Unfortunately, the original version of this article [1] contained an error due to a typographical error in the computer code used. This meant the percentage improvement over baseline of the endpoints CMTNS, ONLS, 9HPT and DML was slightly increased. Therefore some of the values in Tables 3, 4, Additional file 4: Table S4 and Fig. 4 are incorrect.

Table 3 Response to PXT3003 on efficacy outcomes in treatment groups, with comparisons of active doses versus Placebo (Full Analysis Set, n = 80)
Table 4 Response to PXT3003 on efficacy outcomes in HD and in PLI, with comparisons of HD versus PLI (Full Analysis Set, n = 80)
Fig. 4
figure 1

Response to PXT3003 on clinical scales (Full Analysis Set, n = 80). Mean % (s.e.m.) of improvement from baseline per group at 12 months for CMTNS (a) and ONLS (b). Sample sizes: Placebo (n = 19), LD (n = 21), ID (n = 21), HD (n = 19)

However this error has no bearing on statistical outcome of the study, the conclusions or text of the manuscript. Correct versions of Tables 3, 4 and Fig. 4 can be seen below and Additional file 4 accessed using the link below.