Reservoir Delivery System
In a reservoir system the drug is either molecular dissolved or dispersed in a reservoir that is surrounded by a rate limiting inert membrane. In a monolithic or matrix device this membrane is absent. Although controlled delivery from a monolithic device is very well possible, the presence of a membrane normally provides for a better release profile. In principle, a zero order release (constant release in time) can be approached with a reservoir system when the drug is dispersed in the core at a concentration that is far above the saturation solubility. The release of the drug occurs by diffusion of the drug molecules through the core and membrane of the reservoir system. Because of their suitable properties polymers are often used in reservoir systems. The diffusion process of drug molecules in through the membrane is described by Fick’s first law of diffusion:
Where:
J=flux (kgm-2s-1)
dC/dx=concentration difference (kgm-4)
D=Diffusion constant (m2s-1)
In case the drug is dispersed in the core at a concentration that exceeds the saturation solubility and the concentration of drug outside the reservoir system is almost zero (sink conditions) concentration difference (∆C) over the membrane can be regarded as almost constant. For a planar reservoir system the following equation can then be deduced from Fick’s law:
Where A is the area of the membrane and d is the membrane thickness.
K is the interfacial partitioning of the drug between the core and the membrane and is related to the solubility’s in the core (Cc) and in the membrane polymer (Cm) as defined by:
In order to predict the release rate for cylindrical and spherical geometry’s the following equations are described by Baker and Lonsdale:
 |
 |
| a) cylindrical geometry |
b) spherical geometry |
Where:
h= length of the cylinder
ro= outer radius membrane
ri= inner radius membrane
The next figure shows the diffusion process of a drug through the membrane. As a result of the concentration difference over the membrane the drug that is dissolved in the core reservoir diffuses through the membrane to the surrounding medium at a controllable rate and time. In the active layer the permeability should be preferably much higher than the permeability of the membrane polymer. Once immersed in a dissolution medium a thin hydrodynamic layer is present at the boundary of the reservoir system. If the thickness and permeability of this hydrodynamic layer surrounding the membrane is smaller than those of the membrane, the membrane will be rate limiting. Here the permeability is defined as the product of the diffusion coefficient and saturation solubility of the drug (P=D*Cs).
Diffusion process over the membrane
A suitable technology that is often used to manufacture reservoir systems is coextrusion technology. This technology is commercially is used to manufacture multilayer flat films (transdermal delivery) or multilayer coaxial fibers (implants, vaginal rings etc.).
 |
 |
| two layer transdermal patch | three layer transdermal patch |
The advantage of multilayer reservoir system is that they can be designed to deliver one or more drugs at the same time. However multilayer reservoir systems are more complex to design because both physicochemical properties of the drugs and excipients have to be taken in consideration. As mentioned before the drug inside a reservoir system can either be present in a dissolved state or in a dispersed state. If the drug is present in a dissolved state, the concentration in the core will gradually decrease in time and as a consequence the release rate will also decrease. Because the total amount of dissolved drug in the core is limited by the saturation solubility, the release time is normally not very long. The release rate from this system can be adapted by varying the initial concentration of the drug dissolved. In principle, the release rate of more than one drug can be controlled by varying the concentration of each drug.
If on the other hand the drug is present in a dispersed state, the amount of dissolved drug is fixed by its saturation solubility. In this case the release rate can only be controlled by the thickness of the membrane. As a consequence this method can only be used to control the release rate of a single drug. Because the content of dispersed drug can be much higher than the saturation concentration, the release time is normally much longer.
In a multilayer design the thickness of the active layer and drug concentration can be designed in such a way that no more drug is used than is necessary. This is shown in the next figure.
During release of the drug a depletion zone will be formed. Once the depletion zone exceeds the thickness of the active layer the concentration of dissolved drug in the reservoir will decrease and as a result the release rate will decrease. By varying the thickness of the active layer duration of the drug release can be controlled.
Diffusion process over the membrane
The next figure depicts the influence of the storage time on the release of a drug from a reservoir system. Immediately after manufacturing the drug content in the membrane is still zero. If the release rate is determined immediately after manufacturing, the release rate is zero on t=0 and gradually increases in time until steady state is obtained (black symbols). If, however, the reservoir system is first stored at 25 °C, the drug content in the membrane increases upon storage until equilibrium is achieved with the amount of drug dissolved in the core polymer. The distribution of drug between both core and membrane is determined by the partition coefficient and the volume of each compartment. If the release rate is determined subsequently, the release curve shows a burst release. This burst release is attributed to the amount of drug that is dissolved in the membrane. In this case approximately half the amount of drug dissolved in the membrane is released quickly until again steady state release is obtained. If the total amount of released drug is plotted versus time, the so-called time lag and burst effect curves are obtained. The diffusion coefficient of the membrane can be deduced from the intercept of the steady state part of each curve with the time axis.
For a planar reservoir system the intercept of the steady state part of the time lag and burst effect curve with the time axis is respectively l2/3D and -l2/6D. Where l is the membrane thickness and D is the diffusion coefficient.
 |
 |
time lag and burst effect curves
References
R.W. Baker, H.K. Lonsdale, 1974, Controlled Release of Biologically Active Agents, Tanquarry, A.C., Lacey, R.E., Eds., Plenum: New York, 15-71
FastTtrack : pharmaceutics – drug delivery and targeting, London : Pharmaceutical Press, 2012