Facet Biotech - Comment

Line 204: If the negative control is close to the cut point as suggested by the wording “the value obtained for the negative control should closely reflect the cut point…”, then the plate failure rate might be too high (negative control would often screen positive). Can you consider wording such as “the value obtained for the negative control should generally be near but consistently lower than the cut point…” Agree with the statement that negative controls that yield values far below that of the cut point may not be useful…. Line 439: Could you clarify what is meant by “performing several runs of negative samples…” Would it be acceptable to run the 50-100 samples one time each, but over 3 days by several analysts. Such an approach would reasonably characterize variability. Line 461: suggest removing last part of sentence “of the sample for NAB”. Rationale: on page 3, lines 112-121, the multi-tiered approach for immunogenicity assays is described whereby all samples are screened, screen-positive samples are confirmed, and then confirmed positive samples are tested for NAB. On line 461, the statement implies screen positive samples would be tested for NAB, which is one approach to take, but not consistent with p. 3. Removal of the last part of that sentence removes any inconsistencies and leaves it more general as far as further analysis. Page 13: Robustness and Sample Stability section, long-term stability is not mentioned. If FDA agrees that such analysis is not needed, can it be stated that such testing is not expected unless specific circumstances warrant (e.g. unusual matrix). Line 543: Precision: the degree of detail in this section is higher than in any other paragraph, e.g. with regard to replicates. Would it be acceptable for the inter-assay assessment to run duplicates in each assay over 3 day with a minimum of 6 plates?