BACKGROUND OF THE INVENTION AND PRIOR ART STATEMENT A general problem with transdermal delivery devices, including matrix devices, is loss of the material to be delivered, which herein will be called the active material, in storage. The primary cause of this loss is presumably chemical degradation.

Numerous examples of matrix transdermal delivery devices have been disclosed in the art, and many have been placed on the market, containing such drugs as estradiol and levonorgestrel. Some of these devices use a matrix derived from water-based acrylate adhesives. The devices typically have a backing layer on which the adhesive is held when in contact with the

skin and, on the other side of the adhesive layer, a release liner which has an area larger than that of the adhesive layer and is readily peelable by hand from the adhesive layer in order that the adhesive layer may be applied. A release liner consisting of a plastics film or a metallised plastics film, for example metallised polyester, with a silicone coating on the side towards the adhesive layer, has been used successfully. Such a film is thought to provide sealing of the adhesive layer against the atmosphere.

One proposal which has been made to reduce degradation of the active material is disclosed in WO 97/03629, where it is suggested that in a matrix device an adhesive containing carboxylic acid groups should be avoided.

SUMMARY OF THE INVENTION It is an object of this invention to provide transdermal delivery devices in which loss of the active material on storage is reduced or minimised.

According to the invention there is provided a transdermal delivery device having a layer of adhesive which contains material to be delivered transdermally and is to be placed in contact with the skin in use of the device, a backing layer on one side of the adhesive layer and on the other side of the adhesive layer a release liner which is easily removable from the

adhesive layer for use of the device, wherein at least one of the release liner and the backing layer is porous and the material to be delivered transdermally comprises at least one steroid having a 4-en-3-one structure. A porous material is understood to mean a material though which gases are able to diffuse at room temperature and pressure, in the absence of a pressure gradient.

The invention also consists in a method of making a transdermal delivery device containing a steroid having 4-en-3-one structure, wherein a porous sheet is incorporated as at least one of the backing layer and release liner. Particular methods are described herein.

The invention further consists in the use of a porous sheet as a backing layer or release liner in the preparation of a transdermal delivery device, especially a medicinal or veterinary device, containing a steroid having 4-en-3-one structure, in order to achieve reduced loss of the steroid on storage, compared with a corresponding device having the same adhesive layer and gas-impermeable release liner and backing layer.

The inventor has surprisingly found that the use of a porous release liner and/or backing layer provides improved results in storage, i. e. less loss of

the active material. This finding applies to steroids having the 4-en-3-one (4-en-3-keto) structure and not for example to estradiol which does not have this structure. It applies even though there may be a silicone layer, or other material having low adhesivity to the adhesive layer, between the adhesive layer and the porous material. Paper can be regarded as porous, as compared with the non-permeable films used hitherto.

While porosity is considered to be a critical factor, the reason for this is not wholly clear at present.

A particularly surprising finding relating to the improved stability of steroid substances having the 4-en-3-one structure when a porous release liner and/or backing layer is used is illustrated by the results of Test 9. These show that permeation of oxygen which would be expected to result in oxidation of the drug to form substances which are inactive or are less biologically active than the parent drug, in a manner that could be prevented or reduced by the incorporation of chemical antioxidants into the drug-in-adhesive formulation, is not apparently a factor involved in the progressive loss of drug activity of steroids having the 4-en-3-one structure.

Since ammonia is known to be present in some of the types of acrylate adhesive systems used, or is added to the wet drug-in-adhesive formulation, it had

been thought that trace residual levels of ammonia effectively trapped by a non-porous release liner and/or backing layer may have been responsible for the instability seen when these types of components are used.

Experimental work has shown, however, that the ammonia content of the drug-in-adhesive products after drying is below the limits of analytical determination.

Because of this, it is considered unlikely, but is not excluded as a possibility, that traces of ammonia trapped in the drug-in-adhesive layer is a factor in the instability of steroids having the 4-en-3-one structure when a relatively non-porous release liner or backing layer is used.

The results of Tests 10-13 show that the use of some adhesive systems results in a further relative improvement in the stability of drug-in-adhesive on a paper release liner compared with others and further suggests that some other unidentified volatile component may be implicated. It is thus thought possible that a porous release liner and/or backing layer may allow volatile components of the drug-in- adhesive formulation to leave the adhesive layer.

Any suitable porous material may be employed as the porous sheet of the release liner and/or backing layer. This material essentially provides the

structural strength (tensile strength) of the release liner or backing layer. The porous sheet may be a single layer or a laminate. A porosity which allows the detectable passage of air at room temperature (20°C) under a pressure difference is appropriate.

Suitable porous synthetic polymeric materials may be used. Paper is preferred, particularly in the form of a single layer.

By paper we mean a fibrous cellulosic product in sheet form. The preferred type of paper used in this invention is bleached supercalendered kraft. The paper pulp used for making the paper used in the invention preferably complies with Recommendation XXVI of the BGA (German pharmaceutical regulatory authority).

The paper sheet should be of good quality pulp, sufficiently uniform thickness and strength and adequate flexibility to act as the backing layer or as the release liner resistant to tearing on removal from the adhesive layer. Suitable values are weight of at least 80g/m2, e. g. 120g/m2, thickness of at least 70ym, preferably 70-150ym e. g. 100ym, and tear strength of at least 400mN preferably at least 500mN.

As already mentioned, the porous sheet of the release liner may have a layer of material having low adhesivity to the adhesive layer, e. g. a silicone

layer, on at least the side contacting the adhesive layer, and preferably on both sides. Such a silicone layer should conform to pharmaceutical requirements.

The backing layer typically does not have such a silicone layer, since it must adhere sufficiently to the adhesive layer.

The invention is applicable to any steroid having the 4-en-3-one structure, and in particular to:- norethisterone acetate levonorgestrel testosterone, including 17-methyl testosterone norethandrolone norgestrienone.

Preferably the adhesive layer is of the type made by applying an aqueous adhesive system, comprising the steroid or steroids, adhesive and water, and drying the applied layer to remove water and thereby form an adhesive layer (matrix) which allows delivery of the drug in a controlled manner. Particularly, it is preferred to use one or more acrylic adhesives as the adhesive of the adhesive layer. Such a layer may contain residual amounts of water. Spreading of such an aqueous based system onto the release liner (this is the usual practice at present), is preferred for technical reasons relating to the uniformity of spread and the resulting finished product quality. Many

suitable acrylic polymers (which may also be called acrylate polymers) are known, based on derivatives of acrylic acid, for example methyl, ethyl and butyl acrylates and methacrylates. Some are copolymerized.

For example they are anionic dispersions of acrylic polymers in water, which have been used extensively in the manufacture of pressure-sensitive adhesives, e. g. for wound dressings. In addition to the polymers, the adhesive emulsion system may include additional components such as surfactants, tackifiers intended to increase the adhesivity of the finished product to the surface to which it is applied and preservative agents.

Commercially available examples of such adhesive systems are Primal N560J (Rohm and Haas UK Ltd), Rhodotak 315M (formerly known as Vantac, Rhône-Poulenc Chemicals), Permutex (Stahl Holland bv) and Gelva Multipolymer emulsion (Monsanto Europe S. A.).

The adhesive layer may contain other materials, e. g. permeation enhancer and excipient, as is conventional.

In its preparation, the adhesive layer may contain one or more solvents, e. g. evaporating and non- evaporating solvents as described in EP-B-551349.

The concentration of the steroid in the adhesive layer is usually saturated or supersaturated, at least initially.

EXAMPLES Examples of the invention will now be given, together with comparative examples, but the invention is not limited to these examples.

Each of the transdermal delivery devices described below was made up and then subjected to storage under the conditions specified. In each Test, samples were made with a release liner of paper and a release liner other than paper as specified. The paper release liner (identified below as"Paper") was a bleached supercalendered kraft paper made of pulp meeting the requirements of Recommendation XXVI of the BGA, with a conventional silicone coating on both sides. The paper sheet has a weight of 120 g/m2, thickness of 100 ym and tear strength of more than 500 mN. The gas-impermeable release liners used were metallized polyester (MP), unmetallized polyester (UMP), high density polyethylene (HDPE) and metallized paper (M Paper). Unmetallized polyester (UMP) was used as the backing layer, except in a few cases, as indicated.

The steroids incorporated into the adhesive matrices are estradiol (E), norethisterone acetate (NEA), levonorgestrel (LN) and testosterone (T). Their concentration is given in mg/g of adhesive. The storage conditions are expressed in weeks (w),

temperature in degrees Celsius (°C) and relative humidity (or).

In Tests 4 and 6, the relative humidity level is not given. No humidity control was exercised in these tests, and the humidity was that which established itself in the test chambers. In each of these two tests, all the samples were present in the same atmosphere throughout the storage period.

The adhesives used are distinguished by reference numbers but are not individually identified because they appear to have little significant influence on the drug loss performance (compare Tests 1 and 2, Tests 4 and 6, Tests 10-13, for example). All the adhesives were commercially available aqueous dispersion acrylate polymer or multipolymer adhesives.

The mixture of adhesive system including water, drug and other components (in these examples a mixture of evaporating solvent (ethyl alcohol) and non-evaporating solvent (diethyltoluamide) as described in EP-B-551349 was included), was spread on the release liner and dried in a filtered warm air stream, before application of the backing layer to complete the device. In each test, the same adhesive layer was used in all samples.

In each Test, the transdermal device using a porous release liner is identified as B, while comparative devices are identified as A.

Test 1 Table 1 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 1A E 21.5 I UMP MP 26/40/75 1.0 1B E 21.5 2 UMP Paper 26/40/75 1 4 Test 2 Table 2 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RHs) (t) adhesive) 2A E 21.1 II UMP MP 26/40/75-1.9 2B E 21.1 II UMP Paper 26/40/75-1. 9 The results for Tests 1 and 2 show clearly that loss of estradiol on storage was very low as seen in Test 1 or there was an apparent increase as seen in Test 2. In either case, the extent of the change is at about the limit of analytical assessment and it is concluded that both results represent"no change".

This conclusion is supported by the results for estradiol in Tests 7,8 and 9 below.

For the other three drugs employed, all having the 4-en-3-one structure, there were large or very large reductions in the losses on storage, when paper release liners were employed, as shown by Tests 3 to 13.

In Tests 7,8 and 9, two steroid drugs were used in combination in the adhesive matrix.

Test 3 Table 3 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 3A NEA 14.5 III UMP MP 27/40/75 38.7 3B NEA 14.5 III UMP Paper 27/40/75 25.4 Test 4 Table 4 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 4AI NEA 14.5 III UMP MP 8/60/-73.0 4AII NEA 14.5 III MP MP 8/60/-76.5 4AIII NEA 14.5 III UMP UMP 8/60/-41.3 4B NEA 14.5 III UMP Paper 8/60/-38.4 Test 4 was conducted at a higher storage temperature (60°C), well above temperatures encountered during normal storage. Under these conditions, the sample made with unmetallised polyester (UMP) used for both the release liner and the backing showed almost as good drug loss performance as the porous paper sample.

Whether this is due to temperature-related gas permeability of the polyester resulting in higher diffusion rates at this elevated temperature, or other causes, is not known.

Test 5 Table 5 Test Drug, ReleaseStorageDrugBacking No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 5AI NEA 15.5 IV UMP MP 51/40/75 26.0 5AII NEA 15. 5 IV MP MP 51/40/75 27. 7 5B NEA 15.5 IV UMP Paper 51/40/75 13.1 Test 6 Table 6 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/C/RH%) (%) adhesive) 6AI NEA 15. 5 IV UMP MP 8/60/- 39. 3 6AII NEA 15.5 IV MP MP 8/60/-37.8 6AIII NEA 15.5 IV UMP UMP 8/60/-27.3 6B NEA 15.5 IV UMP Paper 8/60/-20.8 The comment made under Test 4 about the result for patches with unmetallized polyester as both backing layer and release liner applies here too.

Test 7 Table 7 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH (%) adhesive) %). 7A NEA 67.1 V UMP MP 12/40/75 39. 7* E 14. 7 6. 5* 7B NEA 67. 1 V UMP Paper 12/40/75 24.2 E 14.7 5.6 * These figures are the mean of results of three samples.

Test 8 Table 8 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 8AI LN 6.7 1 UMP MP 12/40/75 60.2* E 16.2 8AII LN 6.7 I UMP HDPE 12/40/75 54.9* E 16.2 8AIII LN 6.7 I UMP M Paper 12/40/75 58.7* E 16.2 8B LN 6.7 I UMP Paper 12/40/75 16.2 E 16.2 6.3 * These are % losses of LN. Average loss of E was 9.1% Test 9 This test used antioxidants in the matrix. Test 9BIII is the comparison without antioxidants, showing that they have no effect. As indicated above, this suggests that permeation of atmospheric oxygen resulting in oxidative processes in a manner that can be prevented or reduced by incorporating antioxidant substances is not a factor involved in the progressive loss of drug activity of steroids having the 4-en-3-one structure.

Table 9 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/-C/RH%) (%) adhesive) 9AI LN 7.2 I UMP MP 11/40/75 55.3# E 17.7 (+ BHA) * 9AII LN 7.2 I UMP MP 11/40/75 55.2# E 17.7 (+ MA) * 9BI LN 7.2 I UMP Paper 11/40/75 14.3# E 17.7 (+ BHA) * 9BII LN 7.2 I UMP Paper 11/40/75 17.8# E 17.7 (+ MA) * 9BIII LN 7.2 I UMP Paper 11/40/75 16.1 E 17.7 +1.5§ * Antioxidants butylated hydroxyanisole (BHA) or malic acid (MA) were included in the adhesive. <BR> <BR> <P>&num These are % losses of LN. The mean loss of E in 9AI and 9AII was 5.8% and the mean loss of E in 9BI and 9BII was 1.9%. § The small apparent gain in estradiol content indicates a change in drug concentration within the limits of analytical assessment capability. The overall loss of E seen in 9BI-9BIII is therefore bearing in mind the limits of analytical determination this is regarded as a "no-change"situation.

Tests 10-13 Table 10 Test Drug, Adhesive Backing Release Storage Drug No concentration layer liner conditions loss (mg/g (weeks/'C/RH%) (%) adhesive) 10A T 43.0 V UMP MP 27/40/75 15.0 10B T 43. 0 V UMP Paper 27/40/75 8. 6 11A T 42.8 VI UMP MP 27/40/75 20.0 11B T 42.8 VI UMP Paper 27/40/75 9.8 12A T 42.6 VII UMP MP 14/40/75 15.7 12B T 42.6 VII UMP Paper 14/40/75 2.7 13A T 40.0 VIII UMP MP 14/40/75 15.9 13B T 1

Tests 10-13, carried out on patches containing testosterone (T) which has the 4-en-3-one structure, made with four different acrylate adhesive systems, all show that the loss of T on storage for 27 weeks at 40°C/75 RH is at least twice as much from devices made using a metallised polyester (MP) release liner compared with the identical formulation on a paper (P) release liner. There is some difference in the extent of the improvement according to the Type of adhesive system employed. Types VII and VIII gave products having better stability than Types V or VI.