Lactose adduct formation?

[GChad]

We are seeing low assays for a caspule formulation containing lactose. Assays are fine for API itself and API in capsule, but API + excipients (mainly lactose) in capsule are ~10% low. Diluent is 5% methanol in pH 3 buffer, working well below the solubility of the drug. We're not seeing any additional peaks in the chromatogram that would account for the assay loss. The method is HPLC reverse phase using an ion-pairing agent.

I'm reading about Maillard reactions between amines and reducing sugars and wondering if lactose adduct formation could be the culprit here, as the API contains heterocyclic amine groups. My questions are:

- is this a likely mechanism for the low assays in this case?

- if adducts are forming, are they more likely forming in solid state in the capsule, or more likely forming in solution in the flask, i.e. an analytical artefact?

- if forming in the flask, can it be prevented, and what's the best way?

Thanks

[Don_Hilton]

If the nitrogen of the aromatic system is in the ring, Maillard reactions would not be my first guess.

Are you sure the drug was present at the correct levels in the capsules to start with?

If you are concerned about analytical artefacts resulting because of various components being in solution, try mixing the various components together under conditions where you can verify the quantity of drug going into the mix and measure what is left.

Without knowing the various chemicals present, it is difficult to speculate on the potential of them to react with the others - or how to avoid the reactions.

[JMB]

If your N is present as RNH2 or ArNH2, then the Maillard condensation with lactose is almost certain to occur INSIDE the capsule; I agree with Don about no Maillard from lactose + N in a ring.

My feeling is that the 10% loss is in the ballpark for this condensation, and will be increased by forced stressing studies.

I'm concerned that you don't see an additional peak in the chromatographic profile at the ~10% level; one explanation is that a presumed lactose/API adduct is very early-eluting (Void) or very late-eluting (low and broad), while a second suggests that perhaps your API is being degraded and the chromophore is destroyed.

Can you (partially) disclose API structure ??