test. The latter was added after the trial had started. Trends on both primaries and multiple secondary and exploratory endpoints favored crenezumab. After two upward dose adjustments during the trial, participants received 60 mg/kg antibody per month, a sevenfold increase over the beginning dose. Most participants were on the high dose for about two years, with no new safety issues. Ninety-four percent completed the trial. According to data presented at the August 2022 AAIC, treatment slowed declines by about 20 percent on both primary outcomes, but with large variability within treatment groups. Other measures showed a slowing that did not reach statistical significance; they were the CDR-SB and CDR, RBANS, progression to MCI or dementia, FDG-PET, brain shrinkage, tau PET, and CSF p-tau181, total tau, and NfL. The only significant changes were a stabilization of Aβ42 and a rise in Aβ40 in the treated group compared to placebo (Aug 2022 conference news). According to information presented in December 2022 at CTAD, plasma biomarkers trended in the direction of a treatment benefit but fell short of statistical significance.

The trial has been completed; data and samples are being made available (Reiman et al., 2022). Roche has stopped development of crenezumab (Dec 2022 conference news).

This antibody is listed in clinicaltrials.gov under both crenezumab and MABT5102A.

Clinical Trial Timeline