Bioequivalence and Bioavailability Forum

Dear colleagues, we are conducting a pivotal fed-fasted BE study versus a reference listed drug (RLD). In this case, by "pivotal", I mean the study will serve as the sole basis for (US) NDA approval. My CRO states that no "standalone" Statistical Analysis Plan (SAP) is required and that the statistical analysis in the protocol should suffice. However, the information in the protocol does not capture everything I would ordinarily expect to be in the SAP and I am not comfortable with the CRO's response. Is a (detailed) SAP required (or recommended), particularly where the entire NDA will be predicated on its findings?