Amendment 1: The endpoint ‘number of patients with injection site pain’ was removed from other endpoints as this was a safety endpoint. The wording ‘EU-sourced Humira’ was deleted from the description of the single dose PK study in cynomolgus monkeys as this study was not using EU-sourced Humira. Screening period was revised to 46 days and a pre randomization phase of up to ten days was added. It was clarified that the process for randomization and pre-randomization was dependent on site and country. It was clarified that patients were to be randomly assigned to trial drug according to both the randomization ratio and the stratification factors and that the same process was to be used for both the initial randomization and for re randomization. The requirement for patients with a history of hypersensitivity to adalimumab to be excluded from the trial was removed because patients who had received Humira before were not eligible, regardless of whether they had developed an immune reaction. It was clarified that the PPD or IGRA test could be performed at Week 48 for TB assessment; samples for the TB test was updated from 1 to 2 and blood volumes were revised accordingly. Only unblinded personnel were responsible for ensuring drug accountability and compliance; a unique drug number was to be assigned for each drug administration; syringes were not pre prepared but were provided to the sites pre filled; data for primary analysis were to be unblinded only for the team members involved in the primary analysis as well as PPK analysis. New template text was added for the definitions and reporting of cancers and protocol specified AESIs and reporting of AE/SAEs. In the statistical methods, text stating ‘if a patient misses an entire visit the missing data will not be imputed’ was deleted and updated to include ‘Baseline’ to clarify the process for handling of missing data. The power was updated to correct the calculations for DAS28 to reflect the power for 2 sided tests.

Amendment 2: The primary objective updated to replace the co-primary endpoint of ‘change in DAS28-ESR at 24 weeks compared to baseline’ with ‘proportion of patients meeting ACR20 response rate at Week 24’. Primary, secondary and other endpoints were updated based on the changes to the objectives. Sample size was increased to 650 patients to take into account the updated margins resulting from the change in primary endpoints. Safety endpoints were updated to state ‘the proportion’ instead of ‘number and proportion’, as only 1 item should be measured in an endpoint. The endpoint relating to injection site pain was deleted from the safety endpoints; this assessment was not deemed to provide relevant information for the trial as all injection site reactions were to be captured as AEs. A self administration arm was added to collect patient data on real life experience of self injection; the secondary and exploratory objectives were updated accordingly with a reference to this new arm. Injection-related endpoints were added to the ‘other’ endpoints category to allow assessment of self administration. Patients who discontinued the trial early were to be followed for efficacy until Week 48. Only qualified patients were to be offered participation in the OLE trial, not all patients. AEs were to be reported until the last per protocol visit. Patients who discontinued early were required to be followed up until Week 58. The hypotheses for the equivalence test, and the associated statistical margins, were updated to reflect the revised primary objective and endpoints. Secondary analysis was updated to include the new endpoints of the mean changes from baseline in DAS28-ESR after 12 weeks and was based on the FAS and used the LOCF method for missing values. Data reporting was amended to clarify that all efficacy data were to be reported at 24 weeks in a CTR and that a follow up analysis was to be conducted when all data were available and would be reported in follow up CTR.

Amendment 3: It was implemented prior to IRB approval as the changes involved only logistical or administrative aspects. This amendment was implemented after the start of patient enrollment. Based on FDA feedback, the following changes were made throughout the CTP: The term US-sourced Humira was replaced by US-licensed Humira. References to Humira or adalimumab were updated to US-licensed Humira and/or EU approved Humira, as applicable. The term adalimumab was used for general mode of action descriptions of the molecule. The synopsis and trial flow chart were updated in line with the above mentioned changes, as appropriate.

Amendment 4: Based on feedback from the FDA the following changes were implemented: Removal of the self administration arm. This was based on FDA feedback that self administration data from the use of pre filled syringes was not required and could have potentially compromised the blinding of the trial. The process of trial drug administration was changed such that the trial personnel responsible for administration were to be blinded. It was clarified that every effort should be made to follow up patients who discontinued trial drug for efficacy. The description of the analysis of the co primary endpoint of ACR20 at Week 12 was updated to state the use of 90% CIs and an equivalence margin of ± 12%. Additional safety ECG monitoring was implemented at Weeks 4, 12, 24, 40, 48, and 58. The criteria for withdrawal from the trial, discontinuation from trial drug, and the requirements for follow up of patients who discontinued treatment were clarified. The definitions and processes of AE collection and reporting were clarified; pregnancy reporting requirements were expanded to include female partners of male patients; aminotransferase was added as a parameter for identifying hepatic injury.

Amendment 5: It was specified that the Week 48 efficacy endpoints would not be analyzed as part of the primary analysis of efficacy and safety in the current Week 32 CTR as not all Week 48 ACR20 data would be available at the data cut off for the primary analysis. The margins for the Week 12 primary efficacy assessment were updated from [-12%; 12%] to [-12%; 15%]. This slightly higher upper bound (+15%) allowed for variations in the imputation techniques and response rates used in the calculation of the margins and was agreed to by the FDA. The -12% lower bound was maintained in case the response rate for BI 695501 was lower than that for US-licensed Humira. The fixed categorical effect ‘region’ was removed from the primary efficacy endpoint models in order to prevent any potential issues due to sparse data from a region. The primary analysis cut off was changed from Week 24 to Week 32 to allow for an early evaluation of the initial 8 weeks of safety and immunogenicity data following transition for those patients who switched treatments at Week 24. Exclusion criterion 2 was updated to include ‘per Investigator discretion’ for consistency with the rest of the CTP. It was clarified that protocol deviations would be handled on a case by case basis; the term ‘severe’ in relation to protocol deviations was removed accordingly. It was clarified that the primary efficacy endpoint analysis would use two separate logistic regression models to address FDA feedback. The missing data imputation method for the primary and secondary analyses was changed from LOCF to MI to address FDA feedback.