urgent-how to improve low level spiked recovery?

[darasc0211]

I am validating a drug product impurity method. The spiked recovery was performed by spiking a known impurity at LOQ level (1%), specification limit (3%) and 5% to placebo which contains water, API and mannitol. API was also spiked along with the known impurity to demonstrated recovery of unknown impurity. (for ex: spiked both know imp and API at LOQ level to 40ug/ml mannitol in water). The recovery criteria is 85-115% but the spiked recovery for known impurity failed at LOQ level while API passed.
This is a lyophilized drug and suppose to be reconstituted with water before administered.

Why is the known impurity recovery low at low level? I can only think of two reasons: 1 impurity was binding with mannitol, 2: impurity was not completely dissolved. Either way I think I should look into sample preparation? maybe sonicatethe sample (the sample was not sonicated at the time of prep)?
I don't think I can loose the criteria any more.

Pleas advise on how do I improve the recovery? and what could be the reason.

Thanks.

[Don_Hilton]

Step 1) locate tthe loss. If you think the extraction might not be working well, extract several replicates of a sample and of a blank spike several extracts with the analyte and see if you recover the spike plus any analyte extracted from the sample. If the recovery of the spike is good, then yes it is the extraction step.

Step 2) having confirmed the location of the problem - if it is the extraction,
Common solutions to extraction problems include a change of solvent, application of temperature, and mixing. Other approaches can include degredation of matrix compunds, but must be used with care as the analyte may also be degraded or interferences created.

Since you are validating a method, I assume that it has already been developed and shown to provide adequate recovery with the conditions stated by the developer?

To give you a more specific answer to your question, more details would need to be provided.

[Don_Hilton]

And, one other thought - if this is a method that has already been developed, be sure that you use the same types of materials in contact with the samples. I have seen differences in methods when a change has been made from a plastic to a glass vial or vrom a borosilicate to a soda glass disposable pipette.