Immunogenicity risk assessment and in silico tools
Multiple factors can lead to immunogenicity. These include intrinsic
factors like the sequence and structure of a given protein and the post
translational modifications that can occur during the process (Jiskoot,
Rispens, & Kijanka, 2019). In addition, risk factors also include those
from unwanted impurities that can carryover from purification during
drug manufacturing, liabilities due to formulations and excipients and
devices as well as how the drug gets administered (Jiskoot et al.,
2019). To understand the sequence-based risk of residual proteins
associated with the API, multiple algorithm-based tools were used. The
EpiMatrix algorithm evaluate the ability of processed peptides (9-mer
sequences) in a protein to bind with the 8 most prevalent MHC alleles
that represent over 98% of the human population (Bailey-Kellogg et al.;
Goey, Bell, & Kontoravdi, 2018; Jawa et al.). Multiple high-density
T-cell epitopes or clusters were assessed using the ClustiMer tool.
Additionally, the epitopes that could be processed and presented were
further analyzed in the JanusMatrix to determine which predicted
epitopes may be cross-react with epitopes derived from the human genome
on the basis of conservation of T cell receptor (TCR)-facing residues.
ClustiMers with JanusMatrix scores greater than 1 were excluded from the
analysis based on the assumption that the auto-reactive TCR containing
cells were eliminated during T cell development (Bailey-Kellogg et al.).