TRIL acknowledges that it's possible the junction could be recognized as foreign, and that reliable immunogenicity data can only come from patients. They collected immunogenicity data during the initial primate studies, and will do so during the upcoming GLP tox studies. But the human drug is expected to be more immunogenic to mice and monkeys, so the usefulness of the data is questionable, and is likely one reason they don't intend to publish it.
They note that Fc fusion proteins aren't always as immunogenic as antibodies, and supply the example of Enbrel being considered less immunogenic than Humira.
Would a study of a mouse SIRPaFc in mice really convince you that the human drug would be equivalent?
Regards, RockRat