• Stratification criteria modified from steroid use at baseline yes/no to steroid use at baseline <20 mg and ≥ 20 mg and baseline Hgb from < 8 g/dL and ≥ 8 g/dL to < 9 g/dL and ≥ 9 g/dL • Concurrent wAIHA therapies were modified to optimize recruitment with minimal impact on confounding the efficacy endpoints including adding table for allowed concurrent therapies, adding erythropoiesis-stimulating agents (ESA), mycophenolate mofetil, dapsone, and danazol as allowed concurrent steroid therapies. Additionally, dexamethasone added but only as a rescue treatment • Allowance added for steroid taper due to input from wAIHA experts to retain subjects for the full 24-week treatment period • Changes to the primary efficacy endpoint included changing the endpoint to proportion of subjects who achieved a durable response rather than a durable hemoglobin response where a hemoglobin response was defined as a hemoglobin level of > 10 g/dL and ≥ 2 g/dL higher than the baseline (Day 1) value if, during the previous 4 weeks, the steroid dose was maintained at the baseline dose level, and rescue medication was not administered and a durable response was defined as a hemoglobin response on at least 3 scheduled visits during the 24-week evaluation period • Changes to secondary efficacy endpoints included removing average frequency of rescue AIHA regimens used, revising average number of hemoglobin assessments exhibiting a hemoglobin response to average number of weeks with hemoglobin responses, and adding proportion of subjects requiring wAIHA rescue therapy • Inclusion criteria #2, #3, #5, #8 and exclusion criteria #6 revised • Exclusion criteria #9, #11 added • Revised washout AIHA therapies table • Removed interim analysis • Follow-up period updated from Week 28 to Week 26 • Serum pregnancy tests were removed from all study visits to be collected only at screening and baseline visits, and then monthly during the study • Changing AE and SAE reporting periods

• Number of weeks required to be at a stable dose prior to randomization for steroids changed from 3 weeks to 2 weeks • Steroid taper protocol revised such that there would be an option to taper steroid dose to as low as 10 mg/day (rather than 20). • Safety endpoints added including incidence of AEs, incidence of abnormal changes from baseline in laboratory values per CTCAE criteria Version 5.0 (e.g., hematology, chemistry), and incidence of changes in blood pressure compared to baseline • Inclusion criteria #2 revised to include documented positive DAT specific for anti-IgA in addition anti-IgG • Exclusion criteria #6 added exclusion for active HIV infection for patient safety • Statement about washout AIHA therapies removed and replaced with statement such that disallowed AIHA therapies are those that may not be taken within the indicated interval prior to Day 1. Ibrutinib or other BTK inhibitor added to list within 4 weeks prior to Day 1 • Analysis methods for primary efficacy endpoint revised from Pearson-chi-square test to compare the proportion of responders between the 2 treatment groups to Cochran-Mantel-Haenszel test adjusting for randomization stratification factors per advice of HA • Section on potential benefit-risk added based on prior studies per advice of HA • Section for justification of study design added per advice of HA • Section for treatment blinding and unblinding added to provide detailed information about the process of treatment unblinding, unblinding, and how the treatment group is assigned per advice of HA • Screening haptoglobin assessment added • Clarification added to allowed AIHA therapies section such that subjects requiring any AIHA therapies other than those that are allowed will be withdrawn from the study • Statement added for reporting of SUSARs in accordance with regulatory requirements • Planned analyses revised to add statement that all analysis were to be performed at a one-sided significance level of 0.025

• Primary objective revised from “assess the efficacy of fostamatinib in subjects wAIHA” to “compare the proportion subjects with durable hemoglobin response between fostamatinib and placebo in subjects with wAIHA” • Secondary objectives added including to compare the proportion of subjects with hemoglobin response between the fostamatinib and placebo groups, to compare the proportion of subjects with a chain from baseline hemoglobin level of ≥ 2 g/dL between the fostamatinib and placebo groups, to compare the proportion of subjects requiring AIHA rescue therapy between the fostamatinib and placebo groups, and to compare the change in hemoglobin over time between the fostamatinib and placebo groups • Several additional efficacy and pharmacoeconomic objectives and endpoints added to support the primary endpoint including comparing the fostamatinib and placebo groups for the following endpoints: any hemoglobin level > 10 g/dL within the 24 weeks of treatment (yes/no), any change from baseline hemoglobin level ≥ 1.5 g/dL within the 24 weeks of treatment, total duration of hemoglobin response (months), time to first hemoglobin response (days), change from baseline of the total daily dose of steroids (prednisone equivalent) within the 24 weeks of treatment, change in reticulocyte count, LDH, and haptoglobin over time, change from baseline to Week 24 in FACIT-F, EQ-5D-5L at baseline and Week 24, change from baseline to Week 24 in EQ-AS, and hospitalization related to a IHA within 24 weeks of treatment (yes/no) • Similarly, secondary efficacy endpoints were revised • Safety endpoints revised • Sample size revised from 80 to 90 subjects (approximately 45 subjects per treatment group). Justification for sample size revised • Promacta added to list of allowed therapies if given as a concurrent medication and dose is stable for 4 weeks prior to randomization. Dexamethasone added as well dose was stable for 2 weeks prior to randomization

• Description of procedures for blood pressure collection revised to clarify subject should rest 5 minutes before the initial measurements are collected and 3 minutes between measurements. • Added statement for clarity that subjects requiring any AIHA therapies other than those allowed, or an escalation of the allowed medications other than steroids, would be withdrawn from the study. • Guidance for restarting study drug after dose interruptions added such that Restarting study drug after interruption may require performance of procedures for safety that were missed during the interruption. The Medical Monitor should be consulted when considering study drug restart after a prolonged interruption (e.g., > 2 weeks). • Added statement in source documentation requirements that the Sponsor shall maintain the records for a period of 25 years (previously state only in country-specific amendment for Canada). • Appendix 7 added to provide guidance to Investigators on management of subjects during a pandemic.

• In general, changes from Protocol V4.1 to V5.0 reflected updates to align the protocol with the revised Statistical Analysis Plan (SAP) V0.2 (dated 13 October 2020) reviewed by Health Authorities including revision of the primary efficacy endpoint, secondary efficacy endpoints, safety endpoints, and additional efficacy and pharmacoeconomic endpoints. • Added statement to section on safety monitoring to cross-reference the Investigator’s brochure for guidance regarding infections and bone investigations (previously in country-specific amendment for France) • Planned analysis methods for enrollment exceptions and deviations and concomitant therapies added in accordance with the revised SAP. Examples of major deviations include subjects enrolled who do not meet eligibility criteria, subject use of non-permitted rescue medications, improper collection of or failure to collect initial study informed consent, subject receiving treatment contrary to their randomization assignment, and subject who developed withdrawal criteria during the study but were not withdrawn. • Section added for treatment compliance such that subject compliance with the assigned treatment will not be collected in the EDC system. The assessment of compliance will be based on site monitoring and pharmacy drug dispensation logs. All pharmacy logs will be kept in the trial master file. • Added laboratory tests for monocytes • Appendix 7 – Modified the number of extra bottles that can be sent to subjects during pandemic from 2 to 1

• Secondary objectives and secondary efficacy endpoints were revised to include comparison of change in hemoglobin value from baseline to end of treatment (EOT) and change from baseline to Week 24 in Fatigue scale (FACIT-F) (both objectives moved up from additional efficacy objectives to secondary). A description of the FACIT-F scale added • Demographics and Baseline Characteristics will be summarized in ITT not Safety or Efficacy populations • Revised methodology to be consistent with revised objectives and endpoints for extent of exposure, laboratory values, and deaths • Additional efficacy and pharmacoeconomic objectives revised to include time to first rescue medication • Safety endpoint of TEAEs of Interest added to monitor the safety signal for AEs of interest • Statement added to treatment blinding and unblinding such that the subjects, investigators, all other site personnel, and the Sponsor/Representative will remain blinded until all subjects complete the 24 weeks of evaluation or discontinue from the study • The end of study language revised for clarification such that the end of the study occurred when the last subject had completed either the Week 24 visit or their last follow-up study visit, whichever was later • Analysis of primary efficacy endpoint revised where the study will be considered to have met its primary efficacy objective if the lower bound of the 95% CI of odds ratio of durable hemoglobin response between the fostamatinib and placebo treated subjects was >1 • Sensitivity analyses for primary endpoint added including analyses based on subgroup prognostics, multiple imputation for missing data, and a washout period of 6 weeks for rescue treatment exclusion period (rather than 4 weeks) • Methodology for analysis of secondary efficacy endpoints revised • Adjustment for multiple testing revised from rejective Bonferroni method to a hierarchical approach to control the overall Type I error for the study