Pre-analytic optimization of API assays in compounded oral suspensions for pediatrics

Introduction
Quantitative quality controls of Active Pharmaceutical Ingredients (APIs) in compounded oral suspensions for pediatrics (COSPs) are assessed in our hospital pharmacy by UV/Raman spectrophotometry. The used vehicle in these COSPs has heterogeneous and viscous characteristics that make these determinations not easy to perform. In fact, these assays require accurate and repeatable aliquoting, followed by extraction of the aqueous fraction which should be clear and transparent to be analyzed by photometry. In order to optimize these assays, we propose to compare several sampling and extraction modalities and to standardize these pre-analytical steps.

Materials & methods
3 COSPs were prepared using Syrspend^® SF PH4 vehicle (Fagron; Netherland; density = 1.01): amlodipine 1 mg/mL, propranolol 1 mg/mL and thiamine 100 mg/mL.
In order to obtain analyzable aqueous phase, 4 methods were compared: filtration through a 0.45µm hydrophilic filter (Millipore), acidic hydrolysis (5% HCl 37N); centrifugation of undiluted, 1/5^th and 1/10^th water diluted COSPs samples with increasing times and speeds centrifugation.
For aliquoting standardization, samples were taken according to 3 modalities (n=5/modality): samples weighing, direct pipetting (without cone wetting) and reverse pipetting (with cone wetting).
In order to analyze the sampled and extracted fractions, the contents of the aqueous phases were analyzed by UV/Raman spectrophotometer QC Prep^®. The measured contents (m±SD, n = 5) were compared between modalities by ANOVA test (p<5%).

Results and Discussion
Samples clarification by filtration was non-contributory: filter clogging, very low extraction yield (10-12% of the filtered volume) and a cloudy and opaque extract. The acid hydrolysis led to the suspension destruction but with a milky result that could not be read by photometry. Pure COSPs samples centrifugation (1000 to 5500 rpm; 1 to 10 min) did not break up the suspension. 1:5 diluted COSPs samples centrifugation did not result in a satisfactory result either. Only the 1/10^th dilution at 5500 rpm/10min resulted in a usable transparent aqueous extract separated from a solid and stable pellet.
When applied to the samples taken according to the 3 modalities mentioned above, this protocol showed that aliquoting by weighing and reverse pipetting gave the results closest to target concentrations, with the least dispersion and with respective average recoveries of 104.4 ± 4.9% and 103.1 ± 4.6%. The direct pipetting technique gave unsatisfactory results, particularly with thiamine, where the mean recovery of 88.1±2.7% was lower than with the other techniques.

Conclusion
Our study shows that APIs concentration determinations of COSPs requires sampling by weighing or reverse pipetting. The aqueous phase extraction must be performed on a 1/10^th diluted samples submitted at least to 10 min centrifugation at 5500 rpm. In such conditions, the obtained analytical results are satisfactory, accurate and precise.

See the poster