This is a summary of the the framework for impurities per ICH Q3:
Reporting threshold: typically 0.05–0.1% for peptides (lower than small molecules’ 0.05% at low dose, but this varies by dose band)
Identification threshold: where you must structurally characterize the impurity, typically in the 0.1–0.5% range depending on max daily dose
Qualification threshold: where you need toxicological qualification (safety data, not just identity), often 0.5–1.0% for lower-dose peptides, tighter for high-dose/chronic products
Individual specified impurities are usually held to tighter individual limits (often 0.5-1.0% each) while total impurities get a looser umbrella limit (often 2-5%, sometimes higher depending on the peptide’s synthetic complexity. Longer peptides accumulate more truncation/deletion species and can justify higher total impurity allowances if each is qualified).
Who decided >99.5% is the specification for purity?
Bold, considering no one knows:
1. The entire process impurity / degradation profile for any of the peptides. Really an impossible ask since they are all coming from different synthesis processes with their own unique impurities.
2. That any single HPLC method is capturing all of said impurities and degradation products today.
Without those two things, setting a spec is shooting in the dark based on incomplete data.
“Well this lab always gets 99.9%, so it’s possible.”
It’s also possible that their method is trash and sees nothing, and so you’ve set a spec when, if analyzed per 1 & 2 above, 100% of the market will fail.
It’s an unrealistic spec that harms researchers by limiting access to affordable peptides and drives testing to labs with the worst methods.