"CONTROLLED RELEASE COMPOSITIONS FOR ADMINISTRATION OF THERAPEUTIC AGENTS TO RUMINANTS"

This invention relates to improvements in controlled-release compositions for administration of therapeutic agents to ruminants.

Our Australian patent application No. 35908/78 describes a controlled-release intraruminal device (IRD) intended for use in the administration of a wide range of biologically active, powdered, therapeutic agents. In that device (which is described in more detail herein¬ after) , a therapeutic agent of choice is suspended in a cylindrical matrix of surface active agents which is solid at the rumen temperature (39°C) . The cylindrical matrix is placed in the cylindrical barrel of an IRD with one end face held against an orifice at one end of the IRD by the action of a spring and plunger at the other end face. A pair of plastic arms extend from one end of the IRD at 90°C to the axis of the IRD for the purpose of retaining the device in the reticulorumen. Fo,r oral administration, these arms are held against the barrel of the IRD by the action of water dispersible or disruptable tape.

As described in application No. 35903/78, the matrix is prepared by melting the surface active agents, incorporating the therapeutic agent in the molren material and casting the matrix core either directly in

OMPI

the barrel of the IRD or in a cylindrical mould for transfer to the IRD after solidification.

The selection of suitable surfactants and mixtures for use in our earlier invention is dependent upon (i) the melting range and (ii) properties determining the rate of dissolution at the orifice of the IRD. In most instances, this form of matrix production has proved satisfactory, but, since 1978 an important series of failures of. this technique has stimulated further development leading to this present application.

The therapeutic agent "Elfazepam" (Smith. Kline Animal Health Products) dissolved in most matrix surfac¬ tants above 60 C, particularly the preferred surfactant at that time, Teric 18M2 (ICI Australia Ltd.) and recrystallized on cooling giving massive crystals. The resultant dissolution rates were very variable and quite unsatisfactory. The flukicide "Diamphenethide" (Wellcome Ltd.) also recrystallized from Teric 18M2 to give a mat of long, fine crystals which caused the dissolution of matrix to cease shortly after insertion when the crystals formed a felted mass at the dissolution orifice. A developmental therapeutic agent reacted chemically with Teric 18M2 at 60°C and the resultant product remained liquid on cooling to room temperature. These problems limit the potentially wide-ranging applicability of our earlier invention and requires alternative means of matrix production.

We have now found that very satisfactory matrix cores can be made, even with the troublesome therapeutic agents mentioned above, by mixing the selected therapeutic agent with a powdered support or carrier material and pressing the powder mixture in a cylindrical die to give a tablet or bolus.

- OR-E OMPI

- ₃ -

A number of powdered materials have been tested and found to be useful as the carrier material. While it is not presently possible ^' to specify the suitable materials generically in orthodox physical or chemical terms, the following limitations must be observed in selecting the carrier.

(i) It must have a softening point in excess of 39 C. ( _. ii) It must be, or be capable of being milled to, a non-tacky free flowing powder in which 90% of the particles are less than 100 μ (200 mesh sieve) , and which can be freely mixed with a therapeutic agent to give uniform, intimate and reproducible mixtures, which together with such bonding and slip agents -as- are necessary, are capable of forming a coherent tablet or bolus in a die under pressure, (iii) It should preferably be a surface active material or emulsifying agent, or a material which behaves as such. (iv) It must be generally regarded as safe by the appropriate regulatory authorities; preferably with clearance as a food grade material, (v) The resultant core must show the required controlled release characteristics in the IRD of reproducibility of output, linearity of output, and required lifetime or release rate.

Condition (v) is met by selecting a carrier material which when tableted and immersed in an aqueous medium absorbs water at the tablet/water interface, to produce swelling of the material and a resulting water gel. This behaviour ^" is accompanied by marked changes in rheologiσal properties, particularly critical yield stress and visco-elastic properties of the material. There is rapid reduction in viscosity and critical yield stress, due to water absorption up to a particular water

content at which the gel forms, after which further water .absorption results in only a gradual reduction of viscosity.

There is thus a sharp discontinuity in rheological properties at the critical water content and provided this discontinuity occurs at a pressure within the pressure range exerted by the drive spring of the IRD and at a water content which, can be achieved in a portion of the core material which is in close proximity to the release orifice of the IRD, this phenomenon allows the soft water gel structures to be extruded froin the capsule orifice by the action of the spring.

Additionally, water gels reduce the rate of water penetration into the matrix and thus an equilibrium is reached between the formation of the low viscosity gel and the driving spring force. The overall result of this combination of parameters is a high precision of release, and extrusion of the matrix under the influence of the drive spring is sharply limited by the abrupt rheological discontinuity.

Selection of surfactant materials having the above properties can be easily made on the basis of simple experiments, details of which are given herein- after.

As an example of one particularly satisfactory matrix material, we have found that the sucrose esters described in Dai-Ichi Kogyo Seitaka Co. Ltd., Tokyo, Japan and prepared by the Nebraska-Snell process (USP. 3,480,616) , particularly the monostearate ester,

"DK-Ξster-F160", made excellent matrix cores, by the

OMP

- 5 -

method described above, with all of the problem materials described in the previous paragraph and all other therapeutic agents used in our research.

This development can be contrasted with the process described in application No. 35908/78 because (i) _, -the surfactants described in that application are not powders but are waxy materials unsuitable for compressed tablet formation (.ii) the sucrose esters used in this invention melt with decomposition and are unsuitable for cast core formation, and (iii) DK-ester- F160 has an HLB of 15, well above the limit of 8.5 given in our earlier application. This last point is the first of a series of unexpected discoveries arising from our investigations of the use of sucrose esters. Surface active agents of the type discussed in application No. 35908/78, (i.e. nonionic ethyiene oxide adducts of fatty acids and fatty amines) but with an HLB near 15 would dissolve under the conditions of the rumens of sheep and cattle so rapidly that an IRD based on them could not be expected to last for more than a few hours, whereas IRD's based on DK-ester-Flβ 3 can last for considerably more than 100 days, as shown in following examples. At this time we do not have an explanation for this extraordinary difference. Apart from DK-ester-F160, we have also found that other sucrose esters produced by the Dai-Ichi Company, which range in HLB from 1 to 15, can be used alone or in combination to produce compressed powder cores, in accordance with this invention, for use in IRD's.

Our investigations have also shown that other classes of materials are suitable for use in the invention. For example, we have found that milk powder is a suitable support agent for very satisfac ^t ory cores

/00094

— o —

of 6 - 10 day lifetime, at the present stage of -development.

Another class of material which has given satisfactory results in the polyglyσerol stearates, specifically the food-grade product CRESTER KZ (Croda) .

The invention is further described and elucidated by reference to the accompanying drawings in whic :—

Figure 1 is a cross sectional view of a controlled release in ra-ruminal device (IRD) ,

Figure 2 is a graph showing the variation in critical yield stress of DK Ester F160 with water content;

Figures 3 and 4 are graphs showing the in vivo rate of release of material from IRD's containing a core in accordance with the invention.

The results shown in Figure 2 were obtained by the following method.

Varying quantities of water and DK Ester F160 were weighed into vials to make 20g samples containing from 2-20% water in 2% increments. These mixtures were thoroughly stirred, sealed and heated in a water bath at 60 C until all the trapped bubbles had been expelled.

The samples were then transferred to a water bath at 39 C and equilibrated for 1 day. The viscosity was then measured by following the output of a 2kg strain gauge on a chart recorder, where the strain gauge measures the resistance exerted by the mixture to a 1/8" ( ^" 3.2 mm) diameter hemispherical-ended plunger moving at

a velocity of 0,021 mm sec ^" . The measured force used is when either a maximum is reached or steady state achieved.

The results plotted in Figure 2 clearly show the marked discontinuity in critical yield stress which occurs at a water content of about 4.5%.

Figure 1 is a cross sectional view of -a variable geometry device as disclosed in our afore¬ mentioned application No. 35908/78. The device 1 comprises a tubular body 2 having an opening 3 at one end, which opening is restricted by resilient projections 4. Preferably the body has a length of 14 cm. and a diameter of 2.8 cm. for cattle and a length of 9. cm. and a diameter of 1.6 cm. for sheep. The projections 4 protrude inwardly from said one end of the body. The other end of the body is closed. The body contains a cupped plunger 5 which is capable of sliding longitud¬ inally thereof. The plunger is biased by means of a helical spring 6. The helical spring is made from spring steel wire having a circular transverse cross section of 0.5 mm. in diameter. The spring comprises 20 to 30 coils and is capable when fully compressed of exerting a pressure of approximately 100 grams per sq. cm. The body has two resilient arms 7 attached thereto at said one end, which arms project outwardly from said body at an angle of approximately 45° to the side thereof to form a first configuration. In the first configuration the device has the shape of an arrow-head. The arms 7 are capable of being resiliently flexed about an axis corresponding approximately with the junction of the arms with the body, to form a second configuration in which the arms are substantially parallel to the length of the body as shown by the dotted

2/00094

- 8 -

lines in Figure 1. With the arms folded back into the second configuration the device is capable of being administered to cattle per os. As shown in Figure 1, the body contains a preformed matrix core 8.-'made in accordance with this invention and containing a therapeutic material. ^' The resilient projections 4 are sufficiently flexible to allow the precast plug to be inserted into the device but have sufficient rigidity to ^' - retain the plug within the device against the pressure exerted by the spring. Alternatively, a barrier preventing ejection of the plug by action of the spring can be applied after the plug has been inserted, e.g., a strip of polypropylene welded across the opening 3 of the body. The body 2, arms 7 and projections 4 may be integrally moulded from a suitable plastics material such as polyethylene polypropylene or nylon. By choice of the appropriate material of construction a device may be obtained which can be retained in the rumen ^' indefinitely or for lesser periods of time. For example, a device integrally moulded from low density, low molecular weight polyethylene will eventually fail after about 270 days in the rumen by flex cracking of the arms. On the other hand, a device integrally moulded from polypropylene is virtually indestruetable.

By choice of a suitable spring the core can be exposed to an urging force which just exceeds the critical yield stress of the core material at a water content near the discontinuity region (see Figure 2) thereby causing the material with the lowered critical yield stress to be slowly extruded and subjected to mechanical erosion.

O PI

Three unexpected results have been observed in our experiments to date using the present cores in IRD's. The first, as reported above, was the unexpect¬ edly slow rate of dissolution of DK-ester-F160 which has an HLB of 15.

This is a desirable result for long acting IRD's. The other two unexpected results are related to controlled release and are illustrated in the examples which follow.

They are (i) IRD's containing pressed cores have given the most accurately uniform in vivo release rates obtained in our studies (examples 1 and 2) and (ii) between-animal variability, expressed as a standard error of the mean IRD release rate, is an order of ^' magnitude smaller than that usually observed for cast cores in IRD's (example 3) . ^"' •

EXAMPLE 1

In vivo release of compressed matrix from spring driven IRD.

The core was prepared from a spray cried powder CHESTER KZ (Croda) , described as a food grade polyglycerol stearate, with a HLB of 6.5 - 7.5 lOg was weighed into a 1.45 cm diameter dye and pressed with a hydraulic ram to a force of 3000 Ib.wt. on the core material, this pressure was sustained for 10 seconds, resulting in a tablet of density 1.08g/cm . This core v/as loaded into a 9.2 cm IRD with 7.9 mm orifice, and then loaded with a spring giving a spring strength of 162g at 75% compression. The spring chamber vas then sealed with a 9.0mm neoprene stopper to ensure that the

_mnnnΛ 2/00094

- 10 -

chamber was hermetically sealed.

The IRD was placed in the rumen of a fistulated bovine. The output, i.e., rate of release of material, from the IRD was measured by inspecting the IRD at intervals and measuring the distance travelled by the plunger. Figure 3- shows the results obtained. The output from the IRD was substantially linear over a period of.60 days.

O

- 11-

ΞXAMPLE 2

Uniformity of release rates in vivo. (Sheep sized IRD's)

Four groups of capsules were prepared as follows.

Group OP containing Oxfendazole (Syntex) 50% w/w in Teric 18M2 (I.C.I.A.L.) Group 0 containing Oxfendazole (Syntex) 50% w/w. in

DK-ester-F160. Group MP containing Monensin (Eli Lilly) 40% w/w in Teric 18M2.

Group M containing Monensin (Eli Lilly) 40% w/w in DK-e≤ter-F160.

Cores based on Teric 18M2 were prepared as - follows. Matrix components were mixed at 6Q°C, homogenized using a Silverson ho ogenizer for 5 minutes, and the mixtures were then evacuated at 60°C to remove entrained gases. Cores were poured at 60°C into moulds kept at 15°C. Cores were cut to a length of 5.25 cm and ranged in weight from 9 to 10g (Oxfendazole) and 9+ 0.15g (Monensin) . Cores based on F160 were prepared as follows. Matrix components were sieved together through a 35 mesh sieve, mixed on powder mixing paper, weighed into a cylindrical die and pressed at 7000 psi. For Monensin, lg tablets were pressed, 9 tablets being inserted per IRD. For Oxfendazole, 5g and 4g tablets were pressed, with one of each tablet inserted per IRD. Weight precisions were 9+ 0.04g (Monensin IRD's) and 9+ 0.02 (Oxfendazole IRD's) .

In the IRD's spring strengths at 75% compression were 171 to 181 g. Orifice diameter was

2/00094

- 12

7.9 mm and IRD barrel length was 8 cm. In vivo release of representative duplicate IRD's from the four groups was determined with rumen fistulated cattle, the results shown graphically in Figure 4 , being expressed as plunger travel (cm) with time in vivo.

EXAMPLE 3

Between-animal variability studies with sheep.

»

The remainder of the capsules prepared as described in Example 2 were dosed to 12 sheep (one capsule per group to each animal) . Capsules were recovered at slaughter on day 45 and analysed to provide the following data of measured loss of total matrix (g) .

Capsule group 0 OP M MP .

X (Mean matrix loss) g 2.57 2.45 5.32 ^" some

SΞM (Standard capsules error of mean) 0.065 0.772 0.066 empty

SΞM 0.026 0.32 0.013 X

Experiments of this design using cast cores in sheep IRD's gave results comparable with the OP group with one exception, when cast cores containing * Oxfendazole (8% w/w) in the surfactant mixture Span 60/ Tween 80 (60:40) (I.C.I. Australia Ltd.) gave the ratio SEM/X = 0.0231 (6 sheep).