Protocol Version 1.5 was issued in November 2011 (12 subjects were enrolled in Spain only under this amendment) and implemented the following changes in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • Revision of the control group and blinding technique following The Cochrane Library about the invasive “placebo” controls* to avoid the use of sham procedures * Cyna AM, Costi D, Middleton P. Randomised controlled trials using invasive ‘placebo’ controls are unethical and should be excluded from Cochrane Reviews [editorial]. The Cochrane Library 2011 (15 June). • Inclusion of Columbia-Suicide Severity Rating Scale (C-SSRS), Quality of Life AD Measure (QoL-AD) and Resource Utilization in Dementia (RUD-Lite©) Questionnaire • Update to timing for assessment of efficacy measurements from baseline (week -2 to -1), week 6, month 5, 8, 11 and 14 to baseline (week -2 to -1), intermediate visit (week 7-8), months 6, 9, 12 and 14. • Inclusion of Appendix 11 – STUDY SUMMARY TABLE AND STUDY FLOW CHART and modification of Appendix 3 - AUTO-C CLINICAL PROTOCOL

Protocol Version 1.6 was issued in April 2012 (after 12 subjects had been enrolled under the previous version, 6 new subjects were enrolled in Spain only under this amendment) and implemented the following changes in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • Substitution of secondary efficacy variable Boston Naming Test (BNT) to Neuropsychological Assessment Battery Naming Test (NAB) • Update to Appendix 4 - REPORTING OF SERIOUS AND/OR UNEXPECTED AES, elimination of Appendix 5 - STUDY CONTACT INFORMATION • Update to inclusion criteria#3 related to include current stable treatment of memantine and for inclusion criteria #6 related to neuroimaging was clarified. • Extend timing for screening phase to 3 weeks; and provide time-window for FPE visits (window of ± 1 day), Intermediate visit (window of ± 2 days) and LVPE visits window of ± 5 days) • The number of scheduled MRI and FDG-PET assessments for brain structural changes was increased from 3 to 6. • Elimination of AD genetic markers (ApoE and presenilin), update laboratory parameters and AD biomarkers.

Protocol Version 1.7 was issued in November 2012 (after 18 subjects had been enrolled under the previous version, no new subjects were enrolled under this amendment) and implemented the following changes in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • Update to Appendix 3 – AUTO-C CLINICAL PROTOCOL and Appendix 4 - REPORTING OF SERIOUS AND/OR UNEXPECTED AES, elimination of Appendix 8 – SAMPLE INFORMED CONSENT TEMPLATE and Appendix 9 - SAMPLE CAREGIVER CONSENT TEMPLATE • Modification of Investigational Plan since the third group of treatment was changed from 1/3 of albumin dose (13 g) and IVIG (7 g) to albumin dose only group (20 g) • Clarification of exclusion criteria #1 (plasma exchange contraindications), #10 (uncontrolled high blood pressure), #12 (heart diseases), #16 (years of education) and #17 (stable treatment for behavioral disorders) • Update of safety assessments to include a chest X-ray to confirm the correct placement of the catheter; vital signs were assessed at 15-30 min before PE, during and 15-30 min after PE; addition of monitoring of subjects for number and type of bacterial infections requiring antibiotic, and addition of troponin assessment at screening and at each post-PE during intensive FPE period. • Update timing for primary criterion of safety (percentage of PE associated with at least one procedure-related AE) to be assessed within 72 hours after infusion completion (or after the infusion stops). • Incorporation of Data Monitoring Committee (DMC)

Amended Protocol Version 2.0 was issued in January 2013 (after 18 subjects had been enrolled under the previous version, 107 new subjects in Spain and US were enrolled under this amendment) and implemented the following changes in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • The planned enrollment was increased from 350 to 364 subjects to obtain an estimated sample size of 312 subjects (78 per treatment group) for evaluation. This sample size would provide joint power for the 2 co-primary endpoints (implemented in this protocol version; see below) of at least 90%. • Sham procedures mimicking plasmaphereses but with neither fluid exchange nor albumin or IVIG administration were implemented for the control group, following the same visit and assessment schedule used for the treated groups. The control group previously did not undergo these procedures and did not have visits or assessments during the intensive PE period (Weeks 1 to 6). • Exclusion criteria for uncontrolled high blood pressure (despite regular treatment during the previous 3 months) were increased from >140 to ≥160 mmHg (systolic) and from >90 to ≥100 mmHg (diastolic). • Change from baseline in ADCS-ADL inventory was added as a co-primary efficacy variable. • The number of scheduled MRI assessments for brain structural changes was decreased from 6 to 5. • The number of scheduled assessments for variation in FDG-PET patterns was decreased from 6 to 4. • aPTT was added to the scheduled laboratory assessments.

Amended Protocol Version 3.0 was issued in December 2015 (after 125 subjects had been enrolled under the previous versions, 222 new subjects in Spain and US were enrolled under this amendment) and implemented the following changes in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • Guidelines for sham procedures and management of hypovolemia, fever, and thromboembolic events were added. • The exclusion criterion for bradycardia was decreased from <60 to <55/min. • An exclusion criterion was added for subjects being treated with anticoagulants or antiplatelet therapy. • A provision was added that a person accompanying the subject for study procedures should not be in the same room in order to maintain the blind except in cases of agitated subjects. • Premedication with corticosteroids for IVIG administration was added as a permitted concomitant treatment. • ACEIs were added as prohibited concomitant treatments. • Sampling of plasma bags corresponding to each PE (FPE and LVPE) at selected sites was added. • Central interpretation of FDG-PET and MRI assessments was added. • The requirement for immediate notification of unexpected AEs was removed.

Amended Protocol Version 4.0 was issued in November 2017 (after 347 subjects had been enrolled under the previous versions, no new subjects were enrolled under this amendment) and implemented the following change in study conduct (excluding administrative and typographical changes or clarifications to the protocol): • Sites performing LVPE were allowed to use devices based on the Aurora™ device (Fresenius Kabi, Bad Homburg, Germany) for the procedure. This device is an upgrade of the Auto-C device.

Amended Protocol Version 5.0 was issued in February 2018 in the US only, and administrative and typographical changes were implemented to the protocol. No changes to study conduct were specified, and no new subjects were enrolled under this amendment.