This study describes the development of a new esomeprazole (ESO) delayed release gastro-resistant formulation with improved storage stability. A three-step (drug-, sub(seal)- and enteric-) coating process was employed with the aid of a fluid bed coater. Several formulation factors (namely, size and quantity of starting non-pareil sugar spheres, binder quantity during drug-layering, sub(seal)-coating polymer type, and quantity and enteric coating quantity) were evaluated and the whole process was modeled with the aid of feed-forward back-propagation artificial neural networks (ANNs). Results showed that the selection of small-sized starting spheres (45/60 mesh size) leads to pellet agglomeration, while as sub(seal)-coating weight gain increases a reduction in ESO dissolution rate is observed. The enteric-coating applied (Eudragit L30D-55) showed good gastro-resistant performance in both 0.1 N HCl and pH 4.5 media, while immediate release profiles with more than 85% of ESO being released in less than 30 min were obtained. The effect of cellulose-based sub(seal)-coating polymers, (namely, hydroxypropyl cellulose and hydroxypropylmethyl cellulose) on formulation's storage stability at 40 ± 2 °C/75 ± 5%RH indicated that only hydroxypropylmethyl cellulose was able to stabilize ESO delayed-release formulations in terms of assay, dissolution, impurities, and gastro-resistance performance. Finally, scanning electron microscopy (SEM) analysis revealed smooth and homogeneous external surface/coating layers in all three levels (drug-, sub(seal)-, and enteric- coating), while x-ray diffraction showed no polymorphic transformations.
Keywords: Esomeprazole; artificial neural networks; delayed-release; gastro-resistance; storage stability; sub(seal)-coating.