PurposeOsteoarthritis Research Society International (OARSI) Expert Consensus Guidelines recommend topical non-steroidal anti-inflammatory drugs as first-line medications for osteoarthritis (OA) knee pain, but several voluminous daily applications are required to achieve efficacy. There is a need to develop new and improved topical analgesics with a faster onset, longer duration of action, and the requirement to apply less gel. This trial investigated the safety and efficacy of a new 3.06% diclofenac gel (AMZ001) in subjects with knee OA. MethodsIn total, 444 subjects (AMZ001 twice daily (BID) [n = 121], AMZ001 once daily (QD) + placebo QD [n = 121], placebo BID [n = 121], or Voltaren 1% 4-times daily [n = 81]) were enrolled. All except Voltaren 1% (single-blinded) were applied topically in a double-blind manner for a total of 4-weeks. The primary endpoint was the change from baseline to week 4 in the WOMAC pain sub-score in the target knee. Secondary and exploratory endpoints included additional efficacy measures (WOMAC total score, WOMAC function and stiffness sub-scores, WOMAC pain weight-bearing and non-weight-bearing sub-scores, ICOAP, chair-stand test, OMERACT-OARSI responder rate, PGA, WPAI, EQ-5D, rescue medication use, satisfaction questionnaire) and safety. ResultsTreatment with AMZ001 QD was effective at reducing WOMAC pain sub-scores vs placebo (estimated treatment difference [ETD]: −4.61 [95% confidence interval (CI): −9.09, −0.12]; p = 0.0440); however, BID application was not (ETD: −3.76 [95% CI: −8.21, 0.68]; p = 0.0969). For several secondary endpoints, changes from baseline to week 4 conferred nominally statistically significant improvements in favor of AMZ001 vs placebo, including PGA score (AMZ001 BID vs placebo, ETD: −0.61 [95% CI: −1.11, −0.11]; p = 0.0162; AMZ001 QD vs placebo, ETD: −0.63 [95% CI: −1.13, −0.13]; p = 0.0134), WPAI overall work impairment score (AMZ001 QD vs placebo, ETD: −10.44 [95% CI: −20.84, −0.04]; p = 0.0492), and EQ-5D VAS score (AMZ001 BID vs placebo, ETD: 4.70 [95% CI: 0.55, 8.85]; p = 0.0264). Post-hoc analysis excluding 11–14 subjects per group with pain scores that decreased between screening and baseline suggests a consistent effect of both AMZ001 QD (ETD: −5.84 [95% CI: −10.71, −0.97]; p = 0.0189) and BID (ETD: −5.35 [95% CI: −10.16, −0.54]; p = 0.0292) in reducing WOMAC pain sub-scores vs placebo. In general, treatment satisfaction was high, as measured by the satisfaction questionnaire. The frequency and incidence of adverse events (AEs) was greatest in the placebo group. Most AEs (>99%) were of mild or moderate severity. There were no serious AEs. There were no notable effects of any treatment on vital signs, ECGs, physical examination findings, or other laboratory assessments. ConclusionsTreatment with AMZ001 BID for 4 weeks improved WOMAC pain sub-scores; however, only QD application conferred nominally statistically significant improvements vs placebo. AMZ001 was generally well tolerated.