Today, Roche and Genentech announced negative top-line results for their anti-amyloid antibody gantenerumab. In the two Phase 3 GRADUATE trials, the drug failed to slow cognitive decline on the CDR-SB. In addition, it removed less amyloid plaque than expected, according to a press release. More data will come on November 30 at the Clinical Trials on Alzheimer’s Disease conference in San Francisco.

Disappointed researchers noted the contrast with the positive Phase 3 results for lecanemab (Sep 2022 news). They expect the datasets to be discussed at CTAD will shed light on the differing effects of these two antibodies.

“It will be very important to compare the full results of the gantenerumab studies with those of lecanemab to understand the implications,” Paul Aisen at the University of Southern California, San Diego, wrote to Alzforum. “Based on the top-line gantenerumab results, it would seem that insufficient reduction of amyloid in brain, perhaps related to the subcutaneous treatment regimen, resulted in the disappointing clinical findings.”

Lon Schneider at USC, Los Angeles, noted that despite the statistically negative findings, gantenerumab slowed decline on the CDR-SB by 0.31 points in one of the two trials. “[This] is in the same range as the other anti-amyloid antibodies,” he noted. However, this numerical difference represented only an 8 percent slowing of decline, because the rate of decline in the GRADUATE trials was more than twice that seen in trials of aducanumab and lecanemab. “One wonders about exact eligibility criteria and study conduct … secondary and post hoc analyses should be interesting, to say the least,” he wrote (comment below).

To Stephen Salloway at Butler Hospital in Providence, Rhode Island, too, these analyses will be crucial to understanding what happened. “It will be important to determine whether those who became amyloid-negative on PET had a larger clinical response, and whether treating with a higher dose could be safely instituted to produce greater efficacy,” he wrote (comment below).

Of the four anti-amyloid antibodies that reached Phase 3, gantenerumab was the only one that had never demonstrated a cognitive benefit in a placebo-controlled trial. In earlier Phase 3 studies, the antibody was given at low doses that barely budged plaque and had no effect on cognition (Dec 2014 news). In open-label extensions of these trials, however, high-dose gantenerumab did mop up plaque, with hints that greater clearance correlated with better cognition (Dec 2017 news; Dec 2019 news; Aug 2022 news). Furthermore, the antibody moved biomarkers in the DIAN-TU secondary prevention trial, lowering p-tau181 and total tau and slowing the rise of the neurodegeneration marker NfL (Jun 2021 news).

Gantenerumab remains in an open-label extension of the DIAN-TU study, and had been selected for DIAN's planned primary prevention trial. It is unclear how those plans will change. According to Endpoints News, Roche is pausing enrollment in the Skyline secondary prevention study investigating gantenerumab in amyloid-positive, cognitively normal people (Mar 2022 news).—Madolyn Bowman Rogers


  1. As before, it’s chancy to make inferences from press releases that are designed to frame a narrative, avoid presenting data, and set up potential spin.

    There may have been greater expectations for gantenerumab after the lecanemab presser presented a P value = .00005. Yet lecanemab’s CDR-SB difference of -0.45 is still very small and of clinical meaning that remains to be determined, requiring careful and valid subset analyses to try to understand the effects. We’ll soon see how that evolves.

    Here, too, the gantenerumab presser gives a CDR-SB = -0.31 for one of the trials, which is in the same range as the other amyloid antibodies, including solanezumab, which binds to soluble amyloids and not fibrils. As a group, the higher doses of the antibodies tend to give small drug-placebo CDR-SB differences between -0.3 and -0.45 regardless of differences in action.

    The Roche press release sets us up for 17 secondary outcomes and analyses, including blood marker levels. It offers a teaser that plaque reduction—although not to the extent that they expected—is correlated with clinical change. These secondary, and probably some post hoc, analyses should be interesting, to say the least. The problems will be in interpreting and evaluating them in the face of overall non-significant trials.

    A couple of unusual features of these trials were the treatment length of over two years, and the large magnitude of decline on CDR-SB scores. The placebo groups worsened about 3.2 and 3.9 points in the trials. This is more than twice that seen in the aducanumab and lecanemab trials, where worsening was about 1.5 to 1.7 points. It seems doubtful that the six-month longer follow-up can wholly explain this. One wonders about exact eligibility criteria and study conduct, and whether there were relatively more impaired participants included. Again, secondary outcomes might prove interesting.

    There are several ways to interpret Roche’s comments that gantenerumab didn’t lower plaques as much as expected. One is that the dose wasn’t high enough, or the subcutaneous route limited the higher concentrations in blood needed to penetrate the CNS. Another is that the high concentrations of Aβ monomers in blood, and in brain, prevented gantenerumab from binding to oligomers, fibrils, and plaques. (It should be noted that aducanumab, lecanemab and donanemab do not bind Aβ monomers to any great extent.) Or perhaps monomer-bound antibody complexes are transported into brain and are then, of course, unable to engage with oligomers and fibrils. If so, then didn’t Roche know this going into these large trials?

  2. These are disappointing top-line results from a well-run trial. We need to evaluate the safety and efficacy for each compound in this class and learn as much as we can from each study. There appears to be a small clinical benefit with gantenerumab that did not reach statistical significance, and a lower-than-expected amount of amyloid reduction.

    It will be important to determine whether those who became amyloid-negative on PET had a larger clinical response, and whether treating with a higher dose could be safely instituted to produce greater efficacy.

  3. It is very disappointing news—another drug trial for Alzheimer's disease fails. In Phase 3 trials this antibody has no effect on cognition. I think people working on Alzheimer's should focus not only on cognition and reduction in the two pathological hallmarks, Aβ and hyperphosphorylated tau, but also take a serious step to evaluate the safety and efficacy. So far, human clinical trials were halted mostly due to a high incidence of amyloid-related image abnormalities that is presumably due to an auto-inflammatory reaction to immunotherapy with Aβ or Aβ related antibodies. Overcoming the auto-inflammatory side effects while maintaining an effective immune response is a hurdle that must be overcome for the development of a human vaccine for AD.

  4. Passive immunotherapy using anti-Aβ antibodies is being investigated aggressively as an approach for AD treatment. Indeed, a number of anti-Aβ monoclonal antibodies (mAbs) are currently in clinical studies, while some, missing primary endpoints or having toxicity, have been discontinued. Promising recent clinical data for aducanumab and lecanemab have led to renewed enthusiasm for Aβ immunotherapy for AD, but gantenerumab failed. These antibodies recognize different epitopes in Aβ, and, even within a given epitope, bind different conformations of the peptide. The precise molecular basis by which antibodies recognize Aβ may impact the outcome of their use in therapeutic regimes.

    Previous studies showed that mAbs directed toward the N-terminal region of the Aβ peptide can bind to Aβ fibrils, leading to their disaggregation and inhibition of their neurotoxicity (Solomon et al., 1997). Indeed, studies of Aβ fibrils assembled from the synthetic peptides showed that mAbs 6C6 and 10D5 raised against the N-terminal region of the Aβ (residues 1–28) can disaggregate Aβ fibrils, restore Aβ solubility, and prevent neurotoxic effects on PC12 cells. The N-terminal epitope for mAb 6C6 and mAb 10D5 was found to be a sequential epitope within the Aβ molecule composed of only four amino acid residues (EFRH) located at positions 3–6 (Frenkel et al., 1998). 

    In contrast, the inability of mAb 2H3, raised against the N-terminal region 1–12, to modulate fibril aggregation was attributed to the low binding affinity for the EFRH sequence, and this may shed light on the importance of the specific sequence, defined as anti-aggregating epitope, on the behavior of the whole Aβ molecule (Frenkel et al., 2000). The EFRH epitope is involved in modulation of the aggregation process and acts as a regulatory site, controlling both solubilization and aggregation processes. Locking of this epitope by highly specific antibodies affects the dynamics of the entire Aβ molecule, preventing self-aggregation as well as enabling resolubilization of already formed aggregates.

    Mice immunized with an EFRH phage raise a high concentration of high affinity antibodies. These are similar, in their anti-aggregating properties, to antibodies raised by direct injection with fibrillar toxic Aβ. Such antibodies are able to sequester peripheral Aβ, thus avoiding its passage through the blood–brain barrier and, as recently shown in a transgenic mouse model, these antibodies can cross the blood–brain barrier and dissolve already formed Aβ plaques (Frenkel et al., 1999). Our pioneering data were recently confirmed by successful clinical trials.

    Comparison of aducanumab to other antibodies that recognize the N terminus of Aβ enable us to gain further insight into the structural characteristics of aducanumab that is responsible for its high selectivity for aggregated Aβ species and for the binding properties of these antibodies (Arndt et al., 2018). These antibodies recognize different but overlapping N-terminal epitopes in the Aβ peptide—for aducanumab residues 3-7, lecanemab 1-16, and gantenerumab 3–11, and 18–27—which all contain the same four residues (EFRH) in the central region of the shared epitope. Although the linear sequence recognized by aducanumab overlaps substantially with epitopes recognized by several other well-described antibodies, the specific interactions (for example, critical contacts with phenylalanine 4 and histidine 6) are different. Most notably, the interaction between aducanumab and Aβ is very shallow, with a subtle interface, in comparison with those of other anti-Aβ antibodies (Arndt et al., 2018). 

    Others display different selectivity for polymorphic variants, recognize different epitopes in Aβ, and, even within a given epitope, bind the peptide in different conformations and with different binding constants. Lecanemab, which had another successful clinical trial, binds more tightly to diffusible fibrils, and its binding to the immunodominant, B-cell epitope containing EFRH in an extended conformation seems similar to that of aducanumab.

    Antibodies may bind linear epitopes, but they also recognize conformational epitopes represented in short peptide segments. Gantenerumab, which recently failed in clinical trials, recognizes conformational discontinuous epitopes, consisting of an N-terminal domain (residues 3–11) and a central domain (residues 18–27), and it binds to the linear segment of the Aβ peptide with a low binding constant.

    Despite the great success of the mentioned antibodies, a single therapeutic monoclonal antibody may not be able to target all the pathological aggregates necessary to make an impact on the overall disease process.


    . Disaggregation of Alzheimer beta-amyloid by site-directed mAb. Proc Natl Acad Sci U S A. 1997 Apr 15;94(8):4109-12. PubMed.

    . N-terminal EFRH sequence of Alzheimer's beta-amyloid peptide represents the epitope of its anti-aggregating antibodies. J Neuroimmunol. 1998 Aug 1;88(1-2):85-90. PubMed.

    . Immunization against Alzheimer's beta -amyloid plaques via EFRH phage administration. Proc Natl Acad Sci U S A. 2000 Oct 10;97(21):11455-9. PubMed.

    . High affinity binding of monoclonal antibodies to the sequential epitope EFRH of beta-amyloid peptide is essential for modulation of fibrillar aggregation. J Neuroimmunol. 1999 Mar 1;95(1-2):136-42. PubMed.

    . Structural and kinetic basis for the selectivity of aducanumab for aggregated forms of amyloid-β. Sci Rep. 2018 Apr 23;8(1):6412. PubMed.

Make a Comment

To make a comment you must login or register.


Therapeutics Citations

  1. Gantenerumab
  2. Leqembi

News Citations

  1. Finally: Big Win on All Outcomes for Lecanemab in Phase 3 Topline Results
  2. End of the RoAD for Gantenerumab? Roche Declares Prodromal Alzheimer’s Trial Futile
  3. High-Dose Gantenerumab Lowers Plaque Load
  4. Amyloid Clearance: Check. Cognitive Benefit: Um … Maybe.
  5. Crenezumab Secondaries Negative; Gantenerumab OLE Hints at Efficacy
  6. Paper Alert: DIAN-TU Solanezumab and Gantenerumab Data Published
  7. Gantenerumab Prevention Trial in Sporadic Alzheimer's Begins

External Citations

  1. press release

Further Reading