EP1071465A1

EP1071465A1 - Amorphous glasses for stabilising sensitive products

Info

Publication number: EP1071465A1
Application number: EP99910516A
Authority: EP
Inventors: Bruce Joseph Roser; Arcadio Garcia De Castro
Original assignee: Ronai Peter
Current assignee: Ronai Peter
Priority date: 1998-03-18
Filing date: 1999-03-17
Publication date: 2001-01-31
Also published as: AU2945199A; WO1999047174A1

Abstract

Disclosed is a method of drying, without damage, a compound which is subject to deactivation on drying, or a mixture of such compounds, comprising subjecting an aqueous system containing the compound or mixture to drying in the presence of one or more monosaccharide sugar alcohols and at least one additive which is a glass-former or a glass-formation-facilitator, whereby the compound solidifies from solution as an amorphous glass rather than by forming crystals.

Description

AMORPHOUS GLASSES FOR STABILISING SENSITIVE PRODUCTS

Drying of biological molecules such as foodstuffs, vaccines and drugs is an ancient way of preserving them. However, the price that often has to be paid for increased stability is product damage and a decrease in quality. More recently it has been learned that certain additives can prevent this damage. Levine and Slade were amongst the first to realise that the best additives were substances that tended to solidify from solution as an amorphous glass rather than by forming crystals (Levine H and Slade L., Principles of cryostabilisation technology from structure/property relationships of carbohydrate/water systems: A review. Cryoletters 9 21-63 (1988)). Foremost amongst the stabilisers they recommended were certain sugars, notably sucrose. Although not formally proved, the stability of these dried formulations is assumed to be a function of the transformation of the sugar solution into a glass as it dries. As a consequence of the progressive increase in viscosity of the sugar syrup, it is thought that the active molecules are concentrated by a smooth progression from their mobile native state in liquid solution to an immobilised solid solution or glass. In the glass, molecular motion and hence diffusion-based molecular interactions are negligible. The chemistry responsible for degradation is arrested and the product remains stable as long as it remains dry and glassy (Franks F., Freeze drying: From empiricism to predictability. Cryoletters 11 93-110 (1990)).

The simple view that glass formation is the explanation for stability is incomplete. This explanation only holds for glasses that are inherently chemically unreactive and stable. It is incontrovertible that certain substances that form physically excellent glasses are nevertheless very poor stabilisers. Many of these are either chemically reactive, such as reducing sugars or are unstable, like sucrose or sorbitol and break down to reactive intermediates which degrade the product during storage in the dry state (Newman Y.M., Ring S.G. and Colaco C. The role of trehalose and other carbohydrates in biopreservation. Biotechnol. & Genet. Eng. Rev. 11 263-294 (1993)). A common reactivity of sugar glasses is the reducing action of the carbonyl group and the degradation products formed are often the familiar carbonyl-amine compounds of the Maillard reaction (Ellis G.P. The Maillard reaction. Adv. Carbohyd. Chem. 14 63-134 (1959)). Because the Maillard reaction is temperature dependent, it is only slowly progressive at low temperatures. Many glass-forming materials give excellent preservation of activity during the drying process itself but the product subsequently undergoes progressive deterioration unless stored under refrigeration.

The error of ignoring the subtle sugar chemistry that can proceed in dried preparations is widespread in the literature and has led to the advocacy of simple tests of the glass transition temperature of pure sugar solutions as a means of selecting good stabilisers. This approach has actually led to the recommendation of quite useless substances in the past (Franks F. Freeze drying: From empiricism to predictability Cryoletters 11 93-1 10 (1990)). In fact, the efficacy of a formulation is the result of multiple physical and chemical interactions between all the components in the formulation, including the active, on drying. All these interactions may not be predicted by current theories.

The superiority of the disaccharide trehalose as a stabiliser was first indicated by its prevalence in certain rare living creatures, which regularly dried out and could come back to life on rehydration. (Crowe J.H., Crowe L.M. and Chapman D. Preservation of membranes in anhydrobiotic organisms: The role of trehalose. Science 223 701 (1984)). In laboratory studies, trehalose incorporated into the buffer solutions from which active biomolecules were dried, resulted in a product with quite remarkable resistance to denaturation by heat (Colaco C.A.L.S., Sen S., Thangavelu M., Pinder S. and Roser B. Extraordinary stability of enzymes dried in trehalose: Simplified molecular biology. Biotechnol. 10. 1007-101 1 (1992)). Because it is subsequently degraded in the body by the specific enzyme, trehalase, to two molecules of glucose, trehalose possesses many of the properties of the ideal industrial stabiliser for foods and medical products. A large scientific and patent literature has now developed on trehalose stabilisation of foods, vaccines, diagnostics and drugs. The disadvantages of trehalose are that it is as yet not approved by the regulatory authorities, it is expensive and it contains contaminating reducing sugars, especially glucose, in all but the most rigorously purified material.

One of the alternative non-reducing and chemically stable sugar derivatives that might be expected to stabilise effectively is mannitol. Because of its remarkable resistance to water sorption at high atmospheric humidity (Wade A. and Weller P.J. Handbook of Pharmaceutical Excipients second edition p296 (1994)), it is widely used in tablet and powder formulations as a bulking and anti-caking agent. In combination with other excipients such as glycine, it is also widely used in freeze dried parenteral preparations but is added as a carrier to produce a stiff, homogeneous cake that improves the appearance of the lyophilised plug in a vial (Wade & Weller The Excipients Handbook 1994).

In published surveys of the stabilising ability of a wide range of sugars and sugar derivatives (Colaco C.A.L.S., Smith C.J.S., Sen S., Roser D.H., Newman Y., Ring S. and Roser B J. Chemistry of protein stabilisation by trehalose in Formulation and delivery of proteins and peptides Cleland and Langer eds American Chemical Society Washington 222-240 (1994)), it was shown that mannitol and sorbitol were very poor stabilisers. Indeed in several patent applications it has been claimed that mannitol and certain other monosaccharide alcohols cannot stabilise at all. PCT No WO 91/18091 "Stabilisation of biological macro-molecular substances and other organic compounds", Roser B.J. and Colaco C claimed that only non-reducing glycosides of a polyhydroxy sugar alcohol or other straight chain polyalcohol or raffinose, stachyose or melezitose were effective in achieving stability, especially on storage. This patent states "Thus the monosaccharide sugar alcohols galactitol, mannitol and erythritol are not satisfactory protective agents". US patent Number 5,621,094 "Method of Preserving Agarose Gel Structure During Dehydration by Adding a Non-reducing Glycoside of a Straight Chain Sugar Alcohol" by Roser B. and Colaco C states that "glucose mannitol and sorbitol failed after one week" while "lactitol and trehalose were perfect after >12 weeks". It further defined effective formulations as "wherein the non-reducing glycoside of a straight chain sugar alcohol does not form crystals during dehydration". PCT No WO 96/05809 "Improved method for stabilisation of biological substances during drying and subsequent storage and compositions thereof, by Colaco C, Roser B.J. and Sen S. claims methods wherein even reducing sugars are used to stabilise products. This is achieved by preventing the sugars from attacking the product by using Maillard reaction inhibitors. This application states that mannitol has no stabilising effect whatsoever.

We have now found this to be incorrect. In contradiction to the statements in these documents we have found that certain monosaccharide sugar alcohols such as mannitol and inositol can be excellent stabilisers when correctly formulated and in fact have significant advantages over trehalose for some applications. In view of mannitol 's acceptance by regulatory authorities and widespread use in the healthcare industry in both parenteral and oral formulations, it has considerable advantages as a new stabilising excipient. Its low cost and chemical inertness, together with its exceptional stability and its high purity and safety, would make it the stabiliser of choice for pharmaceuticals. We have found that certain substances must be present in a formulation to convert mannitol into an excellent stabiliser. The effect of these substances is dose dependent and below a threshold concentration they do not work. The substances useful in accordance with this invention promote the drying of mannitol solutions as glasses rather than crystals. One of the most potent materials is the borate ion either as boric acid, or tetraborate salts of sodium or potassium. This probably forms a network complex with mannitol or even a covalent compound, sodium mannitoborate. Subsequent to the filing of this application containing the disclosure of the surprising efficacy of small amounts of borate in inducing glass formation in drying mannitol solutions, similar beneficial effects with borate and trehalose were reported by Miller et al. Notably, the molar ratio of borate to trehalose used by these authors was considerably higher than we found to be necessary. (Miller DP, Anderson RE and de Pablo JJ. Stabilisation of lactate dehydrogenase following freeze- thawing and vacuum-drying in the presence of trehalose and borate. Pharmaceutical Research 15 1215-1221 (1998)). Other effective materials such as calcium lactate, and proteins such as serum albumin or gelatin, or polyamine materials such as polyvinyl pyrollidone, or polyvinyl alcohol, intrinsically form glasses themselves when dried from solution. Yet other effective chemicals such as acetate salts, will form glasses but only when quenched from the melt and only when the melt contains several metal cations rather than a single cation, such as sodium and calcium. An additional property of the materials identified to date is that the beneficial actions of these materials are additive so that they can be mixed together in successful formulations which contain sub-threshold doses of each additive alone. Other substances which are either themselves glass-formers (under certain conditions) or are glass-formation-facilitators such as the phosphate salts of sodium and potassium and sodium silicate are capable of being utilised to make stabilising glasses according to this invention.

The quality of the glasses made by this process is high. The glass transition temperature (Tg) of 1 :1 w/w mannitol/calcium lactate glass is around 68°C (Figure 1). This compares with a Tg of around 90°C for a trehalose/sodium sulphate glass dried under the same conditions (Figure 2). Both types of glass have Tg's well above any possible ambient storage temperature and, because the glasses are chemically inert and non-reactive, the entrapped products are stable at room temperatures and require no refrigeration of any kind. It is a particular advantage of this invention that sugars \\ Inch have pre\ lously been known as stabilisers in methods of freeze drying may be used successfully in drying compounds subject to deactivation on drying whilst utilising drying temperatures above freezing point (e g. room temperature or above) Thus US Patent 5,589,167 "Excipient stabilisation of polypeptides treated λ\ ith organic solvents" by Cleland and Jones reports that mannitol and trehalose are both excellent stabilisers for the recombinant peptides human Growth Hormone and human Interferon gamma on freeze-drying and on exposure to organic solvents. These authors found that the ratio of active to excipient was cπtical in achieving the optimum stabilisation Without wishing to be bound by theory, it is likely that the good results reported in this patent are not so much due to the substitution of freeze-drymg for ambient temperature drying but the use of very high ratios of active to excipient m the formulations Under these circumstances, where the mass ratio of peptide to mannitol or trehalose was 1 to 1 or higher, the peptide itself was acting as an enhancer of glass transformation of the drying sugar solutions in a manner analogous to the action of albumin or gelatin shown below. A seπous disadvantage of this approach is that highly potent drugs and vaccines need tiny amounts of stabilising sugars to give a good stabilising glass. Single dose ampoules for example appear to be completely empty; the thin film of dried product being invisible in the container This is confusing to the end user and can be wasteful, if the container is discarded in error, or even dangerous, if the time to complete dissolution of the product is not obvious after reconstitution This greatly reduces the flexibility of formulation and presentation of the product

What is required is a robust formulation that inherently forms a good glass even in the absence of product but which can accommodate a wide range of product concentrations without loss of glass forming capacity and stabilising efficacy A number of sugar alcohols previously rejected as stabilising agents in the prior art listed above such as mannitol, xylitol, mositol, arabmitol and galactitol stabilise very effectively when correctly formulated so as to promote the formation of a glassy matrix, rather than crystals, on drying. A simple method for inhibiting crystallisation is to mix two or more sugars or sugar derivatives together in the same formulation When correctly chosen, these mutually inhibit crystallisation and the mixture dπes as an amorphous glass. In some cases these glasses are more robust on storage and give greater stability to an included product than trehalose itself Brief Description of the Figures

Figure 1 shows differential scanning calorimetry of a 50 / 50 w/w mannitol / calcium lactate glass showing a clear glass transition at a temperature of 68 °C;

Figure 2 shows differential scanning calorimetry of a 50 / 50 w/w trehalose / calcium lactate glass showing a clear glass transition at a temperature of 90 °C;

Figure 3 shows the percentage recovery of alkaline phosphatase activity after vacuum- drying in either trehalose or formula 7 containing mannitol, inositol, galactitol and degraded gelatin (Byco C) followed by storage at 37°C or 50°C for up to 6 weeks. There is no loss with formula 7 but serious losses with trehalose;

Figure 4 shows the percentage recovery of alkaline phosphatase activity after freeze-drying in either trehalose or formula 7 containing mannitol, inositol, galactitol and degraded gelatin (Byco C) followed by storage at 37°C or 50°C for up to 7 weeks. There is little loss with either stabiliser;

Figure 5 shows the percentage recovery of Erythropoietin (EPO) after vacuum-drying in either trehalose or formula 8 containing mannitol, inositol, galactitol and calcium lactate followed by storage at 37°C or 50°C for up to 6 weeks. While there is serious losses with trehalose, no loss occurs with formula 8; and

Figure 6 shows the percentage recovery of EPO after freeze-drying in either trehalose or formula 7 containing mannitol, Inositol, galactitol and degraded gelatin (Byco C) or formula 8 in which calcium lactate was substituted for the gelatin. After storage at 37°C or 50°C for 7 weeks, there is no loss with any of the stabilising formulations.

Figure 7 shows the percentage recovery of alkaline phosphatase activity after spray drying formula 9 to which had been added insoluble calcium phosphate powder to increase the density of the glass microspheres. After storage at either 37°C or 55°C for up to 90 days there was no significant loss of activity. Figure 8 shows the percentage recovery alkaline phosphatase activity after spray drying formula 1 1 to which had been added insoluble barium sulphate powder to increase the density of the glass microspheres. After storage at either 37"C or 55°C for up to 90 days there was again no significant loss of activity.

Examples

Example 1

A solution of mannitol in water 20% w/v was pipetted in 100 μl volumes on to the surface of clean glass microscope slides which were laid flat on a hotplate at 70°C for drying. Within about 5 min. the solution had dried into a mass of crystals. A 20% solution of trehalose or palatinit, dried under the same conditions formed a hard and transparent perfect glass film. Only the latter sugars stabilise actives successfully as described in the patents and publications referenced above. This is considered to be a function of their ability to form amorphous glass on drying.

The addition of a network forming additive such as sodium or potassium tetraborate to the mannitol solution in amounts of less than 10% of the weight of mannitol, completely inhibited crystallisation on drying and resulted in the formation of glasses as perfect as those made with trehalose or palatinit. This demonstrated that mannitol can form glasses under appropriate conditions and a search was then made for less toxic additives to achieve the same effect.

Example 2

Equal weights of trehalose or palatinit were mixed with the mannitol in solution and dried as above. In both cases this yielded perfect glasses showing that these two glass-forming sugars could inhibit the crystallisation of mannitol. Even a sugar which was not itself a glass former, such as galactitol, inhibited the crystallisation of a mannitol / inositol mixture which itself crystallised readily. Similar results were found when equal weights of other glass forming substances such as calcium lactate, albumin, polyvinyl pyrollidone or degraded gelatin (Byco C) were added to mannitol in solution. To establish the longer term stability of these glasses they were held at 70°C overnight and then at room temperature and ambient humidity for several weeks and inspected frequently. All the above glasses were stable under both sets of conditions. When other monosaccharide alcohols such as galactitol, xylitol. arabinitol, adonitol, or inositol were substituted for mannitol. similar results were obtained but the resulting glasses were very soft when alcohols of the pentose sugars were used.

Example 3

More complex mixtures of the monosaccharide alcohols could also be blended together with glass forming substances to yield excellent glasses, which showed good physical stability in the glass phase at 70°C and at ambient conditions for many weeks as described in Example 2. Some good formulations are:-

1. mannitol 33.3%, inositol 33.3% and PVP 33.3%

2. mannitol 31.6%, inositol 31.6% xylitol 5% and calcium lactate 31.6%

3. mannitol 33.3%, inositol 33.3% and calcium lactate 33.3%

4. mannitol 33.3%, inositol 33.3% and Byco C 33.3%

5. mannitol 23.3%, inositol 23.3% calcium lactate 30% and PVP 23.3%

6. mannitol 33.3%), arabinitol 33.3%o and calcium lactate 33.3%

7. mannitol 30%, inositol 15% galactitol 15% and Byco C 40%

8. mannitol 30%, inositol 15% galactitol 15% and calcium lactate 40%

9. mannitol 33%, Byco C 33% and calcium lactate 33%

10. mannitol 50%, and Kollidon 30 (polyvinylpyrollidone (PVP)) 50%

11. mannitol 33%, Kollidon 30 (polyvinylpyrollidone (PVP)) 33% and calcium lactate 33%

12. mannitol 50%, and Dextran 50%

13. mannitol 33%, Dextran 33% and calcium lactate 33%

In short, simple trial and error experimentation will establish a successful formulation from mixtures of monosaccharide alcohols and a glass forming substance. By this method the final concentration of any single ingredient can be kept low. In this way a substantial total solids content can be achieved, even including sugar alcohols, which are individually not very soluble. The high solids content shortens drying times and increases the protection of the active during drying.

In addition to heat-assisted air-drying as above, formulations of this kind have been successfully vacuum dried, spray dried and freeze-dried. Example 4

Stability of alkaline phosphatase enzyme

Affinity purified alkaline phosphatase from bovme intestinal mucosa (Sigma Chemical Co cat No p-8647) was vacuum-dπed or freeze-dπed m 200 μl volumes in formulation Number 7 above or in trehalose. and stored at 37°C or 50°C for 5 weeks Samples were tested for activity at intervals using the Sigma assay with p-nitrophenyl phosphate as substrate Vacuum drying was done at a shelf temperature of 40°C and a vacuum of 30- 100 milhtorr for 4 hr The temperature was then ramped gradually to 60°C over 1 hr and the vials were stoppered and removed from the vacuum chamber for high temperature storage trials Freeze-drying was done in a Labconco dryer at an initial shelf temperature of-40°C for 3 hr at a vacuum of 30-100 milhtorr The shelf temperature was then ramped to 0°C at 5°C / mm and held for 1 hr The shelf temperature was then raised to 40°C at 5°C / min and secondary drying was continued for a further 3 hr when the vials were stoppered under vacuum and removed for storage trials

While there was some variability in the assays of enzyme activity, obvious trends were observed Samples dried by either method without stabilisers lost all activity withm a day or two of storage (not shown) Samples dried in either mannitol alone or a modified formula 7 lacking the glass forming facilitator lost between 75 and 80%> of activity within 3 days at 37°C

There was also a progressive loss of enzyme activity seen with the samples vacuum-dπed in trehalose, which was not seen m the samples dπed in formula 7 (Fig 3) This was particularly marked where the samples had been stored at the higher temperature (50°C) No such loss in the trehalose-dπed samples was seen in the freeze-drying experiments where trehalose appeared to be slightly superior to the monosaccharide alcohols (Fig 4) This result might possibly indicate that a higher residual moisture content may have been responsible for the seπous losses with trehalose in the vacuum dπed samples Whatever the explanation, it is clear that formula 7 gives results which are equivalent to, or supeπor to, the results obtained with trehalose Example 5

Stability of lecombmant human Erythropoietin (EPO)

EPO was vacuum dried or freeze-dπed as above in the same solutions and also in a vaπant of formulation 7 in which calcium lactate was substituted for Byco C (Formula 8), and then subjected to the same stability tests before being assayed by a standard 2-sιte sandwich Enzyme Immunoassay

The results again showed a serious, progressive deterioration in the vacuum dπed samples dπed in trehalose, more dramatic at 50°C storage temperature, which was not seen with the samples dπed in formula 8 (fig 5) The deterioration in trehalose was not seen m the freeze-dπed samples stored at 37°C (not shown) or 50°C (Fig 6) Irrespective of whether formula 7 or 8 was used, all samples showed essentially complete recovery of activity

Example 6

The fluorescent protein R-Phycoerythπn was air-dried in trehalose, formula 3 or formula 4 on a hotplate as described in Example 1 The intensity of fluorescence was gauged visually when illuminated with a UV lamp. In the controls dried in trehalose fluorescence was retained The material dried in formula 3 was masked by an intense silver autofluorescence from the Byco C while the material dried in formula 4 fluoresced with the characteristic orange colour with apparently undimmished intensity

Claims

1. A method of drying, without damage, a compound which is subject to deactivation on drying, or a mixture of such compounds, comprising subjecting an aqueous system containing the compound or mixture to drying in the presence of one or more monosaccharide sugar alcohols and at least one additive which is a glass-former or a glass- formation-facilitator, whereby the compound solidifies from solution as an amorphous glass rather than by forming crystals.

2. A method according to claim 1 in which the aqueous system contains from 0.05 to 90%) by weight of sugar alcohol.

3. A method according to claim 1 in which the ratio of sugar alcohol plus additive to compound is at least 0.25: 1 preferably 0.5: 1 by weight.

4. A method according to any of claim 1 to 3 in which the compound is a protein, polysaccharide or nucleic acid.

5. A method according to claim 4 in which the compound or mixture comprises an enzyme, serum, serum complement, an antibody or antigen (either free or coupled to a support), a nucleic acid, a fluorescent protein, or a vaccine component.

6. A method according to any of claim 1 to 5 in which the system is dried under conditions selected from one or more of the group consisting of ambient temperature or above, chill drying, freeze drying, spray drying, vacuum drying and drying at atmospheric pressure.

7. A dried product which is an amorphous glass containing monosaccharide sugar alcohol and at least one additive which is a glass-former or a glass-formation-facilitator and a compound which is subject to deactivation on drying, or a mixture of such compounds, in a weight ratio of sugar alcohol plus additive to compound of at least 0.25:1 preferably 0.5:1, the product having been dried.

8. A dried product according to claim 7 in which the compound is a protein, polysaccharide or nucleic acid.

9. A dried product according to claim 8 containing an enzyme, serum complement, an antibody or antigen (either free or coupled to a support), a nucleic acid, a fluorescent protein, or a vaccine component.

10. A method or product according to any preceding claim wherein the sugar alcohol is selected from the group consisting of mannitol, galactitol, xylitol, arabinitol and inositol.

11. A method or product according to any preceding claim wherein there is one or a mixture of additives selected from the group consisting of peptide, protein, borate ion, calcium lactate, phosphate, silicate and acetate salts.

12. A method or product according to claim 11 wherein at least one additive is selected from the group consisting of boric acid, tetraborate salt of sodium or potassium and sodium mannitoborate.

13. A method or product according to any preceding claim wherein the amorphous glass is formed from a mixture of 2 or more monosaccharide sugar alcohols.

14. A method or product according to any preceding claim wherein there is an additive which is a protein or denatured protein.

15. A method or product according to claim 1 or 7 wherein the amorphous glass is formed from a formulation having essentially a composition selected from:

1. mannitol 33.3%, inositol 33.3% and PVP 33.3%

2. mannitol 31.6%>, inositol 31.6%> xylitol 5%> and calcium lactate 31.6%

3. mannitol 33.3%, inositol 33.3% and calcium lactate 33.3%

4. mannitol 33.3%, inositol 33.3% and Byco C 33.3%

5. mannitol 23.3%, inositol 23.3% calcium lactate 30% and PVP 23.3%

6. mannitol 33.3%, arabinitol 33.3% and calcium lactate 33.3%

7. mannitol 30%, inositol 15% galactitol 15% and Byco C 40%)

8. mannitol 30%, inositol 15% galactitol 15% and calcium lactate 40%_>

9. mannitol 33%o, Byco C 33% and calcium lactate 33%

10. mannitol 50%, and Kollidon 30 (polyvinylpyrollidone (PVP)) 50%

11. mannitol 33%, Kollidon 30 (polyvinylpyrollidone (PVP)) 33%o and calcium lactate 33%

12. mannitol 50%, and Dextran 50%

13. mannitol 33%, Dextran 33% and calcium lactate 33%