FDA on Thursday offered new draft guidance on statistical approaches recommended for evaluating analytical similarity for biosimilars and functional attributes of the reference product. The guidance includes details on analytical similarity assessment plans, the development of the risk ranking of attributes, the determination of the statistical methods to be used, the statistical analysis plan, and the statistical methods for evaluation. FDA recommends "a minimum of 10 reference product lots be sampled, and the lots "should represent the variability of the reference product." To allow for meaningful comparisons, FDA also recommends a minimum of 10 biosimilar lots to be included in the analytical similarity assessment. FDA recommends the final analytical similarity report should contain differences in age of the lots produced at testing; multiple testing results; assay performance; and differences in attributes that will be considered acceptable. Comments on the draft can be made over the next 2 months.