API ADAD Composite Cognitive Test is a combination of tests, most sensitive to detect cognitive decline & progression. API ADAD Composite Cognitive Test total score were computed from: Consortium to Establish a Registry for AD (CERAD) Word List: Recall (score range=0-10); Multilingual Naming Test (score range=0-15); Mini-Mental State Examination (MMSE) for orientation to time (score range=0-5); CERAD Constructional Praxis (measure of visuospatial ability) (score range=0-11); Raven's Progressive Matrices (measure of nonverbal fluid reasoning & visuospatial abilities), Set A (score range=0-12). ADAD-API Cognitive Composite Test total score = [(Multilingual Naming Test Score/15)+(MMSE Score/5)+(Raven's Progressive Matrices Score/12)+(CERAD Word List Recall Score/10)+(CERAD Constructional Praxis Score/11)]*20. Total score ranges=0-100. Higher score=better results. Annualized rate of change in API ADAD Composite Cognitive Test was estimated using random coefficient regression model.

Modified Intent-to-Treat (mITT) population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

FCSRT assesses immediate & delayed verbal episodic memory, using Controlled Learning to optimize encoding specificity for more effective recall. Participants were to remember a list of 16 items presented on cards. Participant's task was to learn & attempt a free recall of the items, followed by a semantically cued recall of items not spontaneously produced during free recall. There were a total of three learning trials, each preceded by 20 seconds of interfering cognitive task. For first two trials, participants were reminded of any item not recalled in cued recall. Free recall (score range: 0-48) & cued recall (score range: 0-64) assessment occurred immediately, after each of three learning trials, & after a delay of approx. 30 minutes. Higher scores indicate better performance. FCSRT Cueing Index=(FCSRT Total Free Recall score-FCSRT Total Recall score)/(FCSRT Total Free Recall score-48). Score range for FCSRT Cueing Index is 0-1, with higher score=better results.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

Time to progression was calculated from the time at baseline to first confirmed diagnosis of progression from preclinical AD to MCI or from preclinical AD to dementia due to AD whichever occurred first. Preclinical AD was defined as participants who were at certain risk of developing AD dementia but did not have overt symptoms and did not meet criteria for MCI or dementia.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

Time to progression was defined as the time at baseline to first increase in the CDR global score i.e., score of > zero on the CDR Global scale. The standard CDR global scale describes five degrees of impairment in performance on each of six categories of cognitive functioning, including memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. The ratings of degree of impairment obtained on each of the six categories of function was synthesized into one global rating of dementia with a score ranging from 0 to 3 (0 = no dementia, 0.5 = questionable dementia, 1 = mild dementia, 2 = moderate dementia, 3 = severe dementia). A higher score indicates a greater degree of impairment.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

The CDR-SB is a rater administered scale and impairment is scored in the following categories: memory, orientation, judgment and problem solving, community affairs, home and hobbies and personal care. Impairment was scored on a scale in which no dementia = 0, questionable dementia = 0.5, mild dementia = 1, moderate dementia = 2 and severe dementia = 3. The 6 individual category ratings, or "box scores", were added together to give the CDR-Sum of Boxes which ranges from 0-18, with higher scores indicating greater impairment. Annualized rate of change in CDR Scale - SOB was calculated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

RBANS is a clinical tool designed to detect and characterize the earliest neurocognitive changes associated with dementia. The RBANS comprised of 12 subtests which generates age-adjusted index scores for five neurocognitive domains: Immediate Memory, Visuospatial/Constructional, Language, Attention and Delayed Memory, which are used to calculate a Total Scale Index score. RBANS Total Score = Index Score based on the (Index Score for Immediate Memory + Index Score for Visuospatial/Constructional + Index Score for Language + Index Score for Attention + Index Score for Delayed Memory). Index scores and total scores range from 40-160. A higher score indicates better cognitive functioning. Annualized rate of change in RBANS was calculated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

Amyloid deposition in brain is one of the defining neuropathologic findings of AD. Cerebral amyloid burden & effect of crenezumab on cerebral amyloid were assessed using 18F- Florbetapir PET. Florbetapir exhibits high affinity specific binding to amyloid plaques, which allows for a measurement of the relative amyloid burden as the ratio of the standard uptake values (SUVR) in a cortical composite region of interest (including the frontal, temporal, parietal, and cingulate cortices with a subcortical white matter reference region). Annualized rate of change in cerebral fibrillar amyloid accumulation was calculated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

Regional CMRgl and the effect of crenezumab on regional CMRgl was assessed using FDG-PET. FDG PET evaluates the regional brain metabolic rates for glucose as a sensitive, in vivo metabolic index of brain function. Regional CMRgl was derived from FDG PET images using an empirically predefined statistical ROI (sROI) known to be preferentially affected by CMRgl decline in participants with AD. Results are reported as standardized uptake value ratios. Annualized rate of change in CMRgI was calculated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

Cerebral atrophy correlates closely with the gradual cognitive decline in AD and can be visualized by MRI. The change in brain atrophy after treatment with crenezumab was measured by volumetric measurements using MRI which measures volumes of the key brain structures-hippocampus, ventricles, and other brain structures-and compares the volumes to standard norms based on the age, gender and cranial volume. The regions of interest for atrophy measurement included whole brain, bilateral hippocampus, and bilateral ventricles. The annualized rate of change in brain atrophy was calculated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

CSF biomarker tTau has been considered as a general marker of neurodegeneration. CSF phospho-tau is an indicator of neuronal injury and neurodegeneration. An elevation in levels of tau, as well as specific pTau species, is thought to be a marker for progressive cellular degeneration in AD. Annualized rate of change in tTau and pTau was estimated using RCRM.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug. Overall number analyzed is the number of participants with data available for analysis. Number analyzed is the number of participants with data available for analysis at specified timepoint.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. An SAE is any AE that is fatal; life-threatening; requires or prolongs inpatient hospitalization; results in persistent or significant disability/incapacity; is a congenital anomaly/birth defect in a neonate/infant born to a mother exposed to the investigational product(s); considered a significant medical event by the investigator. Pooled data has been reported for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group.

Safety population included all randomized participants who received at least 1 dose of study drug; participants were grouped according to the actual treatment received and the mutation status.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. Participants who withdrew from the study due to AEs are reported here. Pooled data has been reported for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group.

Safety population included all randomized participants who received at least 1 dose of study drug; participants were grouped according to the actual treatment received and the mutation status.

AE=any unfavorable & unintended sign, symptom/disease temporally associated with use of an investigational product/other protocol-imposed intervention, regardless of attribution. Amyloid-related imaging abnormalities (ARIA)=spectrum of image abnormalities detected on MRI. Two types of ARIAs reported are: Amyloid-related Imaging Abnormalities-Edema/Effusion (ARIA-E)=MRI alterations thought to represent vasogenic edema (VE) & related extravasated fluid phenomena, & Amyloid-related Imaging Abnormalities-Hemosiderin Deposition (ARIA-H)=MRI alterations attributable to microhemorrhages (microbleeds [MBs]) & leptomeningeal hemosiderosis. Cerebral macrohemorrhages were read on MRI scans. Occurrence of pneumonia were verified by imaging (e.g., chest X-ray). Other AESIs were drug induced liver injury & suspected transmission of infectious agent via medicinal products. Pooled data has been reported for this outcome measure to maintain blinding of participant's mutation status & treatment group.

Safety population included all randomized participants who received at least 1 dose of study drug; participants were grouped according to the actual treatment received and the mutation status.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. Infusion and injection-related reactions were defined as any AEs that occurred during or within 24 hours after the study drug injection or infusion. Pooled data has been reported for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group.

Safety population included all randomized participants who received at least 1 dose of study drug; participants were grouped according to the actual treatment received and the mutation status.

The number of participants with positive results for ADA against crenezumab at any of the post-baseline assessment time-points were reported. Participants positive at any post-baseline time points were post-baseline evaluable participants determined to have "Treatment-induced ADAs" during the study period. Treatment-induced ADA = a participant with negative or missing baseline ADA result(s) and at least one positive post-baseline ADA result.

Safety population included all randomized participants who received at least 1 dose of study drug; participants were grouped according to the actual treatment received and the mutation status.

PK-evaluable population included all safety-evaluable participants with at least 1 plasma sample, provided sufficient dosing information (dose and dosing time) is available. Overall number analyzed is the number of participants with data available for analysis. Number analyzed is the number of participants with data available for analysis at specified timepoints.

PK-evaluable population included all safety-evaluable participants with at least 1 plasma sample, provided sufficient dosing information (dose and dosing time) is available. Number analyzed is the number of participants with data available for analysis at specified timepoints.

Amyloid beta is a peptide fragment of the amyloid precursor protein. The annualized rate of change in plasma concentration of Aβ1-40 and Aβ1-42 were estimated using a linear mixed effects model with random intercept and slope.

mITT population included all randomized participants who were mutation carriers and were treated with any amount of study drug.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. An SAE is any AE that is fatal; life-threatening; requires or prolongs inpatient hospitalization; results in persistent or significant disability/incapacity; is a congenital anomaly/birth defect in a neonate/infant born to a mother exposed to the investigational product(s); considered a significant medical event by the investigator. During Study Period B, after the results of Study Period A were available, participants had an option to learn their mutation carrier status. Since all participants did not choose to learn their mutation status, pooled data has been presented for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group assigned.

Safety population included all randomized participants who received at least 1 dose of study drug in Study Period B; participants are grouped according to the actual treatment received.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. During Study Period B, after the results of Study Period A were available, participants had an option to learn their mutation carrier status. Since all participants did not choose to learn their mutation status, pooled data has been presented for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group assigned.

Safety population included all randomized participants who received at least 1 dose of study drug in Study Period B; participants are grouped according to the actual treatment received.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. ARIA are a spectrum of image abnormalities detected on MRI. The two types of ARIA reported will include: ARIA-E: refers to the MRI alterations thought to represent VE and related extravasated fluid phenomena, and ARIA-H: refers to the MRI alterations attributable to MBs and leptomeningeal hemosiderosis. Cerebral macrohemorrhages will be read on MRI scans. Occurrence of pneumonia will be verified by imaging (e.g., chest X-ray). During Study Period B, after the results of Study Period A were available, participants had an option to learn their mutation carrier status. Since all participants did not choose to learn their mutation status, pooled data has been presented for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group assigned.

Safety population included all randomized participants who received at least 1 dose of study drug in Study Period B; participants are grouped according to the actual treatment received.

An AE is any unfavorable and unintended sign, symptom, or disease temporally associated with the use of an investigational product or other protocol-imposed intervention, regardless of attribution. Infusion and injection-related reactions are defined as any AEs that occurred during or within 24 hours after the study drug injection or infusion. During Study Period B, after the results of Study Period A were available, participants had an option to learn their mutation carrier status. Since all participants did not choose to learn their mutation status, pooled data has been presented for this outcome measure in order to maintain blinding of the participant's mutation status and treatment group assigned.

Safety population included all randomized participants who received at least 1 dose of study drug in Study Period B; participants are grouped according to the actual treatment received.