Final Protocol Version 1.0 dated 07 MAY 2016 was amended for the following reasons: • To amend the wording of an inclusion criterium to limit the exclusion of patient due to operator error of stopwatch. • To amend the stratification process at randomization, including 2 additional strata based on type of steroids because new data suggest that Deflazacort use is associated with more prolonged time to loss of ambulation and increased frequency of side effects. • To change the MRI cohort in MR cohort because the magnetic resonance spectroscopy (MRS) exam was added at the same time points of MRI test. The reason of the inclusion of MRS is explained in the next bullet point. • To change the Key secondary endpoints in the MR cohort related to imaging data using the data from MRS instead of MRI data by 3 point-Dixon technique. Both teniques are responsive to disease progression, but MRS is considered the gold standard for quantification of drug effects in vivo, and Fat Fraction measures from single-voxel spectroscopy have been shown to be associated with function in DMD. Moreover, MRS is the most direct and sensitive approach to discriminate water and fat proton component signals in MR. • To amend the wording of an exclusion criterium related to MR cohort in order to include also MRS exam. • To specify the type of Raven progressive matrices to be used for assessing the cognitive function. • To better clarify the pharmacokinetics procedures, by limiting the decision of site personnel to collect the PK samples • To allow personnel authorized by the Investigator to perform blood collection during visits 5, 6 and 8 in order to facilitate the management of these visits. • to correct some typographic mistakes existing in the final protocol version 1.0 (dated 07 MAY 2016).

Final Protocol Version 2.0 dated 13 JUL 2016 was amended for the following reasons: · To address relevant GNAs regarding to the VHP procedure VHP934 (VHP2016102); · To add the acceptability/palatability test as suggested by PDCO during the PIP discussion; · To better clarify the pharmacokinetics sampling and procedures; · To better clarify the time window of the Visits; · To update the functional algorithm after an additional discussion was performed with the statistician to include also 6MWT in the algorithm; · To specify the videos procedures which will be followed to ensure and maintain during the study, quality standard of functional evaluators; · To add “age” as covariate in the statistical analysis as discussed with the PDCO during the PIP discussion. . to correct some misleading or unclear sentences existing in the final protocol version 2.0 (dated 13 JUL 2016).

Final Protocol Version 3.0 dated 05 SEP 2016 (For European sites) and Final Protocol Version 3.1 dated 04 OCT 2016 were amended for the main following reasons: · To address relevant FDA comments regarding to the ‘May Proceed Letter’ (dated 18Oct2016); · To modify the exclusion criterion n.1 related to the exposure to another investigational drug (only for the European sites because for the US and Canada sites the criterion was already amended); · To modify the exclusion criterion n. 6 related to the ankle joint contractures to render it more applicable to the target population and to permit the recruitment of a wider number of children affected by Duchenne Muscular Dystrophy; · To better clarify the pharmacokinetics sampling and procedures; · To better clarify the time window of some assessments; · To add a blood sample for the evaluation of serum biomarkers related to DMD and to insert them as exploratory assessments and endpoints; · To add in the schedule of Assessment the possibility to perform the genetic test to confirm the diagnosis of DMD, as requested in the relevant inclusion criterion. . to correct some misleading or unclear sentences existing in the final protocol versions 3.0 and 3.1 (dated 05 SEP 2016 and 04 OCT 2016).

Final Protocol Version 4.0 dated 12 Jul 2017 is amended for the main following reasons: 1. To address safety issue. 2. To safeguard subject safety, a new exclusion criterion is added to the protocol excluding subjects with triglycerides > 300 mg/dL (3.42 mmol/L). 4. To increase the frequency of thyroid function monitoring (TSH, fT3 and fT4 assessments) to ensure patient safety. 5. To address the IDMC recommendation a complete blood count (CBC) test assessment has to be performed weekly for 8 consecutive weeks if the dose is reduced due to platelet count ≤150 x 109/L and/or white blood cell <3.0 x 109/L and/or hemoglobin <10.0 mg/dL for safety reasons. 6. To guarantee a right safety evaluation, spirometry should be repeated if the test does not meet ATS/ERS criteria and/or the evaluator deems that a valid attempt to perform a correct maneuver has not been made by the subject. 7. To update the Inclusion criterion #6 from “Have time to rise from floor of < 10 seconds at screening” to “Have time to rise from floor ≥ 3 seconds and < 10 seconds at screening”. The “time to rise” lower boundary is selected to avoid subjects who will not have enough function decline in 18 months if on placebo. 8. To update the Inclusion criterion #7 from “Have manual muscle testing (MMT) of quadriceps at screening ≥ Grade 3” to “Have manual muscle testing (MMT) of quadriceps at screening ≥ Grade – 3”, to permit recruiting of a wider no. of subjects. 9. To remove the Inclusion criterion (Criterion #8) 10. To update the Inclusion criterion #9 (#8 in the amended protocol). 11. 11. To update the Exclusion criterion #4

Final Protocol Version 6.0 dated 19 October 2018 is amended for the main following reasons: 1. In the original protocol sample size was calculated using publicly available standard deviation (SD), in particular from DMD Phase 3 study data on ataluren and drisapersen, where subjects have specific DMD genetic mutations (ie. Skippable 51 mutations or non-sense mutations). Since in the current protocol, inclusion of subjects is not depending on type of DMD genetic mutation, in case the Standard Deviation (SD) in the current study population will be lower than the estimated SD stated in the protocol, the sponsor believes appropriate to reduce the sample size of the study. Due to this, statistiscal analysis section is updated accordingly. 2. To update the background of givinostat regarding non clinical studies and clinical experience on risks and benefits section according to the new Investigator Brochure version 20 - July 2019. 3. To correct some typographic mistakes and/or clarify some paragraphs of Protocol version 6.0 (dated 19 Oct 2018) that could be unclear.

1. To amend the protocol based on the results of the pre-planned futility interim analysis as described in protocol version 7.0, performed in January 2020: it was concluded that futility on VLFF was not met and the trial should continue. 2. To amend the effect size, in 4SC 18 month change from baseline, from 3 seconds to 2 seconds. The Sponsor considers more appropriate 2 seconds based on the recent analyses performed by C-Tap (Wong B, et al. Action Duchenne International Conference, November 9–11, 2018, Birmingham, UK), where the minimally clinically important difference (MCID) for timed function tests (TFTs) in DMD was estimated at ~1.0 to 1.6 seconds. 3. Update the no. of subjects to be included in the MR cohort following the blinded sample size re-estimation done during the interim analysis in January 2020. 4. To cancel the second interim analysis following the recommendation made by FDA during the Type C meeting in October 2019. Since the second interim analysis has been deleted, the imaging endpoint has been moved to Key efficacy Endpoint. 5. To include an additional analytic method for the NSAA assessment. 6. To update the “Background on DMD” paragraph following the recent NDA approval received by Sarepta for the treatment of DMD boys. 7. To correct some typographic mistakes and/or clarify some paragraphs of Protocol version 7.0 that could be unclear.