PHYSICAL STABILITY TESTING OF DRUGS AND DRUG PRODUCTS
Sometimes a cloud will appear in a product as the storage time progresses, and this is most often due to chemical changes in the system. If for instance an ester (e.g., polysorbate, which is a fatty acid ester) hydrolyzes, then the produced acid may be poorly soluble. If the solubility is denoted S, then the following holds: If the reaction in general is written
Where A is a drug of initial concentration A0 and is the decomposition product with ility S (which is assumed to be limited.
Arrhenius plotting. Such plotting is quite predictive, The precipitation may also occur by the solubility product being exceeded, or from any situation leading to a product with limited solubility. There are other causes for precipitation on storage, one being the original use of a metastable form, so that the solutions in question, in fact, are super- saturated solutions. It was the author's experience, at his tenure at Hoffmann-la Roche in 1965, that a product to be introduced (Taractan Injectable) was in this category. Several pilot batches had been successfully made, but the first production batches precipitated, a more stable polymorph crystallizing out, This necessitated reformulation to a lower strength (corresponding to the lower solubility of the stable polymorph) and subsequent resubmission of data to the FDA. This points out the importance of careful preforrnulation studies of the solubility of compounds. Errors of the above type are costly, both in terms of resubmission and in lost market time. Even official products fall into this category.
The viscosity of these agents is often Ingham bodies, i.e., they posses a yield value. The correct way of checking the is, therefore, with e.g. cup and bob viscometer, so that a rheogram can be drawn. In this fashion it is possible to check both changes in yield value and slope of the rheogram (apparent viscosity). For very liquid solutions (dilute aqueous solutions) this is difficult, and most often it is best followed by the use of an stwald-Fenke pipette two pipettes (with different flow should flora time should be used in is case, because the once in the measured vise erasure of the yield value (although calculation of the yield value from the difference is a priori not not yield value and apparent viscosity are functions of concentration t al., 1980); in a multicomponent system~ there will usually be one moment responsible for viscosity, and it is the breakdown of this one compound that would be of importance. often when drastic occur in viscosity, bacterial contamination can be suspected. precipitation is tied into solubility, as seen in the foregoing.
(b) ORAL SOLUTIONS
The main types of changes in appearance of oral solutions (syrups,elixirs,etc) are loss of dye, precipitation, and bacterial growth. Precipitation has already been dealt with to some degree, but some cases particular to oral solutions will be mentioned. Change in dye content will be treated below.bacterial growth will be treated separately.
3. PHYSICAL STABILITY OF DISPERSE SYSTEM
Disperse systems are suspensions and emulsions. The rationale for the Physical tests carried out on these will be discussed below.
It would be desirable to have a suspension that did not settle (and there are such suspensions), but the general rule is that a suspension will settle, and therefore there are two parameters that are followed in this respect, namely sedimentation rate and sedimentation volume. When the sedimentation volumes are small, then there is a tendency for the suspension to cake, and hence various types of shaking tests are carried out. Tests can be purely subjective, in that a tester notes that e.g. the suspension after three months’ storage at 25°C was “difficult to resuspend, leaving some cake at the bottom,” Such subjective tests should always be included in a pro but more quantitative means are desirable also. Atypical quantitative test is to rotate the bottle under reproducible conditions. The type of setup used for solubility determinations is a good type apparatus for this purpose.
One way of accelerating the settling is to place the suspension product on a shaker at e.g. 37°C. This makes particle movement more rapid and allows the fine particles to slip into the interstices of the larger particles, hence promoting a close packing. This can then be used to judge qualitatively whether caking will take place. It might be thought that centrifugation would be a good way in which to ‘‘accelerate’~ sedimentation, and the Stokes law indeed predicts this. However, it gives only an acceleration of the “initial settling rate,” and the further settling, and the caking phenomena in which the formulator is interested, are not well predicted by this method. Some caking is due to crystal growth, and this is accelerated by the use of freeze-thaw tests, i.e., alternating the temperature every 24 h from e.g. 25°C to -5°C (or some other low temperature above the freezing point of the product). The tem- prelature cycle will promote crystal growth, and the effect of this on the product can be assessed. The freeze-thaw cycle has the advantage of emulating (and overstating) some real conditions to which the product could be exposed during shipping. Zapata et al. (1984) have described the effect of freeze-thaw cycles on aluminium hydroxycarbonate and magnesium hydroxide gels. Coagulation after freeze-thaw cycles led to the formation of aggregates that were visible. These aggregates were particles in a primary minimum, and these were only reseparable by ultrasonic treatment. The freeze-thaw cycle affected content uniformity of both the gels, but the treatment did not alter the surface characteristics or the morphology (as judged by x-ray powder diffraction). It cause a reduction in the acid neutralization rate, and the rate of sedimentation increased. The effect was pronounced after the first cycle (and indeed most of the effect occurred at this point). The duration of freezing was not important, but the aggregate size grew inversely with the rate of freezing. The use of polymers in the suspensions reduced the effects of the freeze-thaw cycle.
Freeze-thaw cycles (aside from being a stability monitoring tool) can be used to screen products as well, the best of a series of suspensions or emulsions being the one that stands up best to the test. This on the surface may be logical, but without a theoretical basis it is difficult to judge the generality of such a statement.
B. SEDIMENTATION VOLUME
If a suspension is particulate, then the particles will (approximately) settle by a Stokes law relation, i.e., the terminal velocity, v, is given by
where the constant g is gravitational acceleration, A p is the difference in densitybetween solid and liquid, q is the viscosity of the liquid, and d is the diameter of the particle. The final apparent volume of the sediment, provided it is monodisperse, would be given by the fact that in cubical loose packing a sphere of diameter d will occupy the space of its confining cube, i.e., the sedimentation volume will be
where n is the number of particles per cm3 of suspension. Since their density is p g/cm3, then (denoting the dosage level Q g/cm3) the following holds:
so that, solving for n,
which inserted in gives
Hence it becomes difficult to separate them, and the precipitate becomes a cake. This would prevent redispersion by shaking and would make proper dispensing impossible. It is a for- mulation goal to prevent this from happening, and this is done by adjusting the zeta potential, as will be discussed shortly. From a formulation point of view, it is better to have the particles at larger distances, e.g., in the secondary minimum occur- ring at longer distances. Adiscussion of the connection between caking tendency and the so-called zeta potential is beyond the scope of this book. Suffice it to state the following: particles are suspended in a liquid, they acquire a charge (and the liquid acquires a similar opposite charge, to maintain electroneutrality). The zeta potential is related to this charge, and caking is prone to happen if the charge potential is outside a range of -10Mvto +10 mV. If the zeta potential is high it can be lowered by the addition of negatively charged ions. Highly valent ions (e.g., citrate) are preferable. On the other hand, if the zeta potential is low, then it can be increased by the addition of positively charged ions (e.g., aluminium ions). The zeta potential is measured with a zeta meter, In this the particles are placed in an electrical field (between two electrodes, the voltage of which can be adjusted), they are tracked under a microscope, and their velocity is determined. The relation of velocity to voltage allows determination of the zeta potential. It is worthwhile occasionally to check the zeta potential in a stability check of suspension (and emulsion) products.
The zeta potential is close to zero, the suspension will be flocculated, i.e., the particles are positioned in the secondary minimum. The floccules are large and hence settle more slowly, but on the other hand the sedimentation volume is large. Since the particles are in the shallow minimum (small potential, i.e., easy to disrupt), they are easy to resuspend.
There are suspensions that do not settle. Here the yield value of the suspension is so large that the gravitational force does not exceed it. In this case it is very important to carry out complete rheological profiles at different time points in the stability program, to insure that the yield value is not changing. In such a system the yield value is a function of the solids content and the viscosity of the medium. If the viscosity imparting substance deteriorates, or if the flocculation characteristic (the “diameter” of the particles) changes, then the yield value may change, and what originally was not prone to cake might at a later time have such a propensity. It has been stated elsewhere that for Ingham bodies, a yield diameter of the bottle can be calculated and below this bottle diameter there will be no settling.
C. TEMPERATURE TESTING OF DISPERSE SYSTEM
A suspension is, as the name implies, a two-component system consisting of a solid and a liquid phase. (Gas phases are considered nonessential in this connection). Obviously, the solubility of the compound is a function of the temperature, and at a given temperature above 25°C this solubility will be reached. Testing about this temperature obviously has no meaning as far as suspension stability (neither physical nor chemical) is concerned. Prior to starting a program, this temperature should be established, so that unnecessary sampling stations can be avoided.
D. SEMISOLID SUSPENSION SYSTEM (OINTMENT, SUPPOSITORIES)
Some semisolid systems (ointments and suppositories) are suspensions. Their testing is not different, in general philosophy, from what is described above, except that the archeology is checked differently.
The factors checked for in stability programs of such products are the following:
1. Consistency fell to the touch
It is mentioned elsewhere that migration of a “disperse” phase within a semisolid product is quite possible when another phase is present. This situation may occur in the case of the use of benzocaine in, for instance, a suppository wrapped in aluminum foil coated with polyethylene. Polyethylene lining of aluminum wraps of suppositories is used to prevent contact between the metal and the suppository, and in most cases this has a positive effect. However, a partitioning of drug or additive between the two phases may be possible if the drug or additive is suspended in the suppository. Denoting its solubility in the polyethylene S, and the solubility in the suppository base Ss, the compound would disappear from solution in the suppository at a rate proportional to Sp - Ss, and “disappeared” compound would be replenished by dissolution from the solid phase. The rate of disappearance would be governed in that the value of Sp would increase by a sigma minus relation (i.e., in the same manner as the appearance of decomposition product in a first-order reaction), and this then would be the over- all “loss” of compound as a function of time. Since this is a first-order overall relationship, the “decomposition” would, initially, appear to be first order.
E. OINTMENT AND TRANSDERMAL
Polymorphism can be followed by x-ray analysis and in some cases by thermal methods. There is, in fat systems, the possibility of trans etherification, and this can be tested for chemically. The problem of morphology changes is often of particular importance and of particular frequency in the case of suppositories. In this type of product, it is also important to check for migration of suspended/dissolved substances. Often a sub- stance is added to a suppository as a suspended particle, which is soluble in the suppository base to some extent. The phenomenon of dissolution will, of course, become evident by checking the particle size as a function of time. If a substance is soluble in the base, then it is preferable (if possible) to saturate the base with it at the onset. For this reason it is necessary to determine the solubility (S gm/gm) of the drug (or other) substance in the base. A Van’t Hoff plot [solubility as a function of temperature (T’K), i.e., plotting ln [q versus 1 /U will allow extrapolation to room temperature. In manufacturing it is advisable to dissolve the drug (or other substance) to the extent of its solubility during the intermediate temperature phase of manufacturing (where the preparation is still quite fluid) and then suspend the rest at a lower temperature. An example is ascorbic acid, which is a good antioxidant in Carbowax bases. To exert its antioxidant action it must, however, be dissolved (and it is quite soluble in polyethylene glycols). Dissolved drug (or other substance, e.g., benzocaine) will diffuse in the suppository base, and can, for instance, partition into polyethylene linings of the suppository wrap.
Release rates are important in many topical preparations in particular in transdermal preparations. Here there are several investigational methods available.
In-vitro methods involve placing the ointment on a membrane and Measuring the appearance of drug in a receptor compartment on the sink side of the membrane. oelgaard and Mollgaard (1983) have, for instance, described the in-vitro release of linoleum acid through an in-vitro membrane. They mounted abdominal human skin in one case and skin from hairless rats in another to open diffusion cells. The dermal side was bathed with a rece tor medium stirred at 37°C. The medium was 75 mL, of 0.05 N phosphate buffer ”-7.4) which contained 0.05% ic F68 and 0.01% butylhdroxytoliene, alter two ingredients added in to increase the lipid solubility. Linear, Fiction diffusion curves were obtained. In a stability program, such tests are obviously useful and should be repeated periodically, but an “internal standard” or “calibrator” should be use i.e., a stable test substance, the diffusion of which is known (e.g., salicylic acid). their pseudo-in-vivo methods involve shaved or hairless rabbits, or cadaver skin. The interaction between ointment and container (patch) should also be part of the stability program. Some of the testing applicable to semisolid emulsion systems is also applicable to ointment systems and will be discussed at a later point.
NOW YOU CAN ALSO PUBLISH YOUR ARTICLE ONLINE.
SUBMIT YOUR ARTICLE/PROJECT AT email@example.com
FIND OUT MORE ARTICLES AT OUR DATABASE