Drug release profile in core–shell nanofibrous structures: A study on Peppas equation and artificial neural network modeling

Abstract

Release profile of drug constituent encapsulated in electrospun core–shell nanofibrous mats was modeled by Peppas equation and artificial neural network. Core–shell fibers were fabricated by co-axial electrospinning process using tetracycline hydrochloride (TCH) as the core and poly(l-lactide-co-glycolide) (PLGA) or polycaprolactone (PCL) as the shell materials. The density and hydrophilicity of the shell polymers, feed rates and concentrations of core and shell phases, the contribution of TCH in core material and electrical field were the parameters fed to the perceptron network to predict Peppas constants in order to derive release pattern. This study demonstrated the viability of the prediction tool in determining drug release profile of electrospun core–shell nanofibrous scaffolds.

Introduction

Nanofibrous webs produced by electrospinning are gaining industrial and scientific attention due to their unique features such as long lengths, small diameters, high surface area and porous structures which make them good choices for a wide variety of applications such as tissue scaffolds, drug delivery, composite reinforcement, chemical sensing, filtration, protective clothing, catalysis, solar cells, and electronic devices [1], [2], [3], [4], [5]. Recent studies have also shown the usage of polymeric nanofibers as drug delivery systems, whereby the drug component such as antibiotics (gentamycin sulfate [6], rifampin [7]), anti-cancers (paclitaxel [8], titanocene dichloride [9]) and also bioactive agents such as DNA [10], enzyme (lactate dehydrogenase [11], [12]), proteins (bovine serum albumin [13], [14], [15], [16], or growth factors [17], [18], [19] were incorporated as the core phase by co-axial electrospinning process. Typically, the drug release behavior was studied in these reports by UV–vis measurements which are subject to several external influences and cannot offer a priori prediction of the release profiles.

In this context, developments of models allowing prediction of drug release behavior represent a crucial challenge in the improvement of advanced drug release systems [20]. Drug release manner largely depends on how well, the drug is encapsulated [8] and in turn, in the case of electrospun fibers, is determined by material and electrospinning parameters. Although, many attempts have been made to study the diffusion of drug molecules through the polymeric solid membranes in order to develop models for their release behaviors [21], [22], [23], [24], formulating and modeling of drug release from electrospun fibers is a new concept. For empirical/semi-empirical mathematical modeling of drug particles from polymeric layer, the Peppas equation is generally favorable [20], which is based on the Fickian diffusive release from a thin polymeric film [25].

$$\frac{Mt}{M\infty }=k{t}^n$$

where Mt is the amount of released drug at time t, M∞ is the total amount of encapsulated drug, k is a system specifics constant and n is the diffusion exponent [25] that corresponds to the transport mechanism. The nanofibrous structures resemble porous networks whose solid parts comprise many intertwined cylindrical fibers containing encapsulated drugs which require a new model to describe drug release profile through fibrous scaffolds. However, Peppas equation used to analyze controlled release of water soluble drugs from the polymers [13] and can be applied to study release profile of hydrophilic drugs like TCH used in this work.

On the other hand, artificial neural network (ANN) models have been widely used in the textile industry and there are numerous publications on neural network applications addressing wide variety of textile problems such as the relationship of the electrospinning parameters with the diameter of electrospun nanofibers or analyzing the dependence of the fiber diameter on the process parameters [26], [27], [28], [29], [30]. Application of ANNs in the design and development of controlled-release systems of oral dosage forms have also been examined to predict or optimize different types of controlled release formulations [31], [32], [33], [34], [35], [36], [37], whereby, the majority of these applications have focused on oral controlled release drug delivery systems.

Electrospinning process involves numerous inputs and possible output parameters showing complex interdependence that poses a challenge on the development of an exact mathematical model to simulate the whole process. Moreover, known and unknown variables cannot be interpolated and extrapolated in a reasonable way due to lack of data on calculation of experimental observations with weighted contribution of the corresponding variables.

Herein, we proceed a combination of Peppas equation and artificial neural network modeling to predict the release profile of TCH from electrospun core–shell nanofiber mats. The study introduces (i) drug release mechanism in electrospun core–shell polymeric fiber mats and influence of diffusion, (ii) application of Peppas equation to formulate the delivery system in order to acquire Peppas constants (k and n), (iii) employing ANN predicting model to estimate the Peppas factors.