For comparing number of adverse events across doses, a repeated-measure generalized linear model was used to assess statistical significance. For comparing proportion of patients with at least 1 adverse event across doses, exact Pearson χ2 test was used to assess statistical significance.