Slide 1: An exploration of amyloid removal measures in relation to clinical benefit
An exploration of amyloid removal measures in relation to clinical benefit: a review and meta- regression of anti-amyloid trials in AD Marzia A Scelsi1, Josephine Jackson1, Matteo Tonietto2, Gregory Klein2, Christopher Lane1, Janice Smith1, Rachelle S Doody2, Paul Delmar2 1 Roche Products Ltd, Welwy
Slide 2: Disclosures
Disclosures § Marzia Scelsi and Josephine Jackson are employees of Roche Products Ltd § Janice Smith and Christopher Lane are employees of Roche Products Ltd and own stocks or stock options in F. Hoffmann-La Roche Ltd § Paul Delmar, Matteo Tonietto, and Gregory Klein are employees of F. Hoffmann-La
Slide 3: Aβ PET reduction versus treatment effect on CDR-SB: a paradox?
Aβ PET reduction versus treatment effect on CDR-SB : a paradox? Comparable PET reduction with gantenerumab and lecanemab, but different study outcomes Gantenerumab [1] Lecanemab [2] Baseline centiloid value Baseline centiloid value B Amyloid Burden on PET Placebo 94.9 Gantenerumab 91.8 Placebo 92.9
Slide 4: Can we gain insights using aggregated data across molecules?
Can we gain insights using aggregated data across molecules ? Objectives of this work ➢ Review the literature about amyloid plaque-targeting monoclonal antibodies to investigate the following questions: ○ As a Correlate/Predictor of CDR-SB Relative Reduction (RR) at the end of study … ■ Which is the
Slide 5: Overview of available amyloid PET data
Overview of available amyloid PET data Bold blue X indicates primary efficacy endpoint read-out Molecule Study Amyloid PET Month Active arms Month 3 Month 6 Month 12 Month 18 24/27 Gantenerumab Marguerite RoAD [1] X X 1 Gantenerumab GRADUATE I & II [2] X X 2 Aducanumab EMERGE & ENGAGE [3] X X 4 Leca
Slide 6: Analysis methods
Analysis methods Literature review Amyloid PET metrics § Publications and conference presentations for aducanumab, § Centiloid change from baseline, treated arm lecanemab, donanemab; Roche internal gantenerumab data § Centiloid change from baseline, difference treated - placebo § Amyloid PET data ha
Slide 7: Numerically higher correlation of CDR-SB RR with absolute Aβ PET level at early timepoint
Numerically higher correlation of CDR-SB RR with measures of absolute Aβ PET level at early timepoint Centiloid value at earlier timepoint vs relative reduction in CDR-SB at latest timepoint Relative Reduction in CDR-SB: p=-0.85, p= 0.0035 30% Latest Timepoint PET timepoint in 20% relation to final
Slide 8: Correlation with CDR-SB change from baseline does not show a clear pattern
Correlation with CDR-SB change from baseline (difference from placebo) does not show a clear pattern Centiloid value at earlier timepoint vs diff. from placebo in change in CDR-SB at latest timepoint Difference from placebo in change from baseline CDR-SB: Latest 0.0 p=0.82, p= 0.0036 PET timepoint i
Slide 9: Predictors of CDR-SB relative reduction at early timepoint differ from late
Predictors of CDR-SB relative reduction at early timepoint are different from those at late timepoint Absolute centiloid and % below positivity threshold are the best predictors early, but have large variability (12 months before final timepoint) 10 Difference between observed and predicted CDR-SB §
Slide 10: Early PET closely predicts results of gantenerumab GRADUATE studies
Early PET closely predicts results of gantenerumab GRADUATE studies Early absolute centiloid and % below threshold give the closest predictions of CDR-SB RR GRADUATE I GRADUATE II 20 Predicted CDR-SB relative reduction (%) at final timepoint (12 months before final timepoint) § Observed CDR-SB RR [1
Slide 11: Early PET closely predicts results of lecanemab CLARITY-AD study
Early PET closely predicts results of lecanemab CLARITY-AD study Early absolute centiloid and % below threshold give the closest predictions of CDR-SB RR CLARITY-AD Amyloid PET metric (12 months before final timepoint) Predicted CDR-SB relative reduction (%) at final timepoint 30 § Observed CDR-SB R
Slide 12: Early PET profile as good as late in predicting donanemab TRAILBLAZER-ALZ 2
Early PET profile as good as late in predicting results of donanemab TRAILBLAZER-ALZ 2 study Amongst the early PET metrics, absolute centiloid gives the closest predictions of CDR-SB RR TRAILBLAZER-ALZ 2 30 (12 months before final timepoint) Predicted CDR-SB relative reduction (%) at final timepoint
Slide 13: Summary
Summary § Early removal of amyloid plaques was evaluated as an indicator of potential clinical benefit based on published summary-level data across clinical trials of Aβ plaque-removing antibodies § Early (12 months before final timepoint) absolute Centiloid and % below threshold were numerically be