Pharmaceutical Development
The situation
A developer preparing a low-dose oral anaesthetic chewing gum came to us with what looked like a complete technical package: a prototype, an analytical release report from a specialist laboratory, a provisional patent filing and a manufacturer brief. They were ready to select a contract manufacturer and move to pilot.
The report showed approximately 79% active release at a thirty-minute endpoint. On the face of it, a working product.
What we found
Reconstructing the calculation from the raw laboratory values showed it had been built around a nominal dose twice the commercial target. The recovered amount was internally consistent with the report’s own reference figure, but that reference figure did not match the product the client intended to sell. There was no release dataset in the package that could be securely attributed to the commercial strength.
Three further gaps surfaced in the same review:
No reproducible commercial formula existed. The package contained a technical handover and a published study example, but not a batch record.
The label basis for the active was undefined. Whether the claim referred to the base, the anhydrous salt or the monohydrate changes dispensing, assay and label text, and none of the documents settled it.
The intended United States route assumed monograph eligibility. The active is not listed in the relevant FDA oral healthcare monograph, so that route was not available as written.
What we did
We rebuilt the formula as a new controlled development baseline with potency correction shown explicitly, rather than trying to recover a batch record that did not exist. We specified a diagnostic assay on retained pieces to test the dose hypothesis before any further spend.
We wrote a manufacturer brief with mandatory capability gates and explicit no-go conditions, so that food-grade and supplement manufacturers would screen themselves out. And we set out a stage-gated programme with budget allowances at each gate rather than a single figure.
The outcome
The client stopped short of committing to a manufacturer on the strength of a number that did not describe their product. The next spend became a bounded feasibility programme with defined closure criteria, instead of a pilot built on an unverified premise.
A single review cycle is usually enough to establish whether the data in hand describes the product you intend to sell.