The amendment was enacted to: • Explicitly allow a home visit to take place instead of a site visit after a case-by-case approval by the sponsor. • Allow more flexibility in selecting the supplier of the rescue medication by allowing both morphine sulfate and morphine hydrochloride to be used, and by using an interactive response technology system (voice and web based) to assign subjects to rescue medication. • Allow the results of a non-protocol blood sample taken within 14 days of Visit 2 to be used for assessing the exclusion criteria and for baseline values without the need to repeat the sample at the Enrollment Visit. This was done to potentially reduce the volume of blood taken from the subject. • Clarify and correct the definitions of the endpoints, populations, and statistical analyses. • Clarify and correct the days on which the SOWS questionnaire is completed, and the questions to be assessed. • Add the measurement of height for the calculation of the glomerular clearance at 3 monthly intervals in the Tapentadol Period (Part 2). • Specify that if a subject turns 18 years old before Visit VE, an additional subject would be allocated to IMP. • Specify that subjects who vomit after IMP intake must not take additional IMP until the time of their next scheduled IMP dose. • Correct the labeling specification for the pharmacokinetic samples. • Add an additional exclusion criterion to exclude female subjects who are breast-feeding a child. Although there is no information on the excretion of tapentadol in human milk, there are pre-clinical data indicating that tapentadol is excreted in animal milk. Morphine is known to be excreted in breast milk, and may thus cause respiratory depression in the newborn.

This amendment introduced the following changes: • The alpha for the primary endpoint analysis was updated based on the methodology proposed by Hlavin et al. (2016). As a consequence of the alpha adjustment, the sample size was reduced. • A definition of long-term pain was added. • Specifications of the trial population were modified, requiring the inclusion of fewer subjects in the lower age group. The expectation that at least 15 subjects would be treated with tapentadol PR for a minimum of 12 weeks was added. • The estimated dates of last subject out for Part 1 and Part 2 were updated. • The main analysis was rescheduled such that it will be performed after all subjects complete the first 12 weeks of Part 2. • The interim analysis for the sample size re-assessment was removed due to the reduced sample size. The adaptive 2-stage design of the trial was simplified to a fixed 1-stage design. • Instructions pertaining to ECG-related discontinuation were added. • The timing of the Early Termination Visit was explicitly defined. • History of CRPS and history of a pain indication that is unlikely to respond to opioids was added to the exclusion criteria. • Safety experience from post-marketing data was updated. • Instructions for calculating the starting dose in Part 1 and Part 2 were revised. Hlavin G, Koenig F, Male C, Posch M, Bauer P. Evidence, eminence and extrapolation. Stat Med 2016; 35 (13): 2117-32.