Unused Indexes with verification across secondary replicas (via Linked Servers or Powershell)
Basically a version of this: https://github.com/MadeiraData/microsoft-dbas-club/blob/master/Indexes/Best%20Practice%20Checks/Unused%20indexes.sql
But done in combination with an identical check on a specified linked server which is assumed to be a secondary replica of the current server (or using a Powershell script that aggregates the check across a list of specified replica servers).