In Section 8.3.2 ‘Secondary Efficacy Variable(s)’ a bullet point was added between bullet 1 and 3 that the percentage of responders (defined as a ≥30% reduction from baseline) in averaged 24 hour pain scores collected for 7 days preceding the study clinic visit at Week 16 was also tested in a hierarchical testing strategy. In Section 14.2 ‘Statistical Considerations’, the percentage of responders in averaged 24 hour pain scores at Week 16 was added to the key secondary endpoints which were tested for confirmatory statistical significance under a hierarchical testing strategy. In Section 14.7 ‘Secondary Outcome/Efficacy Variables, it was added that in the case the analyses described before in this section yielded statistically significant differences at each time point, the percentage of responders in averaged 24 hour pain scores at Week 16 were analysed in a formal, confirmatory manner to test for superiority. Logistic regression analysis was used to determine the odds ratio and 95% confidence interval of OXN PR relative to placebo. The statistical model included terms for treatment and centre as factors, and baseline averaged pain score as a covariate. LOCF was used for this analysis. A supportive analysis was performed on the PP Population. A sensitivity analysis on the FAP was also applied to the analysis of the key secondary endpoint of the percentage of responders in averaged 24 hour pain scores, implementing the same logistic regression model, using OC data at the 16-week time point. The analysis of the key secondary endpoint, CGI-I (as assessed by the Investigator), were conducted in a formal manner in the event that the hierarchical testing structure yielded statistically significant differences at each time point. A statistically significant difference in the percentage of responders in averaged 24 hour pain scores at Week 16 was added as a criterion.