AX-R12 — Unresolved shoulder or excessive peak asymmetry

A purity figure is only as good as the integration that produced it, and an integration is only reliable when the main peak is resolved from its neighbours and roughly symmetric.

The rule applies two criteria. Peak asymmetry above 1.50 — measured at 10 % of peak height, per Ph. Eur. 2.2.46 — means the trailing edge is dragging and the boundary between the main peak and anything eluting after it is a matter of judgement. Resolution below 1.20 to the nearest named impurity means the two peaks share a valley high enough that drop-line and valley-to-valley integration will give materially different answers.

Either condition makes the reported purity a range rather than a number. The index does not discard these records — the material may be perfectly good — but it marks the purity figure as integration-limited and excludes it from the tightest aggregates.

Tirzepatide accounts for a disproportionate share of instances, because the des-Aib2 impurity eluting 0.35 min early frequently appears as an unresolved leading shoulder rather than a separate peak. Retatrutide and the other multi-Aib analogues follow for the same reason. This is the rule that most directly encodes the index's view that method and result are inseparable.