Following regulatory agency input, the following changes were made: Change to single dose level (25 mg), previously dose escalation up to 50 mg was included. Inclusion criteria revised to require unsatisfactory pain control from, or intolerance of, at least 2 prior standard osteoarthritis (OA) treatments. Exclusion of participants considered by the investigator to be non-responsive to corticosteroids. Increase in sample size from 76 to 95 per arm. Requirement for ultrasound guidance of the injection (previously optional).

Change of study location from the United States based to Denmark. Added recording of rescue pain medication use. Requirement added for typical knee pain at Screening to be ≥4.0 out of 10. Requirement for average daily index knee numeric pain rating scale (NPRS) score in the second week of the Washout and Baseline Period changed from ≥5·0 / ≤9·0 to ≥4·0 / ≤10·0. Requirement for non-index knee pain scores at Screening changed from a single NPRS score ≤4 to both weekly Western Ontario and McMaster Universities Arthritis Index (WOMAC) Pain scores and the average daily NPRS scores in the second week of Washout and Baseline Period ≤4.0. Exclusion of chondrocalcinosis relaxed to only exclude chondrocalcinosis likely to affect the outcome of this study in the opinion of the investigator. Removal synovial fluid collection for pharmacokinetic (PK) and biomarker analyses. Added precautions in the event of symptoms and signs of COVID-19.

Changes to align protocol language with that appearing on the electronic Patient Reported Outcomes (ePRO) device, in other study documents, and in typical site procedures. Requirement updated for the injection to be performed by suitably qualified and experienced physician, previously this may have been a suitably qualified and experienced physician’s assistant/nurse practitioner.

Number of sites and countries increased. Removed requirements for average daily index knee NPRS score in the second week of the Washout and Baseline Period ≥4.0 and ≤10.0 and individual daily index knee NPRS scores to not vary by more than 4 points. Requirement for non-index knee pain scores at screening changed from both weekly WOMAC pain scores and the average daily NPRS scores in the second week of Washout and Baseline Period ≤4.0, to only WOMAC pain scores at the end each week both ≤6.0. Morning serum cortisol no longer required at Screening, instead the adrenocorticotropic hormone (ACTH) stimulation test was to be performed in all participants during Screening. Criteria added for selection index knee, covering the event both knees were potentially eligible.

Requirement for weekly WOMAC pain subscale scores for the index knee (end of each week of the Washout and Baseline Period) from both ≥5.0 and ≤9.0 to both ≥4.0 and ≤9.0 (and not vary by >3 points).

Allowed inclusion of well-controlled, non-insulin dependent diabetics (i.e., HbA1c levels ≤7·9% (63 mmol/mol). Previously, patients with diabetes mellitus defined as a haemoglobin A1c (HbA1c) value of ≥6·5% (47 mmol/mol) were excluded. Allowance for re-starting the Washout and Baseline Period (within the 8-week screening period) if baseline WOMAC pain scores were not completed due to reasons beyond the participant’s control eg, technical issues with the ePRO device. Statistical methods revised such that they were to be described in detail in the Statistical Analysis Plan (SAP) which will be finalised prior to unblinding of the study.

Addition of an optional imaging sub-study extending the maximum duration per participant to approximately 60 weeks with additional magnetic resonance imaging (MRI) site visits.

Addition of an optional extended PK sub-study to detect the amount of corticosteroid FP in subjects following Week 24/End of study visit.