DRUG

Drug

Delivery Oral Drug Vasant

The

two main

DELIVERY

Systems Delivery

V. Ranade,

SYSTEMS

SERIES

5A.

PhD

advantages

of controlled drug delivery systems are: maintenance of theraconcentrations in the plasma through zero-order release withand elimination of the need for frequent single dose administrations. The oral and other therapeutic systems in human use have validated the concept that controlled continuous drug release can minimize the daily dose of a drug required to maintain the required therapeutic effect, while minimizing unwanted pharmacological effects. By minimizing patient intervention, a design feature of therapeutic systems, compliance is automatically enhanced. Oral drug delivery systems, in particular, have required innovation in materials science to provide materials biocompatible during prolonged contact with body tissues, bioengineering to develop drug delivery modules, and clinical pharmacology for elucidation of drug action under conditions of continuous controlled drug administration. Recent work in advanced oral delivery has been primarily focused on liposome technology and the concept that substances that are normally destroyed by the stomach can be protected long enough before they could be absorbed downstream. For cost and patient convenience, oral delivery certainly would be an attractive method. The nature of biologic substances, however, with their unique technical problems, will probably limit greatly those that can be delivered orally. Besides, where delivery rate control is critical, oral delivery, even when possible, would probably be insuciently precise. Oral delivery would also limit the substance to bloodstream delivery to the disease site. Even so, oral controlled drug delivery systems will likely find primary usefulness in specific carefully controlled therapies and prophylactic situations with due regard for drug interactions. This system represents a potentially very significant therapeutic modality. These delivery systems will find usefulness primarily in certain well-defined and well-controllable areas with due regard for individual patient variations. The purpose of the present article is to review oral controlled-release drug delivery systems, with particular emphasis on the practical aspects of testing and fabricating these systems and the underlying mechanisms by which control over drug release rate is accomplished. peutically optimum drug out significant fluctuations;

H

istorically, the most employed method oral ingestion. The first maceuticals was achieved pills;

this

dates

back

convenient and commonly of drug delivery has been controlled release of pharthrough the use of coated

over

1000

years.

Coating tech1800s with the

nology advanced in the mid to late discovery in gelatin coatings, sugar-coated made in Europe and used in the U.S., coated made

in the

From the Action

U.S.,

Medical

for reprints:

Vasant

IL, 60048. envelope.

For reprint

2

and

Marketing

V. Ranade,

#{149} J Clin Pharmacol

enteric-coated

request,

PhD,

co., 1219

please

1991;31:2-16

pills.

Libertyville,

Illinois.

pills pills

A major

Address

Deer Trail Lane, Libertyville, send self-addressed stamped

development of coating

in coating technology was a plurality of drug containing

combinations

of fats

first

by

reported

and

Blythe1

waxes.

and

This

was

the

the concept beads with approach

basis

was

for

the

first significant commercially marketed sustained release products in this country. Since the mid 1900s, hundreds of publications, and nearly 1000 patents have appeared on various oral delivery approaches encompassing delayed, prolonged, sustained, and, most recently, controlled release of the active substance.1 Unlike all other drug delivery systems, control of drug delivery by the oral route has a long history.

The first introduced

truly in

effective system, the ‘Spansule,’ This true prolonged release

1952.

was sys-

DRUG

DELIVERY

tern was marketed by Smith Kline & French Laboratories, Philadelphia, PA, and consisted of populations of small coated beads, which were placed in a capsule. The 50 to 100 or more beads per capsule comprised several bead populations, each designed to release at a different rate.2 In the mid to late 1960s, the term controlled drug delivery came into being to describe new concepts of dosage form design. These concepts usually involved controlling and retarding drug dissolution from the dosage form, but had additional or alternative objectives of sustained drug action. A primary objective of a controlled release system has been to enhance the safety of a product to extend its duration of action. Furthermore, today we have controlled release systems which are designed to enhance drug dissolution and increase dissolution rates in order to produce more reliable absorption and to improve the bioavailability and efficiency of delivery from the product, respectively. One such illustration of a controlled release product designed to enhance solubility, absorption rate, and bioavailability is the molecular dispersion of the antifungal drug griseofulvin. Dorsey, Lincoln, ME, has marketed a product called Gris-Peg, which is a molecular dispersion or solid solution of griseofulvin in polyethylene glycol. This molecular dispersion has such enhanced solubility properties that the dose of griseofulvin can be reduced by 50% over previously existing micronized powder forms of the drug. Due to the higher blood levels produced, less frequent dosing of the drug is also possible. An even newer concept of controlled release is that of site-specific release. New technology is being developed which uses drug delivery systems capable of prolonged retention in the stomach or other body cavities, using bioadhesion and other principles to control not only rates of release but also sites of release. Most recently, in the 1970s, yet another concept of drug product design and administration has appeared; the therapeutic system. The objective of the therapeutic system is to optimize drug therapy by the design of a product which incorporates an advanced engineering systems-control approach.3 The modern controlled release system is capable of producing not only sustained release, but controlled release, i.e., a release rate which is not greatly influenced by the gastrointestinal environment. The oral controlled release system is usually made of polymers and the mechanisms of release are generally regulated by (1) diffusion, (2) bioerosion or degradation, and (3) swelling or generation of osmotic pressure. Diffusion occurs when the drugpolymer mixture is exposed to the gastrointestinal

DRUG DELIVERY

SYSTEMS

SERIES

SYSTEMS

5A

fluid, resulting in release of the drug from the tablet or capsule. The polymers are excreted in the fecal contents. Bioerosion or degradation occurs with certain polymer-drug complexes when they pass through the gastrointestinal tract. Swelling or generation of osmotic pressure occurs with certain polymer-drug formulations when they are exposed to the gastrointestinal fluid, resulting in the release or expulsion of the drug. The advantages and disadvantages of the oral controlled release products can be listed. Advantages: Decreased fluctuation of serum concentrations resulting in reduced toxicity and sustained efficacy; decreased frequency of dosing improves patient compliance and reduces patient care time and possibly reduces total amount of drug used. Disadvantages: longer time to achieve therapeutic blood concentrations, possible increased variation in bioavailability after oral administration, enhanced first-pass effect, dose dumping, sustained concentration in overdose cases (after oral administration), lack of dosage flexibility, and, usually more expensive.4 When evaluating different proprietary controlled release drug products, one will find that the absorption characteristics of each product are likely to be different from one another due to different mechanisms of release. Controlled release preparations should generally not be considered bioequivalent or be substituted for one another, even though each product may contain a similar amount of the identical drug and meet the bioavailability requirements of the FDA (Food and Drug Administration). This consideration is especially important for drugs with narrow therapeutic ratios (antiarrhythmics, theophylline products, anticonvulsants). However, if two drug products have similar bioavailability as well as pharmacodynamics, e.g., therapeutic effect, substitution of such products should not cause any problem. Properties of drugs not suitable for controlled release formulations are: very short or very long half-life, significant first pass metabolism, poor absorption throughout the gastrointestinal tract, low solubility, and drug concentration not related to pharmacologic or therapeutic effect. The controlled release system is therefore designed to produce a sustained concentration of a drug in the human body. Many such products, especially in the form of oral controlled release formulations, are now available. Oral drug delivery technology has enjoyed commercial success with sales of products approaching one billion U.S. dollars in the U.S. alone. During the 1980s, more than two thirds of the 20 billion dollar U.S. drug market consisted of

3

RANADE

orally administered drugs and more than 85% of that market was in the form of solid oral dosage forms. Therefore, it is clear that a great opportunity in the pharmaceutical industry lies in converting solid oral dosage formulations to controlled release ones.5 Within the very profitable and large solid oral drug market segment, great opportunities exist for marketing controlled release formulations of several therapeutic categories of drugs. Drugs that are taken on a chronic or extended basis-cardiovascular, arthritic, respiratory, and analgesic products-often have the most potential for controlled-release drug delivery improvements. The oral controlled-release drug delivery segment is currently a small one but is growing very rapidly with total sales of more than $500 million. This market segment is fueled by an emerging trend in the drug industry favoring controlled-release products and improved drug delivery systems. In order to gain a better understanding of the factors involved in developing controlled-release oral drugs, it is worthwhile to understand some of the basic elements of the gastrointestinal (GI) physiology, particularly as it pertains to the mechanisms and factors influencing drug absorption and the GI transit time. Previously, this has had a profound influence on the design of oral controlled drug delivery and as our understanding of GI physiology increases, it should be possible to base strategy for controlled drug delivery on molecular or cellular events that are so central in designing criteria and limitations of this technology.6

FEATURES

OF THE

GI TRACT

While there are a number of drugs in sustained release form that have an intended site of action in a local region of the GI tract, the overwhelming majority of oral drugs are targeted to act elsewhere in the body. Thus, the GI tract is merely a conduit to get drug to the bloodstream with the expectation that blood will then perfuse most areas of the body. For oral dosage forms, it can therefore be assumed that spatial placement of the drug in the body, i.e., targeting, is not going to be a significant design criterion and the focus will be primarily on the temporal aspects of drug release. An assumption is usually made that drug levels in the bloodstream parallel those in the biophase. Further, it is assumed that a constant level of drug in the blood, for a specified period of time, is a desired endpoint. This is most easily accomplished by direct administration of the drug such as by an IV drip, where the rate of drug administered is adjusted,

4

#{149} J Clin Pharmacol

1991;31:2-16

based on the pharmacokinetic properties of the drug in order to achieve an invariant level. In the simplest form the rate of drug administration is computed on the basis of replacement. Thus, the rate of drug eliminated is the same as the rate of drug administered; that is, the product of the desired blood level of the drug, the apparent volume of distribution of the drug and the first-order elimination rate constant for the drug.7 For routes involving drug absorption, such as the oral or intramuscular route, an absorption phase is introduced prior to appearance in the blood. This does not change the approach used in computing the desired rate of release of drug from the dosage form to achieve a constant level of drug in the blood. Thus, the rate constant of drug release from an oral sustained release dosage form will be computed in exactly the same manner as previously described. However, since a lag time has been introduced in getting drug to the blood, i.e., in the absorption phase, it is necessary to make some adjustments in computing the total amount of drug that will be contained in the dosage form. It is therefore common to see a sustained release oral system composed of two parts, an immediately available dose to be used to establish therapeutic levels of the drug quickly, and a sustaining portion that is intended to slowly release drug for eventual absorption and maintenance of constant blood levels. The total drug, therefore, is the summation of immediately available dose and the sustaining dose. It is also clear that a zeroorder rate of drug release will commonly be used to sustain levels of the drug. Thus, the total amount of drug in the sustained dose will be equal to the zeroorder release rate constant times the total number of hours of sustained effect desired. This is therefore an elementary starting point in the design. It is often difficult to obtain a zero-order release, although this is not critical, since a slow first or other order release will look like a zero-order release, in terms of the drug-blood level achieved.8 A decision has to be made that an immediately available dose will not be incorporated into the dosage form. When this occurs, the time to reach therapeutic drug levels will be delayed, as will the actual steady state levels achieved. An important assumption, in the absence of data to indicate otherwise, is that the drug is uniformly absorbed during its transit through the GI tract. For those drugs which are absorbed preferentially in a specific region of the GI tract, i.e., possessing a “window” for absorption, sustained release products are very difficult because of the problem of localizing a drug delivery system. There are a number of additional constraints on the

DRUG

DELIVERY

design of oral controlled drug delivery systems: dose size, drug molecular size, charge and pKa, aqueous solubility, partition coefficient, stability, absorption, metabolism, half-life, margin of safety, side effect, and clinical response.9

Targeting

of Drugs

in the

Gastrointestinal

Tract

The gastrointestinal tract is the preferred site of absorption for most therapeutic agents as seen from the standpoints of convenience of administration, patient compliance and cost. The majority of oral dosage forms consist of tablets and capsules that are often provided as instant release systems designed to disintegrate rapidly in the stomach.1#{176} The dissolved drug substance is usually absorbed from the small intestine. The efficiency of these processes of release and uptake is dependent upon the physicochemical characteristics of the drug (e.g., solubility, stability in acid and alkaline environments, permeability through gastrointestinal membranes) as well as physiological variables such as gastrointestinal transit. Recently Ho and others have attempted to quantify these factors with their ‘reserve length’ concept.11 Briefly, their approach has been to consider the distance a dispersed drug would have to pass down the small intestine before the total available dose was absorbed. Their mathematical analysis has shown that the more efficient the dissolution and absorption processes, the greater the reserve length. Whether the small intestine alone should be taken as the predominant absorption site is debatable. The opinions presented in the literature would suggest that absorption of drugs from the large bowel is often poor and erratic. However, recent studies conducted on beta blockers indicate that the large intestine may

have

a more

significant

contribution

to

total

absorption than hitherto realized.12 Controlled-release (modified release) dosage forms are gaining rapid popularity in clinical medicine. These more sophisticated systems can be used as a means of altering the pharmacokinetic behavior of drugs

in order

to provide

twice-a-day

or

once-a-

day dosage. Other applications include enteric coatings for the protection of drugs from degradation within the gastrointestinal tract or the protection of the stomach from the irritating effects of the drug, and the delivery of drugs to so-called absorption windows or specific targets within the gastrointestinal tract, particularly the colon. Much about the performance of a system can be learned from in vitro release studies using conventional and modified dissolution methods; however, an essential stage in the development must be a

DRUG DELIVERY

SYSTEMS

SERIES

SYSTEMS

5A

subsequent evaluation in vivo. Davis used the noninvasive technique of gamma scintigraphy to follow the gastrointestinal transit and release characteristics of a variety of pharmaceutical dosage forms in human subjects. Such studies not only provide insight into the fate of dosage form and its integrity, but also allow a correlation to be made between the position of a system in the gastrointestinal tract and resultant pharmacokinetic profiles. Davis has also studied various methods for the evaluation of the fate of orally administered dosage forms and the performance of orally administered dosage forms10: radiology (X-ray), endoscopy, radiotelemetry, epigastric impedance, gamma scintigraphy and deconvolution

of pharmacokinetic

data.

A majority

of these

techniques can be considered as direct and either invasive or noninvasive approaches. Deconvolution of pharmacokinetic data represents an indirect means of obtaining information on the performance of an administered dosage form.1#{176} Gastrointestinal motility-a vigorous and variable phenomenon-presents a major impediment to the invention of residential devices necessary for sitespecific drug release. This is most easily overcome in the large intestine where conditions are most predictable and quiescent. Targeting delivery to the stomach is technically more difficult due to the power of gastric movement during both the digestive and interdigestive phases. Buoyancy, dimensional change, mucosal adhesives, and drugs such as propantheline and fatty excipients have all been suggested as methods of ensuring gastric retention of small devices. The carcinogenic nature of nitrosamines derived from the interaction of nitrates in food with secondary or tertiary amines in both food and drugs has prompted the delivery of N-nitrosoblocking agents to the stomach. The ‘hydrodynamically balanced system’ for example, has been the subject of earlier patents and derives its effect from hydration

and

swelling

of

the

outer

layers

which

entrap significant quantities of air and confer a density on the partially wet compact which is less than that of gastric fluid. This buoyancy is claimed to greatly extend the residence time in the stomach, allowing its active principles to diffuse into the gastric milieu and prevent the formation of nitrosamines.13

In appraising these principles of the colonic ticable. These systems radation in the colon or Variation in both factors effectiveness and future devised which ensure

and similar inventions, the delivery systems seem pracemploy either bacterial deggastrointestinal transit time. may confer variability on innovations may well be predictable and catastrophic

5

RANADE

break-up and with arrival in the colon being more predictable. Many of the inventions applied to gastric residence time appear to produce only modest and variable increases. Greater effects, such as might follow a major change in dimension after ingestion, obviously carry a toxicological risk which is never addressed in patent disclosures. However, since the principle is also of value in controlling the intensity and duration of drug action, it will continue to attract inquisition. So far, much of what has been described relates to controlled-release dosage forms intended to give steady state sustained blood levels for more effective drug therapy. The scintigraphic method can be used

to evaluate dosage forms that have been designed to release a drug in designated areas of the gastrointestinal tract. A simple example is an enteric coating that should ensure that the drug will be released when

it reaches

the

small

intestine.

More

specific

targets for delivery within the small intestine include the duodenum, for the preferential absorption of peptides and proteins by exploiting known facilitated transport mechanisms for dipeptides and tnpeptides, as well as the delivery of antigens and allergens to the M-cells residing in the Peyer’s Patch regions.14 Similarly, there is growing interest in the specific delivery of drugs to the colon, either for local treatment such as that of ulcerative colitis and irritable bowel syndrome, or for the systemic delivery of compounds that are normally not well absorbed from the gastrointestinal tract by exploitation of the long residence time in the colon.15 Muranishi and others in Japan have demonstrated that it is possible to modify the absorption characteristics of the colon

using

a variety

of

absorption

enhancers

in-

cluding mixed micelles.’#{176} Clearly for such applications sophisticated delivery systems must be developed that will allow site-specific delivery of not only the drug, but also the absorption enhancer.

Mathematical Models Release Kinetics Controlled agents ther in During during havior,

release of drugs and other bioactive can be achieved by incorporating solutes dissolved or in dispersed form in polymers. the design stage of these formulations experimental verification of their release it is desirable to develop and use simple

sophisticated

release From

mathematical

kinetics. a mathematical

controlled ing to the

6

for Controlled

models

modeling

release systems may controlling physical

#{149} J CIIn Pharmacol

1991;31:2-16

to describe

point

be classified mechanisms

eior beyet the

of view, accordof re-

lease of the incorporated solute. Mathematical modeling of the release kinetics of specific classes of controlled release systems may be used to (1) predict solute release rates from and solute diffusion behavior through polymers, thus avoiding excessive number of experiments, and (2) to elucidate the physical mechanisms of solute transport by simply comparing the release data to mathematical models. Peppas16 has discussed the analytical work on the various systems he classified as diffusion controlled, osmotically controlled, and chemically controlled. Diffusion controlled systems contained reservoir, matrix, and porous systems. In the chemically controlled systems, shrinking core models provide the most accurate description. Cooney developed simple expressions for solute release from cylindrical and spherical devices.16 Hopfenberg17 derived expressions for solute release from erodible slabs, cylinders, and spheres. Mathematical models exist for erodible systems where solute release from the surface is also important; these have been recently discussed by Lee.’8 Pharmacokinetic The primary release is the

Modeling

and Bioavailability

goal of sustained prolonged delivery

or controlled drug of a drug to a partic-

ular body compartment or anatomical target site. This is accomplished by the application of a therapeutic drug delivery system designed to control both the temporal and spatial aspects of drug deposition.19

Previous drug delivery vitro testing perimentation

investigations

in the

field

of controlled

have centered on the design and in of the device. Animal and clinical exhas been largely limited to reports on

the temporal pharmacodynamic performance of drugs released from the implanted, inserted, or surface-applied polymer vehicles. Few studies have provided details of the kinetic behavior of these release systems as well as the distribution, metabolism, and excretion of the controlled release drugs throughout the period of in vivo evaluation. These data are essential for exploration of the pharmacokinetic characteristics of such systems and for providing the development of predictive models. Pharmacokinetic simulations are valuable in that they can be used to determine how the delivery pattern of a given therapeutic system is transformed into blood or tissue levels of any given drug. These models also increase experimental efficiency by allowing the use of preliminary measurements and analyses to optimize the further design of controlled drug release systems. The process of theoretical simulation and experimental validation of these predictions permits

DRUG

a legitimate

extrapolation

while conserving time fore, the mathematical physical and chemical trolled and sustained

tial

in the

overall

systems.

The

of and

animal modeling processes drug release

experimental

bioavailability

Thereprimary conis essen-

analysis

of these

refers

tion of the drug dose administered which reaches the bloodstream. tered intravenously would thus

data

resources. of the governing systems

pharmacokinetic

term

DELIVERY

to that

por-

to the patient A drug adminisbe 100% bioavail-

able by definition; however, a conventional drug administered by the oral, rectal, or transcutaneous routes would not necessarily be 100% bioavailable unless it were completely absorbed. A sustained release oral preparation from which drug is incompletely absorbed during transit of the tablet through the gastrointestinal tract suffers from reduced bioavailability. Similarly, drug from a controlled release implant which is retained in the device indefinitely (such as drug which deteriorates, is irreversibly complexed with the polymer vehicle or has restricted drug diffusion due to a rate-limiting fibrous capsule surrounding the implant device) will not be made bioavailable to the patient. After a drug is administered to a patient it is helpful to be able to predict the therapeutic or toxic response of that individual to a given drug dosage or dosage regimen. For some drugs, the pharmacodynamic response experienced by the patient may be

directly correlated of the drug which the

patient’s

response stances sample other study within

to a concentration

sera

may also by analyzing of biopsied available body of temporal the body and

be

obtained for drug

is defined models are

that, when properly constructed, formation on the process of drug the body and allow predictions

response

in certain concentration

circumin a

tissue, saliva, gastric fluid, or fluid or excretory product. The changes in drug distribution the kinetics of drug absorption

metabolism and excretion netics. Pharmacokinetic

patient

measurement

is most often obtained by sampling or urine. The pharmacodynamic

for a given

dosage

as pharmacokipredictive tools

both provide indisposition within on drug levels and

from

the interval.

Incorporated directly or indirectly into a predictive pharmacokinetic model or simulation are terms to describe the physiological handling of a drug. These terms reflect the pathophysiological status of an individual patient or groups of subjects of similar physiological status.2#{176} Within the classical scheme of pharmacokinetics, a drug is modeled based upon its distribution into a number of ‘compartments.’ These compartments are actually a simulation of the physiological drug han-

DRUG DELIVERY

SYSTEMS

SERIES

SYSTEMS

dling

5A

process

and

there

may

or may

not

be a basis

for

comparison to actual anatomical volume entities or real physiological processes. Notari21 defines such a compartment as a “kinetically distinguishable pool” in which the drug of interest is present in equal concentration throughout the volume of the defined compartment and that the drug, at all points within the compartment, behaves in an identical manner with respect to any kinetic mechanisms operable within the compartment (e.g., drug transfer, metabolic conversion, elimination etc.). The various kinetic mechanisms involved in the disposition of the drug within the body are quantitatively described by kinetic rate constants. For the practical linear pharmacokinetic models of interest these rate constants describe either zero-order or first-order kinetic processes. For nonlinear models the kinetic constants employed may approximate pseudozero-order or pseudofirst-order reactions, depending on the concentration of the active species of interest (i.e., Michaelis-Menton equations describing enzymatic processes). In general, classical pharmacokinetic models are designed to employ the least number of compartments necessary to explain experimental observations of the variation of drug concentration levels (generally as measured in the plasma) with time. The majority of useful classical pharmacokinetic compartmental models are described as “open models” meaning the compartments of interest are in communication with drug input and/or output functions. For the majority of commonly employed compartmental schemes, the drug transfer rate processes are zero-order or first-order kinetic reactions. Anderson has summarized a number of useful simple physiological pharmacokinetic models.’9 Design

and Fabrication

of Oral Systems

The overwhelming majority of controlled release systems rely on dissolution (bioerosion), diffusion, or a combination of dissolution and diffusion to generate slow release of drug to the biological environment. Starting with limited data on a drug candidate for a sustained release system, such as some physical-chemical properties of the drug, the dose, the rate constants for absorption and elimination, and some elements of metabolism, one can compute a desired release rate for the dosage form, the amount of drug required, and the preliminary strategy for the dosage form to be used. While the desirability of having a correlation between in vivo bioavailability and in vitro release is obvious, many, if not most, sustained release prod-

7

RANADE

ucts

do

varies when

not

show

such

a correlation

unless

one

the in vitro a correlation

particular other drug better when rather than test conditions number of

experimental conditions. Thus, is found for a particular drug in a dosage form, it cannot be applied to anor dosage form. The correlation becomes the in vitro test is done in a pH gradient distilled water. Further, optimization of can also help minimize variations. A such systems have been described and

used.2224 Within the scope of this review, a variety of controlled release systems are discussed. Included among these are the following: 1.Dissolution 2. Osmotically 3. Diffusion 4. Chemically 5. Miscellaneous 1.

Controlled

Dissolution

The

most

trolled

Controlled Release Controlled Release Controlled Release Controlled Release Controlled Release

important

drug

delivery

Release

attribute

systems

of membrane-conis their capability

to

maintain a constant rate of drug delivery over a reasonably long period of time. The duration of constant drug deliveries must be compatible with physiologic constraints and the route of administration. For example, while a duration of several weeks may be appropriate for a membrane-controlled implant, it is much too long a time for an oral dosage form. Clearly, the selection of a membrane system and/or its duration of action must be based on an appropriate set of constraints. It may well be that constant

input rate of drug provides little real advantage over well controlled, first-order mechanisms under certain biopharmaceutic conditions. However, there are certainly situations which call for membranecontrolled systems which provide constant rate input for times ranging from several hours to several months.25 The oral own unique time frame

limited

route of drug administration presents its set of problems and constraints. The or “window” for absorption is obviously

to the

total

GI residence

time.

Even

this

time

may be an overestimate if the drug in question is absorbed only in certain segments of the GI tract. Moreover, the interindividual differences in residence time and motility patterns are generally quite large. Taking into account the gastric emptying mechanism and its duration, small intestine transit time, and large intestine transit time, it would seem that a reasonable duration constraint in the GI tract is approximately 24 hours. Another time-related

8

#{149} J ClIn Pharmacol

1991;31:2-16

constraint is associated with absorption through the GI mucosa into the general culation. In order to control the delivery the ultimate target organ via the general

of the drug hepatic cirof drug to circulation

it is necessary that the system release its contents at a slower rate than the physiologic absorption rate. Moreover, when gut wall and first pass liver metabolism are significant, the rate of drug delivery to the GI tract may have significant effects on the amount of unchanged drug which reaches the peripheral circulation and the rate at which metabolism takes place. Finally, the excretion rate or clearance of the drug from the peripheral circulation and/or any tissue compartments which may absorb significant quantities of the drug will ultimately affect the selection of the type of drug delivery system. The absorption, distribution, and elimination of

drugs are normally simplified by considering them all to be simple first-order processes. Given the average 24-hour residence time and high interindividual variability in the GI tract, it would seem that only drugs having relatively short elimination halflives should be considered for membrane-controlled reservoir systems intended for oral administration. There is, however, a means to approach the desired steady state concentrations from above, rather than below their constant values. This can be done by providing a portion of the dose in an immediately available

form,

such

as the

membrane

loading

dose

described by the so-called “burst effect” or by simply including it on the outside of the reservoir. Since the GI tract defines the maximum dosing interval, we are generally constrained to limiting the total dose to that established for the drug itself, independent of the delivery system. This constraint, in itself, may be severe enough to render any controlled release form impractical. In addition, the standards of bioavailability compared to a “really bioavailable standard” dosage form require that virtually all of the dose be absorbed within the time period in which the dosage form is in contact with the GI tract. This latter constraint forces one to consider the kinetics of drug release and effect on plasma concentration during the later periods of time when the reservoir concentration has dropped below the saturation point and steady state delivery is lost. In sustained release formulations employing dissolution as the rate-limiting step, drug release is controlled by dissolution (bioerosion) of a polymer or by a chemical reaction from a soluble subunit. Individual particles or granules containing drug can be uniformly dispersed in the matrix or coated with varying thickness of coating material resulting in dissolution and release of the drug over extended

DRUG

DELIVERY

periods of time. If the dissolution process is assumed to be diffusion-layer controlled where the rate of diffusion from the solid surface to the bulk solution through an unstirred liquid film is rate-limiting, the flux is the product of the diffusion coefficient and the concentration gradient from the solid surface to the bulk solution side. Flux can also be defined as the flow rate of material through a unit area. Encapsulated Dissolution Control. Either the drug or a drug-containing nonpareil seed can be coated with slowly dissolving polymeric (or wax) materials. Once the polymeric membrane has dissolved, all the drug inside the membrane is immediately available for dissolution and absorption. Thus, drug release can be controlled by adjusting the thickness and the dissolution rate of the polymeric membrane. If only a few different thicknesses of the membrane are used, usually three or four, drug will be released at different, predetermined times resulting in pulsed dosing, i.e., repeat action. If a spectrum of different

thicknesses is employed, a more uniformed sustained release can be obtained.26 The membrane-coated particles can be directly compressed into a tablet form, or placed in capsules. If the particles are compressed into a tablet, fracture of some of the surfaces generally occurs with a resultant increase in release rate. It is a common practice to employ one fourth or one third of the particles in nonsustained form, i.e., particles without a barrier membrane, to provide for immediate release of drug. Alternatively, a portion of drug can be placed in a rapidly dissolving coating-membrane to quickly establish therapeutic levels. One of the principal methods of coating a drug is microencapsulation, wherein the drug solution or crystal is encapsulated with a coating substance. The most common approach for microencapsulation is coacervation, which involves the addition of a hydrophilic substance to a solution of colloid. Whether a drug is water sensitive or not, it can be microencapsulated if the drug is protected from the aqueous environment by coating with polymers, such as ethylcellulose, cellulose acetate phthalate, or carnauba wax prior to microencapsulation. The thickness of the coat can be adjusted from less than 1 tm up to 200 m by changing the amount of coating material from 3 to 30% of the total weight, and wall thickness can be theoretically calculated from the known capsule size. Microencapsulation has the additional advantage that sustained drug release can be achieved with taste abatement and better GI tolerability. Good examples of microencapsulations are microencapsulated aspirin and potassium

DRUG

DELIVERY

SYSTEMS

SERIES

SYSTEMS

5A

chloride.

In both

cases,

drug

effects

from

the

mi-

croencapsulated dosage forms are more prolonged and less irritating than the same amount taken as ordinary tablets. Both formulations show the same total drug absorbed, as calculated from the area under the curve. One of the disadvantages in employing microencapsulation is that no single process can be applied to all core material candidates. Moreover, incomplete or discontinuous coatings can cause unstable and irregular release characteristics.27 Sears28 applied synthetic phospholipids as a coating material to obtain sustained release from microcapsules. The synthetic phospholipids, when the polar moiety of the phosphatidylcholine head group was altered, showed a decreased rate of phosphorylase C hydrolysis. These compounds were employed as surfactants and encapsulation agents for such drugs as insulin, which requires protection from hydrolysis in the stomach. Microorganism cells were also used as microcapsules. Yeast, molds, or other fungi which synthesize fat within themselves can

absorb fat-soluble the drugs.29

drugs

and

prolong

the

potency

of

Matrix Dissolution Control. The two general methods of preparing drug-polymer particles are congealing and aqueous dispersion methods. In the congealing method, drug is mixed with polymeric substances or waxes. The wax or polymer drug material can be cooled and put through a screen to obtain the correct particle size or it can be spray congealed. Kawashima et a13#{176} used a modified spherical agglomeration technique as an alternative to the spray congealing method. In the aqueous dispersion method, drug-polymer mixture is simply sprayed or placed in water and then collected. Usually the

aqueous dispersion method shows a higher releaserate than wax congealing or spraying probably due to the increased area and entrapment of water. Recently, Heller and methyl vinyl ether-maleic which has extraordinary

release These

Trescony31 anhydride sensitivity

rate, to the surrounding polymer systems show

synthesized copolymer, and hence drug

environmental a characteristic

pH. pH

above which they are completely soluble and below which they are completely insoluble. The specific pH depends on the size of the alkyl group in the copolymer ester. Thus, the polymer dissolution and drug release can be strictly controlled to fit any desired pH environment. These systems have the potential to be used in oral controlled delivery systems where absorption at a specific site in the GI tract is desired. Zero-order release at a particular site in the

9

RANADE

GI tract can be achieved by maintaining pH of the system. Graffner and coworkers32 prepared sustained release procainamide in matrix forms and compared the release rate to that of IV dosing. The rate of absorption in vivo was well correlated to the in vitro dissolution pattern. A variety of slowly dissolving coatings, such as those based on various combinations of carbohydrate sugars and cellulose, polymeric materials, and wax, is available. 2.

Osmotically

Controlled

Release

In addition to the solution-diffusion mechanism, the drug release from a membrane reservoir device can also take place through an orifice in the membrane via an osmotic pumping mechanism, where a semipermeable membrane such as cellulose acetate is used to regulate the osmotic permeation of water. For a system of constant reservoir volume, the device delivers a volume of drug solution equal to the volume of osmotic water uptake within any given time interval. The rate of osmotic water influx and therefore the rate of drug delivery by the system will be constant as long as a constant thermodynamic activity gradient, usually derived from a saturated reservoir with excess solid is maintained across the membrane; however, the rate declines parabolically once the reservoir concentration falls below saturation. Such an osmotic delivery system is capable of providing not only a prolonged zero-order release but also a delivery rate much higher than that achievable by the solution-diffusion mechanism. The system is also applicable to drugs with a wide range of molecular weight and chemical composition which are normally difficult to deliver by the solution-diffusion mechanism. There are basically two types of osmotic delivery devices, namely, the so-called miniosmotic pump and the elementary osmotic pump. In the miniosmotic pump delivery system, the drug reservoir is separated from the osmotic agent compartment by a movable partition. At the other end of the osmotic compartment is the semipermeable membrane and a rigid impermeable material forms the remaining three sides of the pump, with a delivery orifice at the front. The elementary osmotic pump consists of an osmotic core containing the drug, surrounded by a semipermeable membrane with a delivery orifice. The delivery rate from these devices is regulated by the osmotic pressure of both osmotic agent of the core formulation and by the water permeability of the semipermeable membrane.33 Unlike the solution-diffusion mechanism, the osmotic delivery system involves a volume flux of

10

#{149} J Clin Pharmacol

1991;31:2-16

water

across

a semipermeable

terdependence

has

been

of

treated

the

membrane.

diffusive

rigorously

and

in the

The

volume

realm

in-

fluxes

of nonequi-

librium thermodynamics. In the miniosmotic pump system, as long as a large enough reservoir of osmotic agent is present, the delivery of dissolved drug at any concentration can be of zero order because of the separate compartmental design. Whereas in the elementary osmotic pump, since the core formulation is also the osmotic driving agent, the delivery rate is constant as long as excess solid is present within the drug reservoir. The experimental evaluation of the coefficients appearing in the kinetic equations is necessary regardless of whether one is interested in obtaining

numbers membrane is attempting

release coefficients.

of suitable

accuracy

reservoir drug to understand

sufficiently Many

well

for

the

design

to be

parameters

able

to predict

such

of

or if one of drug

delivery systems the mechanism as

the

the drug

loading, membrane thickness, membrane area, and device geometry are readily measurable during the system fabrication. The resulting drug release characteristics, such as the membrane permeability, expressed as a product of the diffusion and partition coefficients of the drug in the membrane, and the effective releasing period would have to be determined experimentally. Various experimental methods are available for the evaluation of partition and diffusion coefficients in polymers. The measurements of permeation rate and sorption/desorption kinetics are commonly employed for this purpose. This system requires only osmotic pressure to be effective, and is essentially independent of the environment. As a consequence, this should be an excellent sustained release system for oral dosage forms because there are rather inconsistent conditions of pH and mixing in the digestive tract. Thus, the drug delivery rate from an oral osmotic therapeutic system can be precisely predetermined regardless of pH change. In fact, the delivery rate of sodium phenobarbital from this system into artificial gastric juice at pH 2 and intestinal fluid at pH 7.5 (containing no enzymes) was shown to be pH independent.34 Development of OROS System. The quality of a therapeutic system designed to control pharmacologic effects through control of plasma concentrations can be judged through the constancy of drug concentrations attending its use. The flatness of plasma concentrations curves can be expressed by the ratio of

maximum to minimum dosing interval at steady

concentration state for

within repetitive

one injec-

DRUG

tions.

For

plasma

administration

concentrations

obtained

of a therapeutic

system,

DELIVERY

after

this

ratio

the

is,

in addition to pharmacokinetic constants and dosing interval, a function of the system’s design parameters, and is therefore called the dosage form index, DI.35 During selection of a drug substance for delivery via the OROS system (developed by AIza Corporation, Palo Alto, CA) one also must consider the site of entry and factors that can modify the rate and extent of drug absorption in route from that site to the target tissue. Considering that the OROS system is a solid tablet-sized object, it will pass the GI tract within the transit time of foodstuff. To reduce drug plasma concentration fluctuations on repetitive administration of an OROS system, it is also necessary to consider the half-life associated with the distribution phase. Some drugs, like lithium are rapidly absorbed but are distributed slowly in the tissues, giving rise to sharp absorption peaks after administration. For lithium and many other drugs, such peaks are associated with side effects that can be prevented by administration in a controlled delivery dosage form (Figure).

Factors that will affect control over plasma concentrations are gut and liver metabolism, which modify the input rate of drug into the body, and resistance to drug transport across the gut wall, which serves as the port of entry to the body. Only when the rate-limiting step resides in the therapeutic system, and not in such endogenous factors, can control over drug administration be achieved. The drug

OROS that

(Theophylline) appears to have

System. Theophylline the desired attributes

DRUG SOLUTION

f

is a for

DELIVERY .-ORIFICE

SYSTEMS

delivery

Figure. Cross-sectional diagram with permission from Annual Vol. 15, p. 308, 1980, Academic Corporation, Palo Alto, CA).

of an OROS system (reprinted Reports in Medicinal Chemistry. Press Inc., Orlando, FL, and AIza

DRUG

SERIES

DELIVERY

SYSTEMS

via the

OROS

system.

Theophylline

is used

primarily for treatment of obstructive airway diseases. Its pharmacokinetics and pharmacology in man have been well documented. In particular, its pharmacology was studied over a wide range of plasma concentrations during intravenous administration of aminophylline in man by Hendeles et al.36 This and other studies have shown that the OROS system allows safe and effective delivery of theophylline and creates less need for individual dose adjustment than a system with a higher DI. A currently marketed over-the-counter appetite suppressant product, Acutrim (Ciba, Summit, NJ), incorporates Alza’s OROS system. In Acutrim, the active ingredient, phenylpropanolamine, is released at a controlled rate. Another benefit of controlled rate delivery of Acutrim is that in this form, phenylpropanolamine does not produce amphetamine-like side effects that are normally seen in other conventional formulations. In addition to Acutrim, a product called Osmosin (Merck, West Point, PA) has used Alza’s OROS systern to deliver antiarthritic drug, indomethacin. However, this product produced a problem, that was probably caused by the inclusion of potassium in the formulation of this system. Another product using OROS has been Lopressor. For insoluble or extremely soluble drugs, Alza corporation has designed another system, the “push-pull” OROS. This system has two compartments, one containing an osmotic agent and the other containing the drug. A semipermeable membrane surrounds both. In the gastrointestinal tract, water enters each compartment through the membrane at a different rate. In the drug compartment, soluble drugs are formulated into solutions and insoluble ones into suspensions. As water enters the other compartment, it expands and pushes against the drug compartment. This causes the drug solution or suspension to be released at a controlled rate through the minute hole in the membrane surrounding the drug compartment. Elan Corporation of Ireland has developed its own

osmotic

‘OSMOTIC CORE CONTAINING DRUG

5A

pressure

system

called

MODAS.

MODAS

stands for Multi-Directional Oral Absorption System. MODAS is similar to OROS in some respects and quite different from it in others. Like OROS, this system consists of a tablet core surrounded by a semipermeable membrane. However, unlike OROS, MODAS has a multitude of small pores through which the drug solution can exit. The rate of drug solution release can be controlled by the composition of the membrane. Since the drug release is multi-directional, concentration of drug in any one area of the gastrointestinal tract is avoided. Elan has

11

RANADE

identified MODAS, theophylline, chloride,

3.

Diffusion

a number of drugs which including aiphamethyldopa, quinidine, indomethacin, and naproxen.

Controlled

The most commonly rial in drug delivery amorphous

and

are

suitable ibuprofen, potassium

for

Release used type of membrane systems is homogenous

semicrystalline

polymers

matefilms of above

their glass transition temperatures. The drug transport occurs by first dissolution in the membrane at one interface followed by diffusion down a chemical potential gradient across the membrane and finally the release from the second interface into the external medium. Such a solution-diffusion membrane is typically observed in hydrophobic membrane materials such as silicone rubber, and ethylene vinyl acetate copolymer. A similar mechanism is also responsible for drug permeation through most of the swollen hydrogel membranes.16 The rate of drug permeation through solution-diffusion membranes is directly proportional to the product of the drug diffusion coefficient in the polymer, and the polymer/solution partition coefficient. The former is a kinetic or nonequilibrium transport parameter, while the latter is an equilibrium thermodynamic property. Despite the progress in estimating the diffusion and partition coefficients of simple gases in polymers, no reliable method is available so far for the quantitative prediction of both the diffusion and partition coefficients of more complicated organic molecules in polymers. Nevertheless, various trends can be identified based on accumulated experimental evidence in the literature. Above the polymer glass transition temperature, the drug diffusion coefficients in a polymeric medium generally decrease with increasing drug molecular weight, molecular size, crystallinity of the polymer, and the amount of filler in the polymer. However, the drug diffusion coefficient would increase with increasing plasticizer content and solvent swelling in the polymer. Other parameters such as copolymerization, cross-linking, and grafting, as well as the distribution and orientation of crystallites may either increase or decrease the observed drug diffusion coefficient. In many instances, the concentration dependence of the diffusion coefficient may further complicate the picture.37 As to the solubility and partitioning effects, one generally observes that the drug would be more soluble in the polymer phase as the difference in the

12 #{149} J Clin Pharmacol

1991;31:2-16

solubility

parameters

of the

drug

polymer

becomes

smaller. Recently, Michaels and Wong38 have demonstrated that the steroid permeability in polymers can be correlated with thermodynamic parameters such as the melting temperature of the steroid and the solubility parameters of the steroid and polymer. The partition coefficient, defined as the ratio of the drug concentration in the external solvent medium, may also be concentration-dependent. Many of the partition coefficients reported in the literature were measured in saturated drug solutions and subsequently used in situations where the drug concentration may deviate from saturation considerably. Depending on the nonlinearity involved in the absorption isotherm, such practice can lead to appreciable error in the determination of per-

meation

parameters.

Therefore,

for

the

design

of

a

specific membrane-reservoir drug delivery system, it is necessary to determine both the drug diffusion and partition coefficients experimentally. It is preferable that one should also carry out selected experiments over the entire concentration range of interest so that the concentration dependence can also be established. Being cross-linked and hydrophilic, hydrogel polymers are unique in that they are quite glassy in the dry state, whereas in the presence of water they can swell significantly to form an elastic gel. In addition to having good biocompatibility, their ability to release drug in aqueous medium and the ease of regulating such drug release by controlling the water swelling and cross-link density make hydrogels particularly suitable as carrier matrices or ratecontrolling membranes in the controlled release of pharmaceuticals. A typical hydrogel membrane device usually consists of either a solid core of drug or a slightly crosslinked hydrogel matrix containing dissolved or dispersed drug, and a surrounding rate-controlling bydrogel membrane. In both cases the membrane can be either prefabricated or coated and subsequently polymerized. When a hydrogel matrix is used as drug reservoir, a rate-controlling membrane can also be formed by a newly developed interpenetrating network (IPN) technique, where the surface layer of the matrix is first imbibed by heat or UV irradiation or an in situ polycondensation generated by immersion in a second reactant solution to arrive at a less permeable, rate-controlling membrane layer.39’40’43 For different specific end uses, hydrogel membrane devices may be stored in either dry or hydrated states before use. The release of water-soluble drugs from initially dry hydrogel membrane devices generally involves the swelling of the

DRUG

membrane

and

subsequent

dissolution

DELIVERY

or swelling

of

the core. In the case of a membrane originally saturated with drug, a simultaneous absorption of water and desorption of drug via a swelling-controlled diffusion mechanism is also observed. Thus, as water penetrates a glassy hydrogel membrane device, the polymer swells and its glass transition temperature

is lowered. At the same time, the diffuses through this swollen rubbery external releasing medium.41’42

dissolved drug region into the

As to the effect of equilibrium water content on the mechanism of drug permeation through hydrogel membranes, Yasuda et a144 have derived a theoretical expression relating the solute diffusion coefficient in a water soluble polymeric membrane to the free volume and the degree of hydration in the membrane. Conformity of experimental results to the theory suggests that the permeation of solute occurs predominately through the porous regions of the network. As pointed out by Yasuda et al, these porous, water-filled regions through which the transport of permeant can occur may only be conceived as fluctuating pores or channels of the polymer matrix which are not fixed either in size or in location. More recently, Zentner et al,5#{176} studied the effect of cross-linking agent on the progesterone permeation through swollen hydrogel membranes. In

addition coefficient

to the with

decrease increasing

in progesterone cross-linker

diffusion level, they

found that at low concentrations of cross-linker, the chain length of the cross-linker did not affect the “fluctuating pore” permeation mechanism. However, at high concentrations of cross-linker the diffusion coefficient of progesterone in the system with a shorter cross-linker, ethylene glycol dimethacrylate, was relatively independent of the cross-linker

concentration.

This

was

rationalized

as a change

in

permeation mechanism to that of a solution-diffusion controlled process. Such transition from porous to solution-diffusion transport is consistent with the water permeation results previously reported by Chen45 and Wisniewski et al.46 The resistance to mass transfer in the stagnant fluid layer next to a membrane surface is an inescapable consequence of the membrane permeation process. Thus, during the drug release from a membrane-reservoir device, the drug concentration in the upstream diffusion boundary layer can be much lower than that in the adjacent drug reservoir due to the fast transport of drug across the membrane, whereas the drug concentration in the downstream diffusion boundary layer can become much higher than the bulk concentration in the releasing medium due to an insufficient drug removal rate. The

DRUG

DELIVERY

SYSTEMS

SERIES

SYSTEMS

5A

boundary layer effects can alter the rate or even the kinetics of drug release from drug delivery devices depending on the type of device and the environment of use. The frequently observed discrepancy between in vitro and in vivo release rates can generally be attributed to this type of phenomenon. The influence of boundary layer on the release kinetics of monolithic devices has been analyzed by Roseman and Higuchi51 and Roseman52 using a pseudosteady-state approach. In the case of membrane-reservoir systems, the

boundary layers offer additional resistance to mass transfer across the membrane as if the effective membrane thickness has been increased. The steady state release rate from such a membrane device with saturated Similar

membrane centration.

reservoir reduction

devices Several

would in release

therefore be reduced. rates is also expected

with nonconstant approaches were

reservoir proposed

in

conin the

literature, mostly in the area of membrane dialysis, to elucidate the mechanism of this boundary layer effect and to make quantitative calculations of the true intrinsic membrane transport parameters. They either suggest a reduction or elimination of the boundary layer by increasing fluid turbulence, e.g., by stirring or an estimation of the boundary layer resistance by performing transport experiments at different membrane thickness or stirring speed. Estimation of a hypothetical boundary layer thickness has also been attempted. The less permeable polymer or drug selected should still provide sufficient drug release rate to meet the therapeutic requirement. Sometimes this can be difficult to achieve in practice due to the large release rate requirement. In this circumstance, compromises in the membrane permeability, thickness, and area would have to be made in order to

minimize

the

contribution

of boundary

layer

effect

and still maintain the desired rate of release. Aside from the material and system parameters discussed previously, other factors, such as the temperature and membrane porosity, may also affect the rate of drug release. In diffusion controlled release systems, the transport of solute through the polymer is achieved by molecular diffusion due to concentration gradients. Depending on the molecular structure of the polymer, these systems may be classified as porous and nonporous. Porous controlled release systems contain pores large enough so that diffusion of the solute is accomplished through water which has filled the pores of the polymer. These pores are usually of size greater than 200 to 500 A. In the lower limit of this range, hindered diffusion may occur. Therefore, cor-

13

RANADE

rection of the solute diffusion coefficient may have to be made to account for pore wall effects.47’48 Nonporous systems contain “pores” of molecular (solute) dimensions. Molecular diffusion occurs effectively through the whole polymer and the solute diffusion coefficient refers to the polymer phase.

The

macromolecular

structure

of the

polymer

af-

fects solute diffusion according to theoretical analyses. Some of the polymer parameters controlling the solute diffusion coefficient are: degree of crystallinity and size of crystallites, degree of cross-linking, and swelling and molecular weight of the polymer. Many swollen, porous polymer systems retain the main characteristics of the porous structure, so that solute diffusion occurs simultaneously through

water-filled per se.

pores

and

through

the

swollen

polymer

In reservoir (membrane) systems, the bioactive agent is usually enclosed at relatively high concentrations between two semipermeable membranes and placed in contact with a dissolution medium (water or other biological fluid). The bioactive agent may be solvent-free or in the form of a concentrated

solution. The partition coefficient describes thermodynamic rather than structural characteristics of the solute/polymer/solvent system. It is rather easy to determine experimentally, and it is a measure of solute

solubility

derivation

in

of the

a swollen

partition

polymer.

A rigorous

(distribution)

coefficient

is presented by Lightfoot.49 In matrix (monolithic) systems, the bioactive agent is incorporated in the polymer phase either in dissolved or in dispersed form. Therefore, the solubility of the solute in the polymer becomes a controlling factor in the mathematical modeling of these systems. When the initial solute loading is below the solubility limit, release is achieved by simple molecular diffusion through the polymer. However, when the solute loading is above the solubility limit, dissolution of the solute in the polymer becomes the limiting factor in the release process. Park et al37 have listed examples of diffusion-controlled reservoir and matrix devices.

4. Chemically Chemically meric trolled tion of from a

Controlled

Release

controlled systems include all polyformulations where solute diffusion is conby a chemical reaction such as the dissoluthe polymer matrix or cleavage of the drug polymer backbone. In most of the chemically

controlled

systems,

the

geometric

14

#{149} J Clin Pharmacol

shape

solute of the

release device.

1991;31:2-16

is controlled Depending

by on

the

type of degradation reaction, these systems may be classified as chemically degradable (e.g., by hydrolysis) or biodegradable (e.g., by enzymatic reaction) controlled release systems. In chemically controlled drug delivery systems, the release of a pharmacologically active agent usually takes place in the aqueous environment by one or more of the following mechanisms: (1) gradual biodegradation of a drug-containing polymer matrix; (2) biodegradation of unstable bonds by which the drug is coupled to the polymer matrix; and (3) diffusion of a drug from injectable and biodegradable microbeads. In contrast to mechanical and osmotic devices, the main advantages of such biodegradable systems are the elimination of the need for their surgical removal, their small size, and potential low cost. On the other hand, all biodegradable products as well as their metabolites must be nontoxic, noncarcinogenic, and nonteratogenic. These requirements are not easily met and must be subject to careful scrutiny. In a system of type 1, the drug is either dispersed in the biodegradable polymer matrix or encapsulated in it, from which it is released into the sur-

rounding

biological

environment

by

controlled

rates. The particular kinetic behavior depends on the chemical composition of the polymer, the solubility of the drug in the polymer, and preparative aspects of the polymer matrix. Gradual degradation of the polymer can be facilitated by either converting an otherwise water-soluble polymer into a water-insoluble one by cross-links that are nevertheless hydrolytically or enzymatically unstable, or by using polymers that can undergo main-chain cleavage by hydrolytic or enzymatic actions. As is noted above, it is essential that none of the biodegradation products be toxic, carcinogenic, or terato-

genic. Furthermore, all degradation products must be fully metabolized and excreted without excessive or permanent accumulation in the tissues. These requirements pose formidable challenges, especially when they must be combined with drug-release parameters. The discussion on miscellaneous controlled release, survey of controlled release products and recent advances will be reported in the Part 5B of this review article.

CONCLUSION The purpose of this review article has been to review oral controlled release systems, with particular emphasis on the practical aspects of testing and fabri-

DRUG

DELIVERY

we have products

had commercial for more than

oral sustained three decades,

short 15.

retheir

ability to do more than simply prolong drug levels in the bloodstream had been very limited. Given the interest level for oral controlled release systems and associated basic research in the above mentioned aspects in particular, these limitations may appear transient. The oral drug delivery, nonetheless, will certainly continue to exist as one of the major dosage forms in the years ahead.

16. netics, trolled

Banker GS: Pharmaceutical applications of controlled release: An overview of the past, present, and future, in Langer RS. Wise DL (eds): Medical Applications of Controlled Release Vol. II, CRC Press Inc. Boca Raton, 1984; 1-34. U.S.

Patent

3. Anon: Oral Bull 1984;22:57-60.

sustained

4. Shaw JE: nal Chemistry 302-315.

Delivery Vol. 15,

5. Zaffaroni Pharmacology, Kirjapaino 6.

Drug

1956.

products-Pro’s

A: Proceedings Vol. 5. Oy Helsinki,

Heilmann

2,738,303,

Systems Academic of the Tuomisto 1975; 53.

K: Therapeutic

and

con’s.

in Annual Press Inc.

Reports New

Sixth International J, Paasonen MK

Systems-Pattern-specific

ery: Concept and development, G. Thieme, 7. Yates FE, Benson H, Buckles J, Urquhart Advances JF III feds): 8. Urquhart Pharmacology, Kirjapaino

in Biomedical Academic

Engineering, Press, New

J: Proceedings

of the Vol. 5, Tuomisto Helsinki, 1975; 63.

Oy

9. Zaffaroni A: Marcel Dekker,

Drug New

Metabolism York, 1978;

10. Davis SS: Evaluation lease Characteristics of VCH Publishers Chichester, Technological advances

Stuttgart,

J, and

Vol. 5, Brown York, 1975; 1. Sixth

International MK

J, Paasonen Review 191.

Vol.

Drug

Ther

Congress of (eds): Forsan

deliv-

Zaffaroni JHU,

A:

Dickson

Congress (eds): Forsaan

8, DiCarlo

of

F fed):

of the Gastrointestinal Transit and ReDrugs, Johnson P. Lloyd-Jones JG (eds): 1987; 164-179. Also see Ganderton D, in oral drug delivery, 1987; 150-163.

11. Ho NFH, Merkle HP, Higuchi WI: Quantitative, mechanistic and physiological realistic approach to the biopharmaceutical design of oral drug delivery systems. Drug Development and dustrial Pharmacy 1983;9:1111-1184.

In-

12. Antonin H-KH, Bieck P. Scheurlen M, Jedrychowski M, Ma!chow H: Oxprenolol absorption in man after single bolus dosing into two segments of the colon compared with that after oral dosing. Br I Clin Pharmacol 1985;19:1375-142S. 13. Bruck in Controlled CRC Press 14.

John

DRUG

SD:

Pharmacological Drug Delivery Inc, Boca Raton

VA,

Shotton

DELIVERY

PA,

Basis of Controlled Vol. I. Basic Concepts, FL, 1983; 1-13. Moppert

SYSTEMS

J, Theobald

SERIES

Drug Delivery SD Bruck fed): W:

Gastrointesti-

The

use

of scintigraphic

methods

Peppas

NA:

Mathematical

models

in Langer RS, Wise Release Vol. II, CRC

for

volunteers:

for

evaluation

A), Stella Problem,

controlled

DL (eds): Medical Press Inc., Boca

VJ Hu-

release

Applications Eaton, 1984;

ki-

of Con169-187.

From Edible Slabs, CylinPolymeric Formulations. Chemical Society Sympo-

Lee P1: Diffusional release of a solute from approximate analytical solutions. I Membr

Baker

the

Repta

A

a polymeric Sci 1980;7:255.

ma-

JA, Olanoff LS: Pharmacokinetic Modeling and Bioin Medical Applications of Controlled Release Vol. II Wise DL (eds): CRC Press Inc. Boca Raton, 1984; RW,

Lonsdale

rates, in Tanquary logically Active 21. Notari RE: Marcel Dekker

AC, Agents,

HK:

Controlled

Lacey Plenum

RE

release:

Mechanisms

feds): Controlled Press Inc., NY.

Biopharmaceutics NY, 1975.

and

22. Schneider H, Nightingale CH, Evaluation of an oral prolonged-release Pharm Sci 1978;67:1620.

and

Release 1974; 15.

Pharmacokinetics,

Quintrilliani antibiotic

of Bio2nd.

ed.,

R, Flanagan formulation,

DR:

J

23. Beckett AH, Slipper IA: The importance of dissolution test media in in vitro in vivo correlations of formulations of disopyramide, in FIB 81 Abstracts. Programme of Short Communications, 41 st. mt Cong Pharm Sci Vienna Austria. 1981; 133.

in MediciYork, 1980;

drug 1978.

SS:

19. Anderson availability Langer RS, 189-219.

1.

RH:

Davis

systems in healthy 1985;19:203S-2065.

17. Hopfenberg HB: Controlled Release ders and Spheres in Controlled Release Paul DR, Harris FW feds): American sium Series, Washington, DC, 1976;33:26.

20.

Blythe

of Oros drug delivery Br J CIin Pharmacol

report.

of novel delivery systems, in Borchardt RT, feds): Directed Drug Delivery: A Multidisciplinary mana Press, New Jersey, 1985;319-340.

18. trix;

REFERENCES

2.

5A

nal transit

cating these systems. The GI physiology, its mechanisms and factors influencing drug absorption and GI transit time etc. have a profound influence on the design of oral controlled drug delivery systems. While lease

SYSTEMS

24. Turi P. Dauvois M, Michaelis AF: Continuous dissolution determination as a function of the pH of the medium. J Pharm 1976;65:806.

rate Sci

25. Good WE. Lee P1: Membrane-controlled reservoir drug delivery systems, in Langer ES, Wise DL feds): Medical Applications of Controlled Release Vol. I, CRC Press Inc. Boca Raton, 1984; 1-39. See also Wagner JG, Biopharmaceutics and Relevant Pharmacokinetics, Drug Intelligence Publications, Hamilton. IL, 1971. 26. Harris MS: Preparation and release kinetics chloride microcapsules. I Pharm Sci 1981;70:391.

of

27. Bakan JA: Microcapsule RL, User Z, Martin E (eds): Plenum Press, NY, 1975.

in Kronenthal and Surgery.

28.

Sears

29.

Shank

BD:

U.S.

Patent

IL: U.S.

Patent,

4,145,410.

32.

Theeuwes

33.

Theeuwes

PV: Enzyme-mediated

II.

F, Higuchi F:

1977.

H, Takenaka drugs

J, Trescony

systems, Medicine

1979.

4,001,480,

30. Kawashima Y. Ohno matrixes of prolonged-release Pharm Sci 1981;70:913. 31. Heller dissolution, 1979;68:919.

drug delivery Polymers in

H: Preparation from liquid

Controlled

drug delivery

T: U.S.

Elementary

potassium

of spherical suspension,

release device,

Patent,

3,916,899,

osmotic

pumps.

by

polymer

J Pharm

Sci

1975.

J Pharm

Sci

1975;64:1987.

34.

Park

Johnson Ridge

JC: Sustained NJ, 1980.

35. Chandrasekaran Controlled Drug demic Press, NY.

Release

SK, Delivery 1979;

Theeuwes Systems,

Medications, F, Drug

Noyes

Yum SI: The Design Vol.

Data

Corp.,

Design of VIII, Aca-

133-167.

15

RANADE

36.

Hendeles Intell

J: Drug

L, Bighley Clin Pharm

L, Richardson

R, Heples

CD, Carmichael

1977;11:12-18.

37. Park K, Wood EW, Robinson JR: Oral Controlled Release Systems, in Langer ES, Wise DL (eds): Medical Applications of Controlled Release Vol. 1, CRC Press Inc., Boca Raton, 1984; 159-201. 38. Michaels AS, Wang PSL, Prather R, Gale EM: American Institute of Chemical Engineers Journal, 1975;21:1073. 39. Good WE, Mueller KF: Controlled Release rials, Baker R (ed): Academic Press, NY, 1980; 40. Good WE: Polymeric fed): Midland Macramolecular 1978,

Delivery Systems, Monograph.

Mate-

Vol. 5, Kastelnick Gordon & Breach,

Release of Bioactive NY, 1980; 135.

Materials,

Lee P1: Proceedings of the 9th International trolled Release of Bioactive Materials, Controlled 1982; 54.

42.

43.

Mueller

KF,

44.

Yasuda

H, Lamaze

Heiber

SJ, I AppI

CE,

Polym

Peterlin

1971;9:1117.

#{149} J Clin Pharmacol

1991;31:2-16

Sci

Baker

NY, R fed):

Symposium Release

Con-

Division

Polymer

of Polymer

Preprarations.

American

Chemistry,

Chemical

Soci-

1974;15(2):387.

46. Wisniewski SI, Gregonis DE, Kim SW, Andrade JD: Hydrogels for Medical and Related Applications, Andrade JD (ed): American Chemical Society, Washington, DC, 1976.

47.

Ginzburg

48.

Heliferich

49.

Lightfoot

BZ, Katchalsky

A: I Gen

F: Ion Exchange. EN: Transport

Physiol

1963:47:403.

McGraw-Hill, Phenomena

and

NY, 1962; Living

339.

Systems.

John

Wiley

& Sons. NY, 1974. 50. Zentner GM, Cardinal JR. tive through polymer membranes fect on progesterone permeation Pharm Sci 1979:68:794.

52. Sci,

RYS:

51. Higuchi T: bases containing

Society,

1982:27.

A: I Polymer

Chen

ety

R

139.

41. Lee P1: Controlled Academic Press, Inc.

16

of Bioactive 155.

45.

Part

A,

Flynn

nomena 1974:63:479.

CL, and

Gregonis DE: Progestrin permeaIII Polymerization solvent efthrough hydrated membranes. I

Rate of release of drugs in suspension. Yalkowski models:

SH,

medicaments I Pharm

Roseman

Theoretical

TJ: concepts.

Sci Mass

from 1961

ointment ;50:874.

transport

J

Pharm

pheSci