M. Chacon et al., OPTIMIZED PREPARATION OF POLY D,L (LACTIC-GLYCOLIC) MICROSPHERES AND NANOPARTICLES FOR ORAL-ADMINISTRATION, International journal of pharmaceutics, 141(1-2), 1996, pp. 81-91
A rotatable central composite design (RCCD) was applied to optimize th
e preparation of cyclosporine-loaded poly D,L (lactide-glycolide) (PLA
GA) nanoparticles (NP) and microspheres (MS) by solvent displacement a
nd solvent evaporation techniques, respectively. The joint influence o
f needle gauge, polymer amount and the injection rates on the mean par
ticle size, relative standard deviation (RSD), yield and drug encapsul
ation percentage in NP were evaluated. With regards to MS, the polymer
amount and the stirring rate were evaluated. Scan electron microscopy
of MS and NP showed spherical particles with a dense polymeric networ
k in the first case. From the statistical analysis of data polynomial
equations were generated. The mean particle size ranged from 50 to 150
nm for NP and from 1.5 to 30 mu m for MS. Smallest nanoparticles (46
nm) were obtained by using the lowest polymer amounts, the highest inj
ection rates and the lowest needle gauges (r(2) = 0.9443). Under these
conditions the drug entrapment percentage was maximum (85.2%), sugges
ting the drug might be entrapped and adsorbed on the nanoparticle surf
ace. The relative standard deviation was only affected by the polymer
amount (r(2) = 0.8034) and the yield rose with the amount of PLAGA (r(
2) = 0.9016). A very important increase in particle size (r(2) = 0.985
5), relative standard deviation (r(2) = 0.9353) and encapsulation perc
entage (r(2) = 0.9669) were observed for MS by decreasing emulsificati
on stirring rates and increasing polymer amounts, the stirring rate be
ing the most significant independent variable (alpha < 0.0001) in all
cases. The highest experimental encapsulation value (97.69 +/- 0.78%)
correspond to samples prepared from 150 mg of polymer and a global sti
rring rate of 2000 rpm. By using response surface diagrams and the mat
hematical models proposed, it is possible to easily deduce experimenta
l conditions to prepare NP and MS with the desired properties.