GB2500204A - Vaccine comprising aluminium adjuvant having low levels of contaminating heavy metal ions - Google Patents

Abstract

Protein vaccines comprising aluminium adjuvant and methods to prepare them are disclosed wherein an aluminium salt is selected (e.g. by selecting a particularly contaminating metal free commercial batch) such that when formulated into an aqueous vaccine composition comprising aluminium and a protein that vaccine composition comprises less than 350 ppb heavy metal w/v in the aqueous solution. The preparation of a Japanese Encephalitis virus (JEV) vaccine comprising less than 3 ppb copper and less than 350 ppb heavy metals is disclosed and the stability of the vaccine is demonstrated to be better than when the heavy metal concentration exceeds 350 ppb.

Description

Aluminum immunological adjuvants low in heavy metal content

The invention relates to the fields of pharmaceuticals and vaccines. More 5 in particular the invention relates to the field of compounds and compositions that are being co-administered with the medicament and/or antigen.

Aluminium compounds (herein also referred to as "aluminium"), including aluminium phosphate (A1P04), aluminium hydroxide (Al(OH)3), and other aluminium precipitated vaccines are currently the most commonly used 10 adjuvants with human and veterinary vaccines. The adjuvants are often referred to as "alum" in the literature.

Aluminium adjuvants have been used in practical vaccination for more than half a century. They induce early, high-titre, long-lasting protective immunity. Billions of doses of aluminium-adjuvated vaccines have been 15 administered over the years. Their safety and efficacy have made them the most popular adjuvants in vaccines to date. In general, aluminium adjuvants are regarded as safe when used in accordance with current vaccination schedules.

In human vaccinations, of old, aluminium adjuvants have been used in 20 tetanus, diphtheria, pertussis and poliomyelitis vaccines as part of standard child vaccination programmes. Aluminium adjuvants have also been introduced into hepatitis A and hepatitis B virus vaccines and Japanese encephalitis virus (also referred herein as "JEV") vaccines. Other aluminium-adsorbed vaccines against, for example, anthrax, are available for special risk 25 groups. In veterinary medicine aluminium adjuvants have been used in a large number of vaccine formulations against viral and bacterial diseases, and in attempts to make antiparasite vaccines.

Adjuvants typically serve to bring the antigen, the substance that stimulates the specific protective immune response, into contact with the 30 immune system and influence the type of immunity produced, as well as the quality of the immune response (magnitude or duration); Adjuvants can also decrease the toxicity of certain antigens; and provide solubility to some

1

vaccines components. Studies have shown that many aluminium-containing vaccines cause higher and more prolonged antibody responses than comparable vaccines without the adjuvant. The benefit of adjuvants has usually been observed during the initial immunization series rather than with booster doses.

There are three general types of aluminium-containing adjuvants: Aluminium hydroxide, Aluminium phosphate and Potassium aluminium sulphate (collectively often referred to as "Alum")

The effectiveness of each salt as an adjuvant depends on the characteristics of the specific vaccine and how the manufacturer prepares the vaccine. To work as an adjuvant, the antigen is typically adsorbed to the aluminium; that is, it is clumped with the aluminium salt to keep the antigen at the site of injection.

Not all vaccines contain aluminium salts. Sometimes an adjuvant may not have been needed or another adjuvant was selected. Examples of commercial vaccines that do not contain aluminium salts are inactivated Polio Virus (IPV) vaccine, measles, mumps and rubella vaccine (MMR), varicella vaccine, Meningococcal conjugate (MCV4) vaccine, and influenza vaccines. That the commercial vaccines do not contain aluminium salts does typically not mean that an aluminium salt would not work. It just means that for some reason another adjuvant was selected.

Examples of US licensed vaccines for children that contain aluminium adjuvants are: DTP (diphtheria-tetanus-pertussis vaccine); DTaP (diphtheria-tetanus-acellular pertussis vaccine); some but not all Hib (Haemophilus influenzae type b) conjugate vaccines; Pneumococcal conjugate vaccine; Hepatitis B vaccines; Hepatitis A vaccines; Human Papillomavirus vaccine; Anthrax vaccine; and Rabies vaccine.

Aluminium is a very abundant element in our environment. It is in many foods we eat, many personal hygiene products we apply to our skin (deodorants, for example), and many medicines we ingest. Various government agencies establish guidelines for exposure to potentially toxic substances. These guidelines are called "minimal risk levels" —the maximum amount that

one can be exposed to over time-usually on a daily basis-without expected harm.

The US Agency for Toxic Substances and Disease Registry (ATSDR) estimated these levels for infants taking into account the amount of 5 aluminium (e.g. in form of a salt) a child would eat as well as receive by injection of vaccines. The body burden of aluminium from both sources is below the minimal risk level except transiently following vaccinations; since 50-70% of injected aluminium is excreted within 24 hours, this is believed to have no negative effect.

10 Aluminium hydroxide and aluminium phosphate adjuvants are generally prepared by exposing aqueous solutions of aluminium ions, to usually slightly alkaline conditions in a well-defined and controlled chemical environment. Various soluble aluminium salts can be used for the production of aluminium hydroxide. Anions present at the time of precipitation may coprecipitate (for 15 review see, Lindblad, EB (2004) Immunol, and Cell Biol. Vol 82: 497-505).

Aluminium salt is also used in the manufacture and composition of medicaments. For instance, factor VIII is purified from plasma cryoprecipitate. The precipitate is solubilised, absorbed on aluminium hydroxide and then treated to inactivate lipid enveloped viruses. After several other processing 20 steps the concentrate is used to treat hemophilia A patients (Burnouf T, (1991) Vox Sang. Vol 60: pp 8-15).

In the present invention it has been shown that stability of a biological in a composition that also comprises an aluminium salt is not always the same. The present invention, for instance, shows that the stability of a protein 25 component (e.g. as such or within a complex such as e.g. a virus or other pathogen) in the context of an aqueous composition that also comprises aluminium salt is dependent on the content of heavy metals. To estimate a priori whether the protein will be stable in this composition, the present invention provides that it is necessary to determine the residual heavy metal 30 content in the composition (otherwise the aqueous composition comprising a protein is at risk of being degraded over time in particular the invention provides that this risk is considerable when the residual heavy metal content

3

is above 350 ppb (i.e. about 350ng per ml) in said aqueous composition). Further, the invention also revealed that this residual heavy metal content cannot easily removed from the aluminium compound. To this end the invention provides a method for preparing an aqueous pharmaceutical or 5 vaccine composition comprising aluminium, a reactive compound and a protein said method comprising selecting an aluminium-salt that is able to provide an aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition and combining said aluminium salt, said reactive compound, said protein and water to produce said aqueous 10 composition having less than 350 ppb heavy metal based on the weight of the aqueous composition; wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof. Compositions comprising less than 350 ppb heavy metal based on the weight of the aqueous 15 composition, can be stored in a liquid phase at a temperature of between 0 and 30 degrees Celsius, for at least 1 month, such as e.g. 20 months at 2-8°C. The protein component in said composition is stable for at least 1 month in said liquid phase. Compositions comprising more than 350 ppb heavy metal based on the weight of the aqueous composition cannot be stored for a prolonged 20 period under such conditions as the protein component in said composition changes in at least one aspect over the indicated time period. One millilitre or one gram of aqueous composition thus preferably contains no more than 350 nanogram heavy metal. The aqueous composition preferably comprises between 0,1 mg/ml and 2,5 mg/ml aluminium. The average dose of aluminium 25 per administration is preferably not more than 1,25 milligram (mgram). In a particularly preferred embodiment the dose of aluminium per administration is not more than 0,25 mgram aluminium. A dose typically comprises between 0,5 and 1 ml of the aqueous composition.

In a further aspect of the present invention it has been shown that 30 stability of a biological in a composition that comprises an aluminium salt and a reactive compound is not always the same. The present invention, for instance, shows that the stability of a protein component (e.g. as such or

4

within a complex such as e.g. a virus or other pathogen) in the context of an aqueous composition that also comprises aluminium salt and a reactive compound such as e.g. a sulphite is critically dependent on the content of heavy metals. To estimate a priori whether the protein component (such as e.g.

5 protein component within a complex such as e.g. a virus particle; herein also referred to simply as protein) will be stable in this composition, it is necessary to determine the heavy metal content in the composition. To this end the invention provides a method for preparing a clinical grade aluminium-salt precipitate for incorporation into a medicament and/or vaccine, said method 10 comprising preparing an aqueous solution of aluminium ions and precipitating said aluminium-ions from said solution, and determining the level of a heavy metal in the solution and/or the aluminium-salt precipitate, wherein the precipitate is selected that is able to provide an aqueous composition comprising less than 350 ppb heavy metal based on the weight of the aqueous 15 composition. Compositions comprising less than 350 ppb heavy metal based on the weight of the aqueous composition, can be stored in a liquid phase at a temperature of between 0 and 30 degrees Celsius, for at least 1 month, such as e.g. 20 months at 2-8°C. The protein component in said composition is stable for at least 1 month in said liquid phase, such as e.g. 20 months at 2-8°C. 20 Compositions comprising more than 350 ppb heavy metal based on the weight of the aqueous composition cannot be stored for a prolonged period under such conditions as the protein component in said composition changes in at least one aspect over the indicated time period. One millilitre or one gram of aqueous composition thus preferably contains no more than 350 nanogram 25 heavy metal. The aqueous composition preferably comprises between 0,1

mg/ml (milligram per millilitre) and 2,5 mg/ml aluminium. The average dose of aluminium per administration is preferably not more than 1,25 milligram (mgram). In a particularly preferred embodiment the dose of aluminium per administration is not more than 0,25 mgram aluminium. A dose typically 30 comprises between 0,5 and 1 ml of the aqueous composition. An aqueous composition comprising a protein, an aluminium-salt, and optionally a reactive compound, said composition comprising less than 350 ppb heavy metal based

5

on the weight of the aqueous composition is herein also referred to as "an aqueous composition comprising a protein according to the invention" or "a composition comprising a protein according to the invention".

It has been observed that the aluminium component is an important source for the heavy metal in the aqueous composition. Thus one way to control the amount of heavy metal in the aqueous composition is to control the amount of heavy metal in the aluminium source used to generate the aqueous composition. The invention therefore further provides a method for preparing an aqueous pharmaceutical or vaccine composition comprising aluminium, a reactive compound and a protein said method comprising

- selecting an aluminium-salt solution (such as e.g. lOmg/ml aluminium hydroxide liquid such as e.g. alhydrogel® 2% from Brenntag Biosector, catalogue number 843261) that in the final protein formulation comprises no more than 350ppb based on the weight of the aqueous composition (e.g. for the alhydrogel® 2% and final amount of 0,25 mgram aluminium hydroxide in the aqueous composition of 0.5ml (= dose), the selected or prepared aluminium-salt solution should not contain more than 7 microgram heavy metal pro millilitre of the alhydrogel® 2% solution, about 7ppm heavy metal content), and

- combining said aluminium-salt solution, said protein, water and a reactive compound to produce said aqueous composition having less than 350ppb heavy metal in the aqueous composition,

wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof.

It has been observed that the aluminium component is an important source for the heavy metal in the aqueous composition. Thus one way to control the amount of heavy metal in the aqueous composition is to control the amount of heavy metal in the aluminium source used to generate the aqueous composition. The invention therefore further provides a method for preparing an aqueous composition comprising aluminium and a protein said method comprising

6

- preparing or selecting an aluminium-salt solution (such as e.g.

lOmg/ml aluminium hydroxide liquid (such as e.g. alhydrogel® 2% from Brenntag Biosector, catalogue number 843261) that in the final protein formulation comprises no more than 350ppb based on the weight of the 5 aqueous composition (e.g. for the alhydrogel® 2% and final amount of 0,25

mgram aluminium hydroxide in the aqueous composition of 0.5ml (= dose), the selected or prepared aluminium-salt solution should not contain more than 7 microgram heavy metal pro millilitre of the alhydrogel® 2% solution, about 7ppm heavy metal content), and 10 - combining said aluminium-salt solution, said protein, water and optionally a reactive compound to produce said aqueous composition (with no more than 350ppb heavy metal in the aqueous composition).

For example, the aluminium-salt solution, e.g. the aluminium hydroxide liquid (used as a component to be mixed to result in the final aqueous 15 composition) should not have a heavy metal content higher than 7ppm (given as an example herein where the 10 mg/ml aluminium hydroxide liquid will be diluted to 0,5 mg/ml to result in about 350ppb heavy metal content, assuming lppm = about Img/ml) on the basis of weight of the aluminium hydroxide liquid. The limit of acceptable heavy metal content may also be expressed in 20 relation to the weight of the aluminium-salt such as the aluminium hydroxide in solution (referred to also as "starting aluminium compound"). The acceptable heavy metal content in this example then may not exceed 7 microgram of heavy metal for each gram of aluminium hydroxide solution (i.e. about 7ppm), i.e. the alhydrogel® 2% solution. As indicated herein above the 25 aqueous composi tion comprising the protein preferably (such as e.g. if used as a vaccine) comprises between 0,1 to 2,5 mg/ml of the aluminium compound. The concentration of heavy metal in the aqueous composition however should according to the invention not exceed 350ppb, i.e. about 350ng per ml of the final composition and thus the selection or preparation of the starting 30 aluminium compound has to be made accordingly. In order to further illustrate the selection of an appropriate starting aluminium compound solution (e.g. in the form of a concentrated solution (see above alhydrogel® 2%=10mg/ml)? it is

7

shown that an aluminium compound solution of lOmg/ml aluminium hydroxide having about 7 ppm heavy metal impurities corresponds to a concentration of heavy metal in the protein composition when the aluminium concentration is about O.lmg/ml of about 70 nanogram/ml or 70 ppb (so well 5 below the 350ppb provided as the limit of the heavy metal content as taught by the invention). A concentration of 2,5 mg/ml of the aluminium hydroxide in the final aqueous composition starting with an aluminium solution of lOmg/ml aluminium hydroxide having about 7 ppm heavy metal impurities will result in a heavy metal content in the aqueous protein composition that corresponds 10 to a concentration of heavy metal of about 1,75 microgram/ml or about l,750ppm (so well above the 350ppb provided as the limit of heavy metal content as taught by the invention).

A method of preparing an aqueous composition comprising a protein preferably further comprises packaging aliquots of said aqueous composition 15 having less than 350 ppb heavy metal based on the weight of the aqueous composition in separate air-tight storage containers. The protein in the airtight storage containers is stable and can be stored for at least three months, such as e.g. 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 months, preferably 20 or 24 months, more preferably 20 months at a 20 temperature of between 2 to 8°C degrees Celsius.

Without being bound to theory, antigen degradation in aqueous compositions, such as immunogenic composition, comprising heavy metal ions present in an aluminium salt, such as aluminium hydroxide, might be explained with an underlying degradation pathway assuming free-radicals 25 such as e.g. free-radicals of sulphite. Formation of free radicals can be catalysed by heavy metal ions present in, for instance, aluminium hydroxide and this effect (in case of a certain amount of heavy metal as indicated according to the invention) could be the underlying root cause mechanism for stability issues as identified as part of the inventive contribution. The 30 experimental part of this application shows in great detail the evidence of this root cause for the Japanese encephalitis vaccine (also referred to as "JEV") and makes a similar showing for a simple aluminium adjuvanted polypeptide

8

composition that comprises a reactive compound such as sulphite. Thus, it is evident that similar reaction may occur also in other aqueous composition comprising aluminium (with high, e.g. higher than 350ppb based on the weight of the composition, heavy metal content), protein and possibly a reactive 5 component such as sulphite and/or other radical forming particles. Heavy metal-catalysed oxidation is a degradation pathway resulting in covalent modification of proteins. The modified physicochemical properties of the oxidized/modified protein or antigen may result in loss of biological activity (Li et al., 1995; Mayo et al., 2003; Stadtman, 1990).

10 The following reaction schemes (as possibly occurring in the JEV product as described in the experimental part) are indicative for compositions of the invention comprising in addition to the protein and the heavy metal also sulphite such as e.g. sulphite and formaldehyde.

Sodium-metabisulfit (Na2S20s) in solution dissolutes into bisulphite

15 (S2O52"), in an alkaline solution the bisulphite equilibrium is towards sulphite (SO32") and in an acidic solution towards H2SO3/SO2. After dissolution of Na2S20,5 in water, it is hydrolysed into NaHSOs as follows Na2S205 «• 2 NaHSOs

At neutral pH we can assume the following equilibrium:

20 HSOs- «H+ + SO32- pKa=7.2

This means that at pH 7 the equilibrium is shifted towards HS03- and at more basic conditions (e.g. pH 8) towards SO32".

Formaldehyde forms a bisulphite adduct during neutralization according to the following equation:

25 CH2O + HSOs- -► CH2(0H)(S03)-

After bisulphate is used up, the reaction proceeds until equilibrium is reached as follows:

CH2O + SOs2" + H2O -> CH2(0H)(S03)- + OH-

30 Formaldehyde and sulphite react with each other, however, formaldehyde and sulphite have been found to be still present in equilibrium in the JEV vaccine and can be detected in the following range (n=49):

9

Release Results in DS

Free

Free

Sulphite

Formaldehyde

Average (ppm)

113.9

41.9

Average (mM)

1.41

1.36

Standard deviation (ppm)

24

14.3

Min (ppm)

66

10.6

Max (ppm)

174

78.7

According to the literature (Ranguelova et al., 2010), transition metal ions catalyse the auto-oxidation of (bi)sulphite via sulphur trioxide anion radical (*SO.r) formation:

Mn+ + S032- -* M(n"1)+ + •S03-

where M may be copper (Cu2+), iron (Fe3+), oxivanadium (V02+), manganese (Mn2+), Nickel (Ni2+) or chromate anion (Cr042~) (Alipazaga et al. 2004; Berglund et al. 1993; Brandt and Elding 1998; Lima et al. 2002; Shi 1994).

It was shown that such sulphite radicals are highly reactive and can oxidize various substances, such as ascorbate, Hydroquinone and Histidine (Huie at al., 1985). A review of free radical chemistry of sulphite was published by Neta & Huie, 1985. The authors also show that radical formation can be also catalysed by photoionization of sulphite as follows:

SO32- + hv *S03~ + e-

Radical formation catalysed by light might also explain differences observed in potency and ELISA results of unlabeled naked syringes (used for release testing and reference purposes) and fully packaged final vaccine lot samples for the JEV product. Fully packed samples are completely protected from light, whereas unlabeled syringes might be exposed to light during storage and handling.

An important reaction of the sulphite radical in auto-oxidation systems is with molecular oxygen to form a peroxyl radical which is much more reactive:

•SO3- + O2 ®so5-

10

The solubility of O2 in water at 0°C and 20°C is 0.4mM and 0.25mM, respectively. Assuming that O2 solubility in a composition of the invention is in a similar range, a considerable amount of oxygen is present to form the peroxyl radical. This radical is a much stronger oxidant compared to • SO3 and can 5 oxidize certain substrates which are not attached by •SOrr at all and which, in fact, can form radicals that oxidize sulphite ions. In such cases, when the redox potential of the substrate is intermediate between those of ®SC>3_ and • SO5", a reaction chain is likely to develop in presence of 02 following the general pattern:

10 ®S03- + O2 • SOrr •SOs- + X -> SO52- + *x+

®X+ + SO32- -*x+ ®so3-

The intermediacy of a substrate X may enhance the chain process of sulphite oxidation by oxygen. The one-electron reduction of • SOrr yields HSO5" 15 (peroxymonosulfate), a very strong oxidant that is capable to oxidize many organic compounds (Lambeth et al., 1973; Ito & Kawanashi, 1991). Peroxymonosulfate is also a precursor sulphate anion radical • SO4-. •SO5- + HSOs- *S04~. + HSO4-

The ® S04_.radical is a very strong oxidant, nearly as strong as the 20 hydroxyl radical (*OH), and is very likely to oxidize other biomolecules by one-electron oxidation.

In a preferred embodiment, the composition comprising a protein according to the invention is a therapeutic composition or an immunogenic composition, such as a vaccine. Therapeutic compositions are administered to 25 individuals, such as a human or an animal. In particular for such compositions, it is important that the protein within the composition still has its therapeutic effect at the time it is administered to said individual. Degradation of the protein or changes to the protein in its structure may result in the protein to lose its therapeutic activity. Similarly, degradation or 30 structural changes of the immunogenic composition will also lead to a reduction in the effectivity of the composition in inducing and/or boosting an immune response in an individual. An immunogenic composition is preferably

11

administered to an individual to counteract or prevent a viral or bacterial infection. Protein contained within the aqueous immunogenic composition can be a single protein or a multimeric protein or part of a complex comprising said protein (e.g. such as part of a virus or a cell, e.g. bacterial cell). In a preferred 5 embodiment said complex comprises a live attenuated or inactivated virus or bacterium or a immunogenic viral or bacterial protein or an immunogenic part of such protein (e.g. an immunogenic peptide). If said immunogenic composition is administered to provide protection against a viral or bacterial infection, degradation of protein may result in loss of protective capability of 10 the immunogenic composition. The term "immunogenic viral or bacterial protein" refers to a viral or bacterial protein which is capable of eliciting an immune response. The term "immunogenic part" as used herein refers to a part of a viral or bacterial protein which is capable of eliciting an immune response. Preferably the immune response elicited recognizes both said part of 15 the protein and the entire protein. Therefore, in one embodiment, an aqueous composition comprising a protein according to the invention is an immunogenic composition. Said composition is preferably a therapeutic composition and/or prophylactic composition such as a vaccine. Also provided is a vaccine that is an aqueous composition comprising a protein according to 20 the invention

An "immunogenic composition" is herein defined as a composition that is capable of eliciting an immune response when administered to an individual. The elicited immune response can be humoral, cellular or a combination thereof and includes, but is not limited to, the production of antibodies, B cells 25 such as activated B cells, and T cells such as activated T cells. An immune response as used herein is preferably directed specifically to one or more immunogens within a composition comprising a protein according to the invention. An immunogenic composition of the present invention can be administered to an individual by any technique known in the art including, 30 but not limited to, intramuscular (IM), intradermal (ID), subcutaneous (SC), intracranial (IC), intraperitoneal (IP), or intravenous (IV) injection, transdermal, oral, intranasal, or rectal administration, and combinations

12

thereof, preferred are intramuscular (IM), intradermal (ID), subcutaneous (SC), intracranial (IC), intraperitoneal (IP), or intravenous (IV) injection. In a preferred embodiment an immunogenic composition comprising a protein according to the invention is used for eliciting an immune response that may 5 be useful in chronic setting (such as cancer treatment) or prophylactic setting (such as a typical vaccine). It is preferred that the immunogenic composition is used as a vaccine, i.e. prophylactic use. In this embodiment the aluminium is typically present as the adjuvant.

An "adjuvant" as used herein refers to a pharmacological or 10 immunological agent that modifies the effect of other agents, such as an immunological agent that increases the antigenic response. Adjuvants typically serve to bring the antigen—the substance that stimulates the specific protective immune response—into contact with the immune system and infl uence the type of immunity produced, as well as the quality of the immune 15 response (magnitude or duration); decrease the toxicity of certain antigens; and provide solubility to some vaccines components

An "individual" is herein defined as a human or an animal. Individuals include but not limited to chickens, ducks, geese, turkeys, swans, emus, guinea fowls and pheasants, humans, pigs, ferrets, seals, rabbits, cats, dogs and 20 horses. In a preferred embodiment of the invention an individual is a mammal, preferably a human.

An aluminium adjuvant is often prepared by controlled exposure of an aqueous solution of aluminium ions, to alkaline conditions (for review see, Lindblad, EB (2004) Immunol, and Cell Biol. Vol 82: 497-505). In the present 25 invention it has been found for the JEV product (see experimental part) that a large amount of the heavy metal in this aqueous solution of aluminium ions ends up in the aluminium salt precipitate for the aluminium adjuvant. It has further been found that the amount of heavy metal in the aluminium precipitate affects the stability of the vaccine during storage of the vaccine. 30 The amount of heavy metal that is present in the aluminium-salt can thus be controlled by determining the amount of heavy metal in the salt but also, and preferably, by controlling the amount of heavy metal in the aqueous solution of

13

aluminium ions. The invention thus further provides a method for preparing a clinical grade aluminium-salt precipitate for incorporation into a medicament and/or vaccine, said method comprising preparing an aqueous solution of aluminium ion and precipitating said aluminium-ions from said solution, and 5 determining the level of a heavy metal in the solution and/or the aluminium-salt precipitate, preferably wherein said solution and/or the aluminium-salt precipitate is determined to comprise an amount that results in less than 350ppb heavy metal in the final composition e.g. when re-suspended in the final composition.

10 Also provided is a pharmaceutical composition comprising a protein according to the invention, optionally further comprising a pharmaceutically acceptable carrier and/or diluent. "A pharmaceutically acceptable diluent" as used herein is defined as any solution, substance or combination thereof that has no biologically or otherwise unwanted activity, meaning that it can be 15 administered to an individual together with other components of an immunological composition without causing a substantial adverse reaction. Examples of suitable carriers for instance comprise keyhole limpet haemocyanin (KLH), serum albumin (e.g. BSA or RSA) and ovalbumin. In one preferred embodiment said suitable carrier comprises a solution, like for 20 example saline.

In one aspect, a method according to the invention is used for prolonging the storage life or shelf life of an aqueous composition comprising a protein according to the invention. As used herein, the term "shelf life" is defined as the period of time a composition comprisi ng a protein according to the 25 invention can be stored without becoming unsuitable for use, for instance due to degradation of protein (e.g. is within the potency specification of the composition, e.g. vaccine, as required by the regulatory agency that approved or will approve the vaccine). During storage of aqueous compositions as described herein, degradation of the protein may occur, in particular when a 30 certain level (as described herein) of heavy metals is exceeded . Degradation generally increases with time when such aqueous compositions are stored. Now that it is found that degradation of protein is reduced in an aqueous

14

composition comprising in addition to said protein an aluminium-salt, if said composition comprises less than 350 ppb heavy metal based on the weight of the aqueous composition, it has become possible to counteract degradation of protein in aqueous compositions. By counteracting degradation of protein with 5 a method according to the invention, the stability of said protein within said composition is increased and the storage life of an aqueous compositions comprising said protein is prolonged. An aqueous composition according to the present invention provides the advantage that it is stable and does not undergo degradation of protein during a prolonged period. Such aqueous 10 composition comprising a protein according to the invention is stable for at least one month at elevated temperature such as e.g. 20 or 37°C, preferably for at least three months at elevated temperature such as e.g. 20 or 37°C.

An aqueous composition comprising a protein according to the invention is preferably stored at a temperature of between 0 °C and 20 °C to contribute 15 to an increased shelf life, more preferably between 2 °C and 15 °C, more preferably between 2 °C and 10 °C, most preferably between 2 °C and 8 °C. The shelf life at temperature between 2 °C and 8 °C is stable preferably for at least three months such as e.g. 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 months, preferably 20 or 24 months, more preferably 20 20 months at a temperature of between 2 to 8°C degrees Celsius. Provided is thus an aqueous composition comprising a protein according to the invention having a shelf life of around 12 to 24 months. The invention further provides an aqueous solution according to the invention which has been stored for at least one month, preferably for at least two months, more preferably for at least 25 three months, more preferably for at least six months.

As used herein "a stable protein composition" means that when compared to a starting composition not more than 50%, preferably not more than 40%, even more preferably not more than 30%, even more preferably not more than 20%, even more preferably not more than 10, even more preferably not more 30 than 5% of the protein in said composition is degraded. "Degraded" in this context refers to any detectable modification of the protein when compared to the protein in the starting composition. For instance, a decrease in detection of

15

the protein with a monoclonal antibody such as e.g. an antibody recognizing a neutralizing epitope is suitable and can be measured with any method known in the art, such as ELISA (see Example 4). The level detected can for instance be compared with the level detected with a polyclonal specific for the same 5 protein such as e.g. representing the amount of total protein (changed or unchanged). Also provided is a method of prolonging the shelf life of an aqueous composition comprising a protein and an aluminium-salt, said method comprising selecting and/or preparing an aluminium-salt resulting in an aqueous composition with a heavy metal content of less than 350ppb based on 10 the basis of weight of the aqueous composition and combining said aluminium salt, said protein and water to produce said aqueous composition.

Further provided is a method of improving the shelf life reproducibility of preparations of aqueous compositions comprising a protein and an aluminium-salt, said method comprising obtaining at least two different aluminium-salt 15 preparations, determining the amount of at least one heavy metal in said aluminium-salt preparations, selecting from said aluminium-salt preparations, aluminium-salt preparations that comprise less than 350ppb of said at least one heavy metal; and combining aluminium salt of said selected preparations with said protein and water to produce said aqueous compositions. 20 In another aspect the invention provides a method for analysing the storage stability of a composition comprising aluminium and a therapeutic or prophylactic compound, said method comprising combining into a composition a pre-determined amount of therapeutic or vaccine and a pre-determined amount of an aluminium salt, said method further comprising storing said 25 composition for at least 2 weeks, preferably at least 4 weeks and preferably at least one month, at a temperature of more than 20°C, preferably at a temperature of about 22 °C and determining the stability and/or the amount of protein, preferably therapeutic or prophylactic compound, in said composition. As demonstrated in Examples 1 and 2, a temperature of 22 °C, which is higher 30 temperature compared to normal storage conditions or about 2-8°C, results in accelerated degradation of protein in an aqueous composition comprising protein, an aluminium-salt and more than 350ppb of said heavy metal based

16

on weight with respect to said composition. Thus, a temperature of about 22 °C and a storage duration of at least 2 weeks, preferably at least 4 weeks, are suitable to determine the storage stability of aqueous compositions comprising a protein according to the invention. The storage stability can be determined 5 by any method known in the art. The storage stability of an aqueous composition comprising a protein according to the invention is preferably analyzed by determining the storage stability of said protein, preferably by determining a storage sensitive epitope on said protein. For instance, as described in Example 1 and 2, the stability of a protein, preferably an antigen, 10 is determined by determining the ratio of intact storage sensitive epitope, such as intact antigenic epitope (e.g. epitope of a neutralizing epitope) content, and total protein content. "Intact storage sensitive epitope" or "intact antigenic epitope" as used herein means that degradation has not occurred within said epitope. The intact antigenic epitope content is for instance measured by 15 determining protein bound to a monoclonal antibody specifically directed against said epitope in, for example, an ELISA. The total protein content is for instance measured by determining protein bound to polyclonal antibody which is directed against various epitopes within the protein, for example by ELISA. The relative specific epitope content can then be expressed as the ratio of the 20 total antigen content determined by binding to monoclonal antibody divided by total antigen content determined by binding to polyclonal antibody. A high ratio indicates high antigenic epitope content and a low ratio indicates a low antigenic epitope content. A low ratio measured for an aqueous composition after storage at least 20°C, preferably 22°C, for at least 2 weeks, preferably 4 25 weeks, as compared to the ratio measured for said aqueous composition before storage, indicates that structural changes have taken place within the antigenic epitope. Structural changes within said antigenic epitope indicate reduced storage stability of the aqueous composition.

The heavy metal content of an aqueous composition prepared according to 30 the invention is less than 350 ppb based on the weight of the aqueous composition. Generally, the less heavy metal such aqueous composition contains, the less degradation of protein occurs. Preferably, therefore, the

17

heavy metal content is less than 325 ppb based on the weight of the aqueous composition, more preferably less than 300 ppb, more preferably less than 275, more preferably less than 250 ppb and more preferably less than 235 ppb based on the weight of the aqueous composition.

5 As used herein the term "heavy metal" refers to the total amount of elements that exhibit metallic properties and includes the transition metals, metalloids, lanthanides, and actinides. Transition metals are elements whose atom has an incomplete d sub-shell, or which can give rise to cations with an incomplete d sub-shell, and include zinc, molybdenum, cadmium, scandium, 10 titanium, technetium, palladium, vanadium, chromium, manganese, iron, cobalt, rhodium, hafnium, copper, nickel, yttrium, niobium, ziorconium, rughenium, silver, tantalum, rhenium, thungsten, osmium, meitnerium, platinum, iridium, mercury, bohrium, seaborgium, hassium. Metalloids are Boron (B), Silicon (Si), Germanium (Ge), Arsenic (As), Antimony (Sb), 15 Tellurium (Te), Polonium (Po). The lanthanides are the fifteen metallic chemical elements with atomic numbers 57 through 71, i.e. Lanthanum, Cerium, Praseodymium, Neodymium, Promethium, Samarium, Europium, Gadolinium, Terbium, Dysprosium, Holmium, Erbium, Thulium, Ytterbium and Lutetium. The actinides are the fifteen metallic chemical elements with 20 atomic numbers from 89 to 103, actinium, thorium, protactinium, uranium, neptunium, plutonium, americium, curium, berkelium, californium, einsteinium, fermium, mendelevium, nobelium and lawrencium. Preferably said heavy metal is selected from the transitional metals. In another preferred embodiment, said heavy metal is a metal having a molar mass of between 21 25 and 83, more preferably from Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V, Cr, Pb, Rb and Mo. In an even more preferred aspect, the heavy metal is selected from the heavy metals Cu, Ni, and Fe.

As demonstrated in the Examples, Aluminium hydroxide (Alum) Lot 4230 was identified to contribute significantly to antigen degradation in JEV 30 vaccine FVL09L37. In Example 3 it is shown that this Alum lot comprises at least the following metals: Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V. Higher levels of Fe, Ni and Cu ions were noted in Alum lot 4230 when compared to other

18

investigated lots. Lot 4230 was the only one where residual Cu ions were detected. Therefore, preferably said heavy metal is selected from the group consisting of Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V, more preferably from Fe, Ni and Cu.

5 The amount of heavy metal in an aqueous composition of the invention is defined herein above. This amount is typically for the total of determined heavy metals, or for the heavy metals Fe, Cr and Ni, or a combination thereof which constitute the major heavy metals by weight in the aqueous composition of the invention. For specific heavy metals different maximums amounts may 10 be preferred. For instance, it is preferred that the amount of Fe in the aqueous composition of the invention is less than 350 ppb based on the weight of the aqueous composition. In a preferred embodiment the amount of Fe is less than 250 ppb Fe based on the weight of the aqueous composition.

There is strong evidence that many of the pro-inflammatory effects of 15 aluminium adjuvants are mediated via the formation of reactive oxygen species (ROS). Aluminium can, under physiological conditions, promote the reduction of Fe(III) to Fe(II) and the oxidation of the latter. Thus the combination of Fe and Al in the adjuvant will potentiate the formation and activities of ROS (Exley, C (2010). Trends in Immunol. Vol. 31: pp 103-109). In 20 the present invention it has been found that Fe can be present in an aqueous composition of the invention without significantly affecting the storage stability of the composition. In this embodiment of the invention it is preferred that the aqueous composition of the invention comprises between 5 ppb and 250 ppb Fe based on the weight of the aqueous composition. In these amounts 25 the formation of ROS during storage due to the presence of such amount of Fe (if any ROS) does not significantly affect the storage stability of the aqueous composition as defined elsewhere herein. However, the amounts are sufficient to allow pro-inflammatory effects following administration of the vaccine in vivo.

30 In the present invention it has been found that particularly the presence of heavy metal Cu seriously affects the storage stability of the composition of the invention. In a particularly preferred embodiment an aqueous composition

19

of the invention therefore comprises less than 3 ppb Cu based on the weight of the aqueous composition. Preferably less than 2,5 ppb. In a particularly preferred embodiment said aqueous composition comprises Cu at a level that is below the detection limit of the method for the detection of copper as described 5 in the Examples.

In the present invention it has been found that particularly the heavy metal Ni affects the storage stability of the composition of the invention. In a particularly preferred embodiment an aqueous composition of the invention therefore comprises less than 40 ppb Ni based on the weight of the aqueou s 10 composition. Preferably less than 30 ppb, more preferably less than 20 pbb and more preferably less than 15 pbb Ni based on the weight of the aqueous composition. In a particularly preferred embodiment said aqueous composition comprises Ni at a level that is below the detection limit of the method for the detection of nickel as described in the Examples.

15 The heavy metal can be present in electronic neutral form or it can be ionised. Typically and preferably the heavy metal is present in ionic form in an aqueous composition of the invention.

Metal content of a composition can be determined in various ways. In one aspect, a method according to the invention comprises determining the level of 20 a heavy metal in an aqueous composition and/or the aluminium-salt present in said aqueous composition. Methods for measuring the level of one or more heavy metals in an aqueous sol ution are known in the art. Examples of such methods inductively-coupled-plasma mass spectrometry (ICP-MS), flame atomic absorption spectrometry (F-AAS), and/or graphite furnace atomic 25 absorption spectrometry (GF-AAS).

An example of an assay which can be used to determine the content of heavy metals is described in Example 3. The assay involves treating a sample of an aqueous solution containing an Aluminum-salt with concentrated HNO3 under heat until a clear solution is obtained. The clear solution can then be 30 further diluted and analyzed, for instance by ICP-MS, F-AAS and/or GF-AAS., for the presence and content of metal ions including Pb, Cd, Cr, Co, Fe, Cu, Ni, Ag, W and Al.

20

Examples 1 and 2 demonstrate that the JEV antigen shows higher stability at pH 7.5-8 as compared to pH 7. In the Examples antigen stability is expressed as the ratio of monoclonal/polyclonal ELISA. The monoclonal antibody used (clone 52-2-5) was shown to recognize a neutralizing epitope in 5 the Japanese Encephalitis Vaccine (JEV). The relative specific epitope content can be expressed as the ratio of the total antigen content determined by "monoclonal ELISA" divided by total antigen content determined by "polyclonal ELISA". Without being bound to theory, the effect of a higher antigen stability at pH 7.5-8 can be explained based on the underlying 10 assumed complex reaction chemistry of sulphites. pH might influence the related to equilibrium reaction conditions of the sulphite / formaldehyde reaction and surface charge of certain proteins/amino acid side chains accessible to modification. The pH may affect oxidation by direct influence on redox potentials of the amino acid residues and the oxidizing agents, e.g. free 15 radicals. Therefore, in one embodiment, a method according to the invention comprises buffering said aqueous composition at a pH of between 7.5 and 8.5.

Various aluminium salts are being used in compositions for administration of an individual. Aluminium adjuvant typically contains an aluminium oxide or sulphate or a combination thereof. In a preferred 20 embodiment the aluminium salt comprises aluminiumoxide (AI2O3), aluminiumhydroxide (Al(OH)g) or aluminiumphosphate (AIPO4).

The aqueous composition of the invention comprises a reactive compound. Typically though not necessarily the reactive compound is present as a result of a manipulation of the aqueous composition, for instance to treat or 25 inactivate infectious agent if any in the composition. The reactive compound can also be present for another reason. Sulphite, for instance, is sometimes present to inactivate any residual formaldehyde in the aqueous solution. Formaldehyde is typically a chemical that is often used to inactivate any infectious agent.

30 The reactive compound is a redox active compound, radical building compound and/or a stabilizing compound. In a preferred embodiment said aqueous composition of the invention comprises formaldehyde, ethanol,

21

chlorophorm, trichloroethylene, acetone, triton-X-100, deoxycholate, diethylpyrocarbonate, sulphite, Na2S20s, beta-proprio-1 acton, polysorbate such as Tween 20®, Tween 80®, O2, phenol, pluronic type copolymers, or a combination thereof.

5 Sulphite is preferably present in an amount of between 0,1 mM and 5

mM, or preferably between 0,5-2mM. Formalin is preferably present in an amount of between 0,1 mM and 5 mM, more preferably between 0,5 mM and 2mM. Oxygen is preferably present in an amount that is equivalent to the solubility of O2 at the measured temperature, O2 is preferably present in an 10 amount of between 10 and 250 uM when measured at 20 degrees Celsius.

When measured at a temperature of 0 degrees Celsius O2 is preferably present in an amount of between 10 and 400 uM. A stabilizing compound is preferably present in an amount of betweenlO and 400 uM. Similarly a redox active compound is present in an amount of between 0,1 mM and 5 mM, or preferably 15 between 0,5-2mM. A radical building compound is preferably present in an amount of between 0,1 mM and 5 mM, or preferably between 0,5-2mM. In this context and for the sake of clarity it is important to note that redox active compound, the radical building compound and/or the stabilizing compound is consumed in the production of a radical, whereas the heavy metal is indicated 20 herein above, is a catalyst in the production of a radical and is not consumed, as such. The redox active compound, the radical building compound and/or the stabilizing compound is therefore not a heavy metal.

The total amount of redox active compound, the radical building compound and/or the stabilizing compound although small in absolute 25 amounts can still be significant in relation to the antigen or protein in the aqueous composition of the invention. The antigen/protein is preferably present in an amount of between 0,1 nmol to 1 umol, more preferably between 1 nmol and 100 nmol.

The concentration of protein, preferably a therapeutic or vaccine protein, 30 in an aqueous composition comprising a protein according to the invention is preferably between 1 ng/ml and 10 mg/ml, preferably between 10 ng/ml and 1 mg/ml, more preferably between 100 ng/ml and 100 ug/ml, such as between 1

22

ug/ml and 100 ug/ml. The concentration is preferably at least 1 ng/ml to ensure that the therapeutic or vaccine protein is in a concentration sufficient to exert its therapeutic effect when administered to an individual. The concentration should, however, preferably not exceed 10 mg/ml in order to prevent or reduce 5 the occurrence of possible side effects associated with administration of said protein to an individual. In particular, the concentration of viral protein in an aqueous composition according to the invention comprising JEV is preferably between 0.01 pg/ml and 1 mg/ml, more preferably between 0.1 pg/ml and 100 ug/ml. In an exemplary embodiment, of the invention, an aqueous composition 10 according to the invention comprises about 10 ug/ml of JEV. The dose of a single administration of an aqueous composition comprising a protein, preferably a therapeutic or vaccine protein, according to the invention is preferably between 0.1 ml and 10 ml, preferably between 0.5 ml and 5 ml, such as 0.5 ml, 1 ml, 1.5 ml, 2 ml, 2.5 ml, because such dose allows for convenient 15 administration to an individual, such as a human.

A method according to the invention is preferably used to increase stability of an immunogenic composition, preferably a vaccine, comprising an aluminium-salt based adjuvant. Non-limiting examples of such vaccines are those directed against infection with Bacillus anthracis (causing Anthrax), 20 Corynebacterium diphtheriae (causing diphteria), Clostridium tetani (causing tetanus), Pseudomonas such as Pseudomonas aeruginosa, Staphylococcus such as Staphylococcus aureus or Staphylococcus epidermidis, Haemophilus influenzae type B bacteria (Hib), polio virus, hepatitis A virus, hepatitis B virus, Human Papillomavirus, influenza virus, Japanese encephalitis virus, 25 Rotavirus, Rickettsiae bacteria (causing Typhus), yellow fever virus, Varicella Zoster Virus, Meningococcus, or combinations thereof, such as, but not limited to, DTP (diphteria, tetanus, polio). An aqueous composition comprising a protein according to the invention therefore preferably comprises a protein which is a viral or bacterial protein, preferably a protein of Bacillus anthracis, 30 Corynebacterium diphtheriae, Clostridium tetani, Haemophilus influenzae type B bacteria (Hib), polio virus, hepatitis A virus, hepatitis B virus, Human Papillomavirus, influenza virus, Japanese encephalitis virus, Rotavirus,

23

Rickettsiae bacteria, yellow fever virus, Varicella Zoster Virus and/or Meningococcus. In a preferred embodiment, said protein contained in a composition comprising a protein according to the invention is a viral protein from a virus of the Flaviviridae family, preferably of a Japanese encephalitis 5 virus (JEV). As demonstrated in the Examples, the stability of aqueous compositions comprising a JEV protein, an aluminium-salt and comprising less than 350 ppb heavy metal based on the weight of the aqueous composition, and in particular wherein the amount of Cu is less than 3 ppb based on the weight of the aqueous composition, is increased as compared to aqueous 10 composition comprising more than 350 ppb of heavy metal and more than 3

ppb of Cu. Thus, a method according to the invention is particularly suitable to increase stability of an aqueous composition comprising a JEV protein.

In another preferred embodiment or aspect of the invention, said protein contained in a composition comprising a protein according to the invention is a 15 bacterial protein from a bacterium of the Pseudomonas family, preferably of Pseudomonas aeruginosa. As demonstrated in the Examples, the stability of aqueous compositions comprising Pseudomonas aeruginosa fusion protein (SEQ ID NO: l) and an aluminium-salt is reduced when more than 350 ppb heavy metal based on the weight of the aqueous composition is present. 20 An aqueous composition comprising a protein according to the invention is particularly suitable for use as an immunogenic composition or vaccine. For instance, such compositions are particularly useful for immunize an individual to treat or prevent a viral or bacterial infection. In one embodiment, the invention therefore provides a method for the treatment of an individual 25 comprising obtaining an immunogenic aqueous composition comprising a protein and an aluminium-salt, said aluminium-salt having less than 350 ppb heavy metal based on the weight of the aqueous composition, preferably less than 3 ppb of Cu, and administering the immunogenic aqueous composition to an individual in need thereof. Also provided is a method for the prophylactic 30 treatment of an individual comprising obtaining an immunogenic aqueous composition comprising a protein and an aluminium-salt, said aluminium-salt having less than 350 ppb heavy metal based on the weight of the aqueous

24

composition, preferably less than 3 ppb of Cu, and administering the immunogenic aqueous composition to an individual in need thereof. Further provided is a method for inducing and/or boosting an immune response towards an antigen in an individual, said method comprising obtaining an 5 aqueous composition comprising a protein comprising said antigen and an aluminium-salt, said aluminium-salt having less than 350 ppb heavy metal based on the weight of the aqueous composition, preferably less than 3 ppb of Cu, and administering the aqueous composition to an individual in need thereof. In another aspect the invention provides a method for immunizing an 10 individual comprising administering to said individual at least two immunogenic compositions at an interval of at least two weeks between each administration, and wherein each of said at least two immunogenic compositions compri se the same antigen, and wherein at least one of said immunogenic compositions further comprises an aluminium salt having less 15 350 ppb heavy metal based on the weight of the aqueous composition,

preferably less than 3 ppb of Cu, and administering the immunogenic aqueous composition to an individual in need thereof. Nucleic acid compositions are sometimes also administered together with aluminium. Thus for the present invention it is possible to replace "protein" in an aqueous composition of the 20 invention with nucleic acid. Thus in one embodiment the invention provides a method for preparing an aqueous composition comprising aluminium and a nucleic said method comprising - combining an aluminium-salt, said nucleic acid and water to produce said aqueous composition and - determining the level of a heavy metal in the aqueous composition and/or the aluminium-salt. 25 The invention also provides a method for preparing an aqueous composition comprising aluminium and a nucleic acid said method comprising - preparing or selecting an aluminium-salt having less 350 ppb heavy metal based on the weight of the final aqueous composition, preferably less than 3 ppb of Cu, and administering the immunogenic aqueous composition to an individual in need 30 thereof and

- combining said alumini um salt, said nucleic acid and water to produce said aqueous composition. In a preferred embodiment said methods further

25

comprising buffering said aqueous composition at a pH of between 7.5 and 8.5. In a particularly preferred embodiment said methods, further comprise packaging aliquots of said aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition in separate air-5 tight storage containers. The nucleic acid may be administered for therapeutic purposes. For instance, in the form of an antisense RNA, RNAi or mimic thereof. The nucleic acid may also be administered in the form of an infectious agent, typically a virus or a modified virus, as is the case in many gene therapy approaches. In that case the nucleic acid is enclosed in a particle that 10 comprises protein. The invention thus further provides an aqueous composition comprising a nucleic acid and an aluminium-salt, said composition comprising less than 350 ppb heavy metal based on the weight of the aqueous composition.

The invention is further explained in the following examples. These 15 examples do not limit the scope of the invention, but merely serve to clarify the invention.

References

20 Alipazaga MV, Moreno RGM, Coichev N. 2004. Synergistic effect of Nifll) and

Co(II) ions on the sulphite induced autoxidation of Cu(II)/tetraglycine complex. Dalton Trans 13:2036-2040.

Arunee Wittayanukulluk, Dongping Jiang, Fred E. Regnier, Stanley L. Hem, "Effect of microenvironment pH of aluminum hydroxide adjuvant on the chemical 25 stability of adsorbed antigen", Vaccine 22 (2004) 1172-1176

Berglund J, Fronaeus S, Elding LI. 1993. Kinetics and mechanism for manganese-catalyzed oxidation of sulfur(IV) by oxygen in aqueous solution. In org Chem 32:4527-4538.

Brandt C, Elding LI. 1998. Role of chromium and vanadium in the atmospheric 30 oxidation of sulfur (IV). Atmos Environ 32(4):797-800.

Exley, C (2010). Trends in Immunol. Vol. 31: pp 103-109.

26

Ito, Kimiko and Kawanashi, Shosuke. Site-specific fragmentation and modification of Albumin by sulphite in presence of metal ions or peroxidase/ThCh:

Role of Sulphate radical. Biochem and Biophys Res Comm., 1991, 176,1306-1312

Huie R.E., Neta P. One-electron redox reaction in aqueous solutions of sulphite 5 with hydroquinone and other hydroxyphenols. J. Phys. Chem., 1985, 89 (18), 3918-3921

Kalina Ranguelova, Marcelo G. Bonini, and Ronald P. Mason: (Bi)sulphite Oxidation by Copper,Zinc-Superoxide Dismutase: Sulphite- Derived, Radical-

Initiated Protein Radical Formation. Environmental Health Perspectives 2010, 118 10 (7), 970-975

Lampeth D.O., Palmer G. The kinetics and mechanism of reduction of electron transfer proteins and other compounds of biological interest by dithionite. J.Biochem. Chem. 1973, 248, 6095-6103

Li S, Schoneich C, Borchardt RT. Chemical instability of protein 15 pharmaceuticals: Mechanisms of oxidation and strategies for stabilization. Biotechnol Bioeng. 1995 Dec 5;48(5):490-500

Lindblad, EB (2004) Immunol, and Cell Biol. Vol 82: 497-505.

Lima S, Bonifacio RL, Azzellini GC, Coichev N. 2002. Ruthenium(II) tris(bipyridyl) ion as a luminescent probe for oxygen uptake on the catalyzed 20 oxidation of IISO:r. Talanta 56:547-556.

Mayo JC, Tan DX, Sainz RM, Natarajan M, Lopez-Burillo S, Reiter RJ. Protection against oxidative protein damage induced by metal-catalyzed reaction or alkylperoxyl radicals: comparative effects of melatonin and other antioxidants. Biochim Biophys Acta. 2003 Mar 17; 1620(1-3): 139-50.

25 Neta P., Huie R.E.: Free Radical Chemistry of Sulphite. Environmental Health

Perspectives 1985, 64, 209-217

Shi X. 1994. Generation of ^SOs" and OH radicals in SQ32~ reactions with inorganic environmental pollutants and its implications to SO32" toxicity. J Inorg Biochem 56(3): 155-165.

30 Stadtman ER. Metal ion-catalyzed oxidation of proteins: biochemical mechanism and biological consequences. Free Radic Biol Med. 1990;9(4):315-25.

Brief description of the drawings

27

Figure 1: RP-HPLC elution profiles of chlorobutyl stopper extract (1:4 diluted) and JEV09L37 SN.

Figure 2: SEC HPLC elution profiles of PS (2mg/mL) before and after trypsin cleavage.

5 Figure 3: SEC HPLC elution profiles of Trypsin treated PS and degraded PS as present in NTV11A74. Note that elution profiles were normalized to similar peak height to allow better comparison.

Figure 4: DOE evaluation of the ratio monoclonal/polyclonal ELISA by Pareto chart analysis and main effects plots (4 weeks at 22°C).

10 Figure 5: DOE evaluation of the ratio monoclonal/polyclonal ELISA by Pareto chart analysis and main effects plots (8 weeks at 22°C).

Figure 6: Contour plot of the estimated response Figure 7: Residual plot of the estimated response

Figure 8: ELISA ratio (monoclonal/polyclonal) for JEV formulations at pH 7 in 15 presence of Ni, Cu and Cr. Samples were stored at 22°C for 5 weeks

Figure 9: Summary of results obtained after 7 weeks at 22°C. Shown are the raw data of ratio as a function of pH and metal ion type and combined results for each parameter.

Figure 10: Mean ratio of DP formulations prepared with different Alum lots. Samples 20 were stored for 6 weeks at 22°C. Error bars represent the 95% confidence interval calculated based on pooled standard deviation. Samples from left to right: Alum 3877, Alum 4074, Alum 4230 nonGI, Alum 4230 GI, Alum 4470, Alum 4563, Alum 4621, Alum Mix 4074_4230.

Figure 11: Particle size distribution of Alhydrogel® samples.

25 Figure 12: Alhydrogel® titration curves in PBS. ■ non-irradiated AlOH (RQCS0890), A GI AlOH (RQCS1200), ♦ GI AlOH (RQCS1342), + GI AlOH (RQCS0448)

Figure 13: Overview of tested Alhydrogel® batches. Total concentration of contaminating metal ions in ng/mL and share of the main metal ions Fe, Cr and Ni are shown.

30 Figure 14: Comparison of metal ion contamination in analyzed Ammonia Alum raw material batches; lots 6-80578-28, AA0427 and 91480 were used in JEV vaccine production.

Figure 15: Amino acid sequence of Ala-(His)6-OprF 190-342 -OprI21-83 (SEQ ID NO: l) - herein also referred as "protein A".

28

Examples

Example 1

5 Aluminium hydroxide (Alum) Lot 4230 was previously identified to contribute significantly to antigen degradation in FVL09L37. In this particular Alum lot much higher residual metal ion content was observed compared to other Alum lots used for formulation of the inactivated JEV antigen. This Example demonstrates additional studies carried out to further identify the underlying root-cause mechanism and 10 influence of metal ions on degradation pathway of JEV. Design of experiments (DOE) was performed to work out the influence of individual parameters on antigen stability.

Parameters tested in a 25 full factorial DOE were 15 • Aluminium hydroxide Lot 4230 vs. Aluminium hydroxide Lot 4074

• Presence of excess Protamine sulfate fragments

• Presence of leachables from chlorobutyl rubber stopper

• pH range 7 to 8

• Residual formaldehyde content

20

Alum lot 4230 contains much higher levels of residual metal ion impurities compared to other Alum lots used for formulation of JEV. A "Design-of-Experiment" (DOE) was selected to further investigate the potential root cause mechanism and interaction of parameters that finally could lead to product degradation. In factorial designs, 25 multiple factors are investigated simultaneously during the test. As in one factor designs, qualitative and/or quantitative factors can be considered. The objective of these designs is to identify the factors that have a significant effect on the response, as well as investigate the effect of interactions (depending on the experiment design used). Predictions can also be performed when quantitative factors are present, but 30 care must be taken since certain designs are very limited in the choice of the predictive model. For information about DOE in general see (Siebertz, Karl; van Bebber, David, Hochkirchen, Thomas: Statistische Versuchsplanung: Design of Experiments (DoE). Publisher: Springer Berlin Heidelberg; 1st Edition (2010), ISBN-10: 3642054927).

29

1.1 DOE Study Design

1.1.1 Definition of Parameters and levels for DOE Design

The following parameters and levels were taken into consideration for designing an 5 appropriate DOE experiment:

• Residual metal ion content of Alum: Aluminium hydroxide Lot 4230 and Lot 4074 were selected as representative of the two extremes quality with regard to residual metal ions content of Aluminium hydroxide. The center point was a mixture of 50/50% of both Alum lots. Initial analysis for remaining metal ion impurities in 2%

10 Aluminium hydroxide stock solution by ICP-MS showed significant differences in Cr, Fe, Ni and Cu ion content between these two lots (see Table 1).

• Protamine Sulphate fragments: Protamine sulfate (PS) fragments are present at low quantity (<5pg/mL) in the final vaccine lot. It was tested if PS fragments could contribute to virus surface modification (e.g. interaction/ covalent

15 linkage to the virus surface proteins) in combination with Alum and other factors used in this study. Therefore a stock solution of PS fragments was prepared by digestion with Trypsin followed by heat inactivation and ultrafiltration using a 5kDa membrane for protease inactivation and removal of the enzyme. This stock solution was used for spiking additional PS fragments into the respective formulations at the 20 high level of 50 pg/mL. In low level samples no additional PS fragments were spiked and the actual level in formulations was < 5 pg/mL according to HPLC analysis.

pH: Lower and upper level of pH in formulations was 7 and 8 with the center point at pH 7.5.

• Leachables/Extractables from syringe plunger: Syringe plunger (made of 25 chlorobutyl PH701/50 black) that are currently used in the container closure system.

It was tested if leachables from the chlorobutyl rubber in the formulation could contribute to antigen modification. Therefore a stock solution of leachables was prepared and used for spiking experiments. The high level of spiked leachables in formulation was estimated to be on average 1.4 x higher compared to commercial 30 Final Vaccine Lot (FVL). Due to the harsh extraction conditions additional peaks were detected not present in FVL samples. Therefore the spiked formulations represent a "worst case" with regard to leachables and extractables. Formulations at the low level did not contain any leachables from chlorobutyl rubber.

30

• Residual formaldehyde: For low level formulations no additional formaldehyde was spiked into the formulation samples. The lower level was the residual formaldehyde that was still present in diluted N1V sample after inactivation/neutralization and 2-fold dilution was in the range of approx. 37 ppm 5 (recalculated from commercial DS release GMP analytical certificate). For the high level additional 40 ppm formalehyde were spiked into the corresponding formulation (total final content approx. 77 ppm). It was tested if residual formaldehyde in combination with higher level of metal ions present in Alum 4230 and possible other factors could further react with the virus leading to hyper -cross linking of surface 10 proteins and loss of relevant epitopes.

Determination of other residual process related impurities:

Residual formaldehyde, sulphite and sucrose in final formulations were estimated based on GMP certificates for commercial drug substance JEV11A74. Results were 15 recalculated by the actual 2-fold dilution of NIV to DS used within the DOE experiments.

Residual Sulphite: The concentration of residual sulphite was constant in all formulations with approx. 93 ppm.

Residual sucrose: The concentration of residual sucrose was constant in all 20 formulations with approx. 1% v/w.

1.1.2 DOE Design

These 5 factors were combined in a 25 DOE plan resulting in a total number of 34 experiments including 2 center points with the following base design. DOE planning 25 and evaluation were carried out with an appropriate software (Statgraphic Plus 3.0) Base Design: Factorial 25 Number of experimental factors: 5 Number of blocks: 1 Number of responses: 1 30 Number of centerpoints per block: 2 Number of runs: 34 Error degrees of freedom: 18 Randomized: Yes

31

Factors Continuous1'

Low

High

Units pli 7.0

Yes

5 Alum2) 0.0

Yes

Spiked PS Fragments3 0 No

Spiked leachables 4) 0

10 Spiked Formaldehyde55 0 No

8.0

100.0

50

Alum 4230

pg/mL

1.4 relative content compared to FVL No 40 ppm

Responses

Units

15

20

25

30

ELISA (monoclonal, polyclonal) of desorbed antigen AU/mL

1) Continuous means that a center point (mean value from high and low level) is present in the study design. No Center point means that only low and high levels are present in the study design.

2) Low level (0%) means that formulation was prepared with Alum Lot 4074. High level (100%) means that formulation was prepared with Alum lot 4230. For center point formulations an equal mixture (50/50%) of both Alum lots was used.

3) Since PS is present in NIV used for preparation of drug product samples, actual PS concentration in non-spiked formulations was <5pg/mL and ~50-55pg/mL for PS-spiked formulations.

4) No leachables were assumed in non-spiked formulations since samples were prepared/stored in low-bind Eppendorf tubes. Total content of leachables from chlorobutyl rubber syringe plunger in spiked samples was approx. 1.4 times higher compared to FVL.

5) Actual formaldehyde concentration in non-spiked DP samples was approx. 37ppm, total formaldehyde concentration in spiked samples was approx.77ppm.

2 Definitions & Abbreviations

AcCN Acetonitrile

DOE Design of experiments

DS Drug substance

32

10

FVL Final Vaccine lot

HPLC High performance liquid chromatography

ICP-MS Inductively-coupled-plasma mass spectrometry

PS Protamine sulphate

RP Reversed Phase

SEC Size exclusion chromatography

SN Supernatant

TFA Trifluoroacetic acid w/o without

3 Materials and Methods

3.1 DOE Studies

3.1.1 Materials

• Syringe plunger stoppers: PH701/50/C black Sil6 7002-1051 (obtained from 15 West, Order No. 2116)

• 100 mL Glass bottle (Schott) + Teflon coated screw cap

• Aluminium foil

• HQ water

• Electrical water bath (IKA, HBR 4 digital)

20 • LoBind Eppis 2ml (Eppendorf, Cat. No. 0030 108.132)

• Speed Vac (Christ, RVC-2-25)

• HPLC vials, clear glass, 900 pL, Chromacol (VWR, Cat.no.: 548-1124)

• HPLC vials, PP, 900 pL, (Agilent, Item no. 51.82-0567)

• Caps for HPLC vials, pre-cut (VWR, Cat.no.: 548-1260)

25 • 15ml Falcon tubes (Greiner, Cat. No. 188724)

• Alum batch 4470 (RQCS 1342); Alum Lot 4230 (RQCS 1200)

• lOxPBS (Gibco, Order No. 14200-091)

• Parafilm

• Waters Atlantis T3 column; 3pm particle diameter; column diameter/length 30 2.1x100mm (Order No. 186003718; Lot 0107372331)

• Acetonitril (Merck, Cat No. 1.13358.2500)

• TFA (Sigma, Order No. 302031

• HPLC system Dionex 3000

• Solvent Rack SR-3000

33

• Pump UltiMate-3000, analytical low pressure gradient pump

• Autosampler WPS-3000 TSL, analytical autosampler - temperature controlled

• Column compartment TCC-320Q. temperature controlled

• PDA-Detector PDA-3000

5 • Formaldehyde solution 37% (Merck, Cat No. 1.040031000)

• Protamine sulphate (Intercell Biomedical Ltd, Batch no. 086056)

• Ultrafilatrion device (Amicon® Ultra 3 kDa) (Millipore, Cat No. UFC900324)

• Incubator Infors HT Incubator Multitron Standard (InforsAG)

• Trypsin (Sigma, Order No: T0303)

10

3.1.2 Procedure for preparation of extractables from syringe plunger Syringe plunger (made of chlorobutyl PH701/50 black) that are currently used in the FVL container closure system were obtained from West (Germany). Therefore a stock solution of leachables was prepared by heat treatment of syringe plunger in water 15 (90°C/2h) followed by concentration in a speed-vac. The relative content of leachables in this stock solution was estimated by RP-HPLC using a C18 column (Atlantis T3 column) and compared to FVL JEV09L37 supernatant.

Extraction method

20 A 100 mL Schott glass bottle with a Teflon coated screw top and a piece of aluminum foil were washed with hot water and thoroughly rinsed with HQ water. 30 stoppers were filled in the bottle and 30ml of HQ-water were added. The bottle was closed with the aluminum foil fitted between bottle and screw and sealed additionally with Parafilm. The bottle was heated in the water bath to 90°C for 2 hours and allowed to 25 cool to room temperature. The extract was transferred to 14 low-bind Eppendorf tubes (a total of 28 mL extract was recovered). Twelve vials (total of 24 mL) were concentrated in a Speed Vac for approximately 44 hours and pooled into a falcon tube to obtain 6 ml of 4x concentrated stopper extract. A control sample containing 30 mL HQ water w/o stopper was prepared in the same way to evaluate any possible 30 contamination.

C18 RP-HPLC method

Leachables were separated by RP-HPLC C18 column (Atlantis T3) operated at 40°C and 0.25 mL/min. Solvent A was 0.1% TFA in H20, solvent B was 0.1% TFA in AcCN.

34

Separation was performed by linear gradient ranging from 0 to 95% B in 30 min. Detection was done at 214 nm, 254 nm and 280 nm. The total relative concentration of concentrated stopper extract was estimated to be 80 fold higher compared to peaks detected in Final Vaccine Lot supernatant (FVL SN; obtained by removal of Alum 5 particles by centrifugation at 5000g/5min) as detected at 254 nm. Therefore a total relative content of 80 U/mL (arbitrary Units U) were assigned for the stock solution, whereas the total relative concentration of leachables in FVL SN was set to lU/mL. For DOE studies, the stock solution was diluted 16-fold into the respective formulations yielding approx. 5 U/mL of total extractables.

10

3.1.3 Preparation of Protamine Sulfate fragments

A stock solution of PS fragments was prepared by digesting a PS solution (2mg/mL in PBS) with Trypsin (200 ng/mL for 60 min at 37 °C). The enzyme was subsequently inactivated by heat (90°C for 10 min) followed by ultrafiltration using a 3kDa 15 membrane (Amicon® Ultra centrifugal filter). Due to the cut-off of the membrane Trypsin remained in the retentate, whereas the PS fragments were present in the permeate. Complete inactivation of the enzyme was evaluated by spiking 500 pg/mL of full length PS into an aliquot of the obtained PS fragment followed by incubation at 37°C for 18 h. No degradation of full length PS was observed indication complete 20 inactivation/removal of Trypsin. Degradation was monitored by PS-SEC HPLC.

3.1.4 DOE Plan

Samples were prepared according to the pipetting scheme as shown in Table 2. NIV Batch JEV11A74 obtained from a commercial production run was used as starting 25 sample. NIV was diluted 2-fold to DS using PBS buffer followed by pH adjustment. 5 mL aliquots were removed and adjuvanted with the corresponding Alum lot 4230, 4074 or a 50/50% mixture of both. The final amount of Alum stock (2% A1203) added was 500pg/mL Aluminium (0.1% A1203). Each formulation (5 mL) was split into two parts (2 x 2.5 mL) using Lo-bind Eppendorf tubes. One aliquot was stored at 2-8°C, 30 another aliquot stored at 22 ± 1°C (Infors HT Incubator) under gentle shaking (20 rpm).

3.2 Inactivated JEV ELISA (polyclonal based)

Desorption of the antigen from Alum and ELISA analysis was carried out using

35

polyclonal sheep anti JEV antibodies for coating the 96 well ELISA plates as described in Example 4.

3.3 Inactivated JEV ELISA (monoclonal based)

5 A monoclonal (mAb) based JEV ELISA was developed. The assay is primarily based on the "polyclonal JEV ELISA" assay format, only a monoclonal anti-JEV antibody (clone 52-2-5) is used for coating. The employed mab 52-2-5 was shown to be specific for JEV and to recognize a neutralizing epitope. Mab clone 52-2-5 was obtained by subcutaneously immunizing BALB/c mice with commercially available vaccine lot 10 JEV08 J14B. Spleen cells of the mice were fused to myeloma cells. From resulting hybridoma cells single clones were selected and sub-cloned. The clones were negatively screened against Bovine Serum Albumin, Protamine sulphate and an extract of the production cell line of the JE-vaccine (Vera cells). A positive screen was done against Neutralized Inactivated Virus (NIV) of vaccine lot JEV08M20. For 15 screening, microtiter plates were coated with the relevant antigen and reacted with supernatant of cultures of the selected clones. For detection a goat anti mouse polyclonal antibody conjugated with alkaline phosphatase was used. Mab clone 52-2-5 was shown to recognize a neutralizing epitope on domain III of the envelope (E) protein of JEV containing Ser331 and Asp332 (Lin C.-W. and Wu W.-C. J Virol. 2003 20 ;77(4):2600-6). Binding of the mab to the indicated neutralizing epitope is for instance determined as described in Lin and Wu (2003) by site-directed mutagenesis of the domain III at position 331 (for instance: S—>R), and/or by alanine mutations at or near position 331 of domain III, for instance of residues Ser 331 and Asp332, followed by immunoblots to determine binding of the mab to the mutated proteins. Negative 25 binding results indicate that the epitope of the mab is the neutralizing epitope identified by Lin and Wu (2003). The neutralizing characteristic of the epitope gives rise to the assumption that the epitope might be of importance for the antigen to elicit a protective immune response.

JEV samples were analyzed by both ELISA assays, polyclonal and monoclonal. The 30 relative specific epitope content can be expressed as the ratio of the total antigen content determined by "monoclonal ELISA" (clone 52-2-5) divided by total antigen content determined by "polyclonal ELISA". Any differences in the ratio may indicate differences in specific epitope content 52-2-5. Results close to 1 would correspond to high epitope contents, and results close to 0 correspond to low relative epitope

36

content. A low ratio indicates presence of structural changes of the neutralizing epitope.

In the course of development of this "mAb ELISA", differences between vaccines lots were detected, which could be correlated with potency results of these lots.

5

3.4 Protamine sulfate SEC-HPLC

PS (full length) and its fragments were analyzed by size-exclusion HPLC (SEC-HPLC) using a Superdex Peptide 10/300 GL, 10 x 300 mm, 13 pm (GE Healthcare) using 0.1 % (v/v) Trifluoroacetic acid (TFA) in 30 % acetinitrile (CAN) as mobile phase at a flow 10 rate of 0.6 mlVmin. PS containing samples were prepared in duplicated, i.e. diluted with mobile phase before injection.

4 Results

4.1 Analysis of Stopper Leachables used for spiking Experiments

15 RP-HPLC elution profiles of concentrated stock solution obtained after extraction of stoppers under heat compared to FVL SN is shown in Figure 1. Similar peak pattern as observed for both samples. Due to the harsh extraction conditions additional peaks were detected in the concentrate that were not present in FVL samples or present only at a much lower relative content Therefore the spiked formulations represent a 20 "worst case" with regard to leachables and extractables. The total relative content of individual peaks in extract concentrate and spiked formulation in comparison to FVL SN is summarized in Table 3. The total amount of leachables in the stock solution was calculated as the sum of all peaks detected and expressed in arbitrary units as 67 U/mL. Since the stock solution was diluted 16 times into the respective formulation, 25 the resulting total content of leachables was estimated as 4.2 U/mL. This corresponds on average 1.4 fold increase compared to FVL JEV09L37 supernatant (3.0 U/mL).

4.2 Analysis of Protamine sulphate fragments

PS fragments obtained after cleavage of full length PS by Trypsin are shown in Figure 30 2. Similar peak profiles of Trypsin treated PS and already degraded PS present in NIV11A74 were obtained by HPLC (see Figure 3).

4.3 DOE Evaluation

Formulations prepared for this DOE were analyzed after 4 weeks and 8 weeks of 35 incubation at accelerated conditions (22°C). It was considered that any degradation

37

reaction would be accelerated when stored at higher temperature compared to normal storage conditions (2-8°C). However, samples are still stored at 2-8°C and will be analyzed on a later time point (-4-6 month). First analysis of samples stored at 22°C for 4 and 8 weeks are shown in the Table 4.

5

4.3.1 DOE Evaluation after 4 weeks at 22°C

Statistical evaluation of the DOE matrix results obtained after 4 weeks at 22°C showed that the specific epitope content 52-2-5 (expressed as the ratio of desorbed antigen analyzed by monoclonal/polyclonal ELSIA) was statistically significant 10 influenced (95% confidence level, see Table 5) by the following factors:

• lower specific epitope content 52-2-5 in presence of Alum lot 4230

• lower specific epitope content 52-2-5 at lower pH 7

• Higher specific epitope content 52-2-5 at increased concentration of formaldehyde

15 Presence of higher concentration of PS fragments and chlorobutyl rubber leachables did not show any influence on specific epitope content. No 2nd or higher order interactions between individual parameters were detected.

The ANOVA table partitions the variability in "Ratio 4 weeks" into separate pieces for 20 each of the effects. It then tests the statistical significance of each effect by comparing the mean square against an estimate of the experimental error. In this case, 3 effects (Alum, pH, Formaldehyde) have P-values less than 0.05, indicating that they are significantly different from zero at the 95.0% confidence level. The R-Squared statistic indicates that the model as fitted explains 74.85% of the variability in Ratio 4 weeks. 25 The adjusted R-squared statistic, which is more suitable for comparing models with different numbers of independent variables, is 51.27%. The standard error of the estimate shows the standard deviation of the residuals to be 0.063. The mean absolute error (MAE) of 0.0353 is the average value of the residuals. The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based 30 on the order in which they occur in your data file. Since the DW value is greater than 1.4, there is probably not any serious autocorrelation in the residuals. Effects are also displayed by standardized Pareto chart and main effect plots as shown in Figure 4.

38

4.3.2 DOE Evaluation after 8 weeks at 22°C

Statistical evaluation of the DOE matrix results obtained after 8 weeks at 22°C were similar to results obtained after 4 weeks. Evaluation shows that the specific epitope content 52-2-5 (expressed as the ratio of desorbed antigen analyzed by 5 monoclonal/polyclonal ELSIA) was statistically significant influenced (95% confidence level, see Table 6) by the following factors:

• lower specific epitope content 52-2-5 in presence of Alum lot 4230

• lower specific epitope content 52-2-5 at lower pH 7

Presence of higher concentration of PS fragments, chlorobutyl rubber leachables and 10 formaldehyde did not show any influence on specific epitope content. Note that P-

value for formaldehyde (P=0.08) is quite close to be significant. No 2nd or higher order interactions between individual parameters were detected.

The ANOVA table partitions the variability in "Ratio 8 weeks" into separate pieces for 15 each of the effects. It then tests the statistical significance of each effect by comparing the mean square against an estimate of the experimental error. In this case, 2 effects (Alum and pH) have P-values less than 0.05, indicating that they are significantly different from zero at the 95.0% confidence level. The R-Squared statistic indicates that the model as fitted explains 75.9% of the variability in "Ratio 8 weeks". The 20 adjusted R-squared statistic, which is more suitable for comparing models with different numbers of independent variables, is 53.3%. The standard error of the estimate shows the standard deviation of the residuals to be 0.095. The mean absolute error (MAE) of 0.057 is the average value of the residuals. Effects are also displayed by standardized Pareto chart and main effect plots as shown in Figure 5.

25

Regression analysis was also performed to the fitted data and calculated regression coefficients are shown in Table 7. The regression equation is displayed below which has been fitted to the data including pH, Alum and Formaldehyde. The equation of the fitted model is 30 "Ratio 8 weeks" = 0.0228125 + 0.113125:|<pH - 0.00185625*Alum +0.0315625*Formaldehyde where the values of the variables are specified in their original units, except for the categorical factors which take the values -1 for the low level and +1 for the high level. The contour of the estimated response and residual plot is shown in Figure 6 and

39

Figure 7. The ratio increases when relative Alum content Lot 4230 decreases and pH increases.

Table 8 contains information about values of "Ratio 8 weeks" generated using the 5 fitted model. The table includes:

(1) the observed value of "Ratio 8 weeks"

(2) the predicted value of "Ratio 8 weeks" using the fitted model

(3) 95.0% confidence limits for the mean response

10 As shown the experimental results are well predicted by the regression model. 5 Summary

Out of the parameters tested, Alum lot 4230 was shown to contribute significantly to antigen degradation as analyzed by monoclonal/polyclonal ELISA under accelerated 15 conditions (22°C, testing time points 4 and 8 weeks). DOE results obtained after 4 and 8 weeks at 22°C show that Alum 4230 is the most significant factor with regard to antigen degradation as detected by the ratio of monoclonal/polyclonal ELISA. Formulations made with Alum 4074 (much higher purity with regard to residual metal ions) show in general much higher specific epitope content. 20 Formalaldehyde and pH also contributed to antigen stability, but to a lower extent. The effect of increased antigen stability in samples formulated with Alum 4230 at higher formaldehyde level was well demonstrated (e.g. samples #19 and 29). However, influence of formaldehyde was less pronounced after extended storage period (8 weeks at 22°C).

25 Better stability of the antigen was observed at pH 8 compared to pH 7.

Protamine sulphate and leachables from chlorobutyl rubber stopper did not contribute to the antigen degradation.

30

40

Example 2

In previous studies (see Example 1) Aluminium hydroxide Lot 4230 was identified a significant contributing factor to the observed antigen degradation in FVL09L37. In 5 this particular lot of Aluminium hydroxide (Alum), a much higher residual metal ion content was observed compared to other Alum lots used for formulation of the inactivated JEV antigen. This Example summarizes additional studies carried out to evaluate the influence of metal ions on stability of inactivated JEV. Spiking studies were conducted with the antigen either present in inactivated neutralized virus (NIV) 10 solution or in drug product (DP) suspension following formulation of the antigen with Aluminium hydroxide.

1 Study Description

It was previously shown that Alum lot 4230 contains much higher levels of residual 15 metal ion impurities compared to other Alum lots used for formulation of JEV (see also Example 3). Additional studies were performed to evaluate the influence of metal ions on the stability and on a potential surface modification of JEV. The inactivated antigen was either present in neutralized inactivated virus (NIV) solution or in drug product (DP) suspension following further dilution of NIV and formulation with 20 Aluminium hydroxide. In another set of experiments, different Alum lots covering a broad content range of residual metal ions were used and formulated with a single defined NIV lot.

All of these formulations still contained residual formaldehyde and bisulphite at representative concentrations compared to commercial product. Stock solution of 25 metal salts were dissolved in water and spiked to the samples to the desired final concentration.

2 Definitions & Abbreviations

AcCN Acetonitrile 30 ANOVA Analysis of variance DOE Design of experiments DP Drug product DS Drug substance FBV Final bulk vaccine

41

FVL Final Vaccine lot

GI Gamma irradiated

HPLC High performance liquid chromatography

LSD Fisher's least significant difference

5 mAb Monoclonal antibody

NIV Neutralized inactivated virus

PS Protamine sulphate

RP Reversed Phase

SEC Size exclusion chromatography

10 SN Supernatant

TFA Trifluoroacetic acid w/o without

3 Materials and Methods

15 3.1 Materials

• Iron(II)chloride tetrahydrate(Sigma, Order no. 44939)

• Iron(III)chloride hexahydrate (Sigma, Order no. 31232)

• Nickle(II)sulphate hexahydrate (Sigma, Order no. N4882)

• Cobalt(II)chloride hexahydrate (Sigma, Order no. 31277) 20 • Copper(II)chloride dehydrate (Sigma, Order no. 807483)

• Zinksulphate heptahydrate (Sigma, Order no. 24750)

• Crom(III)chloride hexahydrate (AlfaAesar, Order no. 42114)

• Ethylenediaminetetraacetic acid disodium salt dehydrate (EDTA) (Sigma, E5134)

25 • Aqua bid est. (Fresenius Kabi, Art no. 0712221/01 A)

• lOxPBS (Gibco, Order No. 14200-091)

• Formaldehyde solution 37% (Merck, Cat No. 1.040031000)

• Protamine sulphate (Intercell Biomedical Ltd, Batch no. 086056)

• LoBind Eppis 2 ml (Eppendorf, Cat. No. 0030 108.132)

30 • 15 mi Falcon tubes (Greiner, Cat. No. 188724)

• Incubator Infors IIT Incubator Multitron Standard (InforsAG)

• 0.2 pm filter Mini Kleenpak 25 mm (Pall)

42

• NIV11A74 and Final bulk vaccine (FBV, formulated with Alum lot 4539) JEV 11D87 from commercial production runs was obtained from Intercell Biomedical (Livingston, UK) and stored at 2-8°C until further processing

• Stock solutions of metal salts in water (final concentration ImM) used for 5 spiking experiments were prepared and stored at 2-8°C until usage

• Aluminium hydroxide samples (2% A1203, Brenntag Biosector) were either retain samples obtained from Intercell Biomedical or purchased directly from Brenntag. Alum samples were stored at 2-8°C. The following Alum lots were used in this study: 4470, 4563, 4621, 3877, 4230 (non gamma irradiated and gamma

10 irradiated)

3.2 Preparation of metal stock solutions

3.2.1 Iron(II) stock solution

15 20 mM Iron(II) stock solution was prepared by dissolving 397 mg of Iron(II)chloride tetrahydrate in 100 mL aqua bidest.

3.2.2 Iron(III) stock solution

20 mM Iron(III) stock solution was prepared by dissolving 540 mg of Iron(III)chloride 20 hexahydrate in 100 mL of aqua bidest,

3.2.3 Nickle(II) stock solution

20 mM Nickle(II) stock solution was prepared by dissolving 525 mg of Nickle(II)sulphate hexahydrate in 100 mL of aqua bidest.

25

3.2.4 Cobalt(II) stock solution

20 mM Cobalt(II) stock solution was prepared by dissolving 476 mg of Cobalt(II)chloride hexahydrate in 100 mL of aqua bidest.

30 3.2.5 Copper(II) stock solution

20 mM Copper(II) stock solution was prepared by dissolving 341 mg of Copper(II)chloride dihydrate in 100 mL of aqua bidest.

3.2.6 Zink stock solution

43

20 mM Zink stock solution was prepared by dissolving 575 mg of Zinksulphate heptahydrate in 100 mL of aqua bidest.

3.2.7 Crom(III) stock solution 5 20 mM Crom(III) stock solution was prepared by dissolving 533 mg of Crom(III)chloride hexahydrate in 100 mL of aqua bidest.

3.3 Preparation of working solutions

Working solutions of metal ions (1 mM final concentration if not otherwise stated) 10 were prepared by dilution of metal ion stock solutions with aqua bidest and sterile filtration via 0.2 pm syringe filter.

3.4 Preparation of Formulation

All formulations were prepared under sterile conditions. NIV and FBV obtained from 15 commercial production runs were adjusted to the desired pi I and spiked with aliquots of metal stock solution. All samples were stored in plastic tubes if not otherwise stated. In all formulations using Alum, the final Al content was 500pg/mL, corresponding to 0.1% A1203. It has to be noted that metal ions, especially iron (II), iron (III) and to a certain extent Cr (III), formed a precipitate with the phosphate ions 20 present in the buffer resulting in partial co-precipitation of the inactivated virus represented by the low recovery determined by size-exclusion HPLC (SEC-HPLC).

3.4.1 Experiment 20110913(NIV): NIV Formulation at different metal ion concentration of Ni(II), Cu(II), Cr(III) with or w/o presence of PS fragments

25

NIV 11A74 was adjusted to pH 7 and pH 8 followed by spiking of metal ions (Ni(II), Cu(II), Cr(III)) at 100/500/1000 ng/mL final concentration. All formulations were stored in low-bind Eppendorf tubes at 22°C. Aliquots of all formulation were also prepared in presence of protamine sulphate fragments (50pg/mL). This was done to 30 evaluate for any effect of PS fragments on JEV stability in presence of metals. The preparation of Protamine Sulphate (PS) fragments is described in Example 1.

Samples were prepared on the same day (see Table 9) and analyzed three weeks later. All samples were analyzed by SEC-HPLC, but only samples at pH 8 (#21-40) were analyzed by ELISA.

44

3.4.2 Experiment 20110913(DP): DP Formulation at different metal ion concentration of Ni(II), Cu(II), Cr(IlI)

5 FBV 11D87 (formulated with Alum Lot 4539) was used in this study. FBV was adjusted to pH 7 and pH 8 and spiked with Ni(II)/Cu(II)/Cr(III) at 100, 500 and lOOOng/mL to evaluate any metal ion concentration /pH depended effect. Table 10 shows the experimental design of this experiment. All formulations were stored in Falcon tubes at 2-8°C and 22°C. Samples stored at 22°C were analyzed by SEC-HPLC 10 and ELISA after 5 weeks.

3.4.3 Experiment 20110812-Metal Spiked DP

Final Bulk Vaccine 11D87 (formulated with Alum Lot 4539) was obtained from a 15 commercial production run and used in this study. Residual formalin in DS was analyzed as 28. lppm, residual sulphites was 92.2ppm. Actual content in DP can be considered to be in the same range. FBV JEV11D87 was adjusted to pH 7.0/7.4/7.8 and spiked with 500ng/mL (final concentration) of Fe(ll), Fe(III), Ni(II), Co(II), Cu(II), Zn(II). A metal ion mix formulation was also prepared containing all of the individual 20 metal ions together in solution. Formulations with Cr(III) were prepared later on and Cr(III) was not included in metal ion mix. Control formulations were only adjusted to the desired pH, but not spiked with metals. All formulations (#1-24) were prepared on the same day and stored in Falcon tubes at 2-8°C and 22°C.

Additional Cr(III) spiked samples (#25-27) were prepared by taking aliquots of the 25 control samples stored at 2-8°C and spiked with Cr(III) to a final concentration of 500ng/mL. Formu lations were stored at accelerated conditions (22°C) only. Table 11 shows the experimental set-up of this experiment. All samples stored at 22°C were analyzed by ELISA (monoclonal and polyclonal) after 4 weeks and 7 weeks.

30 3.4.4 Experiment 20110819: DP Formulation using various Alum lots

Spiking studies as described above can give first evidence of possible instability of the formulated antigen in presence of certain metals, but might not be completely representative of the real conditions where metals present in Aluminium hydroxide

45

are incorporated in the three-dimensional structure of the gel resulting in different local concentration and orientation/accessibility. To overcome these limitations an initial study was started to simulate the real conditions. A single NIV batch (11A74) obtained from a commercial production run was formulated with various Alum lots 5 produced by Brenntag covering a broad range of residual metals. 4.75mL of NIV was mixed with 0.25mL Alum (2%) in Falcon tubes. The final Aluminium hydroxide concentration was 500pg/mL (=0.1% A1203). Formulated vaccine samples were stored at 2-8°C and under accelerated conditions at 22°C. All of these Alum lots contained residual metal ions at different concentrations. Alum lot 4230 has the highest level for 10 Fe, Cu, Ni and V (see Example 3). Note that metal ion valences cannot be specified by ICP-MS. A mixed Alum sample containing equal amounts of 4230 and 4074 was also prepared to get an "intermediate" level for Ni(II) and Cu(II). Samples were analyzed after 6 weeks of storage at 22°C. The residual amount of formaldehyde and sulphite estimated by recalculation from available DS analysis results corrected by dilution 15 factor of NIV to DS was 76ppm formaldehyde and 192ppm sulphite respectively.

3.1 Antigen Desorption from Aluminium Hydroxide for SEC-MALLS analysis

Viral particles were desorbed from Aluminium hydroxide. -625 pL of DP was spun 20 down (8°C, 5 min, 3300 x g) and the supernatant was either discarded if not otherwise stated or analyzed by JEV-SEC-MALLS to detect the unbound antigen concentration. Viral particles were desorbed by suspending the Aluminium hydroxide particles with 62.5 pL 0.8 M potassium phosphate buffer (pH 8) containing BSA (50 pg/mL). BSA was added to the desorption buffer for SEC-MALLS analysis to minimize losses 25 caused by unspecific adsorption of the antigen. After shaking (500 rpm) the

Aluminium hydroxide particles for 10 min at room temperature, particles were removed by centrifugation and the supernatant was collected into a LoBind Eppendorf tube and the desorption procedure repeated on remaining sample. The pooled desorbed antigen (~5x concentrated sample; final volume 125pL; starting 30 volume -625 pL) was then further analyzed by SEC-MALLS.

3.2 SEC-MALLS HPLC Method

Desorbed antigen was analyzed by SEC-MALLS. In brief, following desorption of the antigen from Aluminium hydroxide 100 pL of the pooled desorbed material (~5x

46

concentrated) were subsequently loaded onto a Superose 6 10/300 GL SEC column, lx PBS + 250 mM NaCl was used as mobile phase. Ultraviolet (UV) 214 nm and MALLS signals of viral particles were recorded and analysed using Chromeleon and ASTRA software packages.

5

3.3 Inactivated JEV ELISA (polyclonal based)

Desorption of the antigen from Alum and ELISA analysis was carried out using polyclonal sheep anti JEV antibodies for coating the 96 well ELISA plates as described in Example 4.

10

3.4 Inactivated JEV ELISA (monoclonal based)

During course of this investigational testing, a monoclonal (mAb) based JEV ELISA was developed. The assay is primarily based on the "polyclonal JEV ELISA" assay format, only a monoclonal anti-JEV antibody (clone 52-2-5) is used for coating and the 15 current polyclonal antibody for detection. The employed mab 52-2-5 was shown to be specific for JEV and to recognize a neutralizing epitope. Mab clone 52-2-5 was obtained by subcutaneously immunizing BALB/c mice with commercially available vaccine lot JEV08J14B. Spleen cells of the mice were fused to myeloma cells. From resulting hybridoma cells single clones were selected and sub-cloned. The clones were 20 negatively screened against Bovine Serum Albumin, Protamine sulphate and an extract of the production cell line of the JE-vaccine (Vero cells). A positive screen was done against Neutralized Inactivated Virus (NIV) of vaccine lot JEV08M20. For screening, microtiter plates were coated with the relevant antigen and reacted with supernatant of cultures of the selected clones. For detection a goat anti mouse 25 polyclonal antibody conjugated with alkaline phosphatase was used. Mab clone 52-2-5 was shown to recognize a neutralizing epitope on domain III of the envelope (E) protein of JEV containing Ser331 and Asp332 (Lin C.-W. and Wu W.-C. J Virol. 2003 ;77(4):2600-6). Binding of the mab to the indicated neutralizing epitope is for instance determined as described in Lin and Wu (2003) by site-directed mutagenesis of the 30 domain III at position 331 (for instance: S—>R), and/or by alanine mutations at or near position 331 of domain III, for instance of residues Ser 331 and Asp332, followed by immunoblots to determine binding of the mab to the mutated proteins. Negative binding results indicate that the epitope of the mab is the neutralizing epitope identified by Lin and Wu (2003). The neutralizing characteristic of the epitope gives

47

rise to the assumption that the epitope might be of importance for the antigen to elicit a protective immune response.

JEV samples were analyzed by both ELISA assays, polyclonal and monoclonal. The relative specific epitope content can be expressed as the ratio of the total antigen 5 content determined by "monoclonal ELISA" (clone 52-2-5) divided by total antigen content determined by "polyclonal ELISA". Any differences in the ratio may indicate differences in specific epitope content 52-2-5. Results close to 1 would correspond to high epitope contents, and results close to 0 correspond to low relative epitope content. A low ratio indicates presence of structural changes of the neutralizing 10 epitope.

In the course of development of this "mAb ELISA", differences between vaccines lots were detected, which could be correlated with potency results of these lots.

3.5 Statistical Evaluation

15 Statistical evaluation was done with Statgraphic Plus 3.0.

4 Results

4.1 Experiment 20110913(NIV): NIV Formulation at different metal ion concentration of Ni(II), Cu(II), Cr(III) with or w/o presence of PS fragments

20 SEC-HPLC results of NIV formulations (pH 7 and pH 8) containing metal ions [Ni(II), Cu(II), Cr(III)] w/ and w/o PS fragments are summarized in Table 12. SEC-HPLC results show that antigen recoveries of most of the samples was >80%. Some samples (#7, #36, #38) showed slightly reduced recoveries in the range of 70-80%. It has to be noted that the actual virus content is quite low and precision of HPLC results can be 25 estimated as approx.± 20%. Since for samples #36 and #38 the recoveries for following formulations (#37, #39) at next level of individual metal ion content were higher again, these differences might be caused by assay variability and were not considered as significant. Based on the results obtained it was not possible to clarify the influence of metal ions with respect to the recovery of soluble inactivated JEV. 30 However, SEC-HPLC only gives information about content of soluble virus, but no information about any potential surface modification. Only formulations prepared at pH 8 were also analyzed by ELISA (duplicate analysis). The ratio of monoclonal/polyclonal ELISA was calculated and can be used for comparison purpose of results. Analysis of samples by ELISA (see Table 13) do not show any significant 35 influence of tested metals on degradation of inactivated JEV at pH 8 after three

48

weeks at 22°C. There might be a trend of decreasing ratio in presence of Cu(II), but overall it appears that an incubation time of three weeks at 22°C seems not be sufficient to detect any significant degradation. As also shown in DOE experiment (Example 1) inactivated JEV appears to have higher stability at pH 8 when stored at 5 accelerated conditions at 22°C and this would also contribute that significant effects were not observed. In this experiment it was also shown that PS fragments do not have any influence on JEV stability. This is also well in agreement with DOE results. NIV samples 1-20 formulated at pH 7 showed significant reduction in monoclonal epitope content in presence of Cu(II). At the highest tested concentration (lOOOng/mL) 10 the ratio was close to zero indication significant structural changes of the antigen.

4.2 Experiment 20110913(DP): DP Formulation with different metal ion concentration of Ni(II), Cu(II), Cr(III)

Analysis of desorbed JEV antigen is summarized in Table 14 (SEC-HPLC) and Table 15 15 (ELISA). Antigen recoveries for all samples as determined by SEC-HPLC was

>80% after 5 weeks at 22°C indicating no significant influence of tested metal ions on desorption recovery. As shown in Figure 8, there is a trend of decreasing ratio as analyzed by ELISA in presence of Cu(II) and Cr(III) at pH 7. Formulations at pH 8 appear to be more stable.

20

4.3 Experiment 20110812(DP): Metal Ion Spiked DP

In this experiment FBV11D87 was used as starting material. The formulation pH value was adjusted in a more narrow range (pH 7.0, 7.4 , 7.8) and additional metal ions were used for spiking, each at 500ng/mL (final concentration). The metal ion mix 25 contained all individual metal ions with the exception of Cr(III) in single formulations (each metal at 500ng/mL). ELISA results obtained after 4 weeks and 7 weeks at 22°C are summarized in Table 16. Results are also displayed as graphs in Figure 9.

Statistical evaluation of stability samples stored at 22°C for 7 weeks was performed.

30 ANOVA (analysis of variance) showed significant effects of parameters (pH and metal type) on antigen stability, that is expressed as the ratio of monoclonal/polyclonal ELISA (see Table 17). The ANOVA table decomposes the variability of Ratio into contributions due to various factors. Since Type III sums of squares have been chosen, the contribution of each factor is measured having removed the effects of all other

49

factors. The P-values test the statistical significance of each of the factors. Since the P-values for pH and metal ion type are less than 0.05, these factors have a statistically significant effect on Ratio at the 95.0% confidence level.

5 In Table 18 a multiple comparison procedure is applied to determine the significance of the differences observed with respect to the means. Significant effect on ratio was shown for Cu(II) and the metal mix compared to the non-spiked control formulations. The bottom half of the output shows the estimated difference between each pair of means. An asterisk has been placed next to 7 pairs, indicating that these pairs show 10 statistically significant differences at the 95.0% confidence level. At the top of the page, 3 homogenous groups are identified using columns of X's. Within each column, the levels containing X's form a group of means within which there are no statistically significant differences. The method currently being used to discriminate among the means is Fisher's least significant difference (LSD) procedure. With this method, 15 there is a 5.0% risk of calling each pair of means significantly different when the actual difference equals 0.

Significant effect on ratio was shown for Cu(II) and the metal ion mix. The influence of other metal ions might become significant at longer storage period. The metal ion mix contained the highest total concentration and might represent a worst case. 20 However, it was concluded that several metal ions present in Alum might contribute to degradation, each to different extent. These results further support the proposed root cause of metal ion-catalysed antigen degradation. It has to be noted that spiking experiment might not fully simulate the real conditions of residual metal ion impurities present in Alum 4230. Metal ions are incorporated in the Alum structure 25 and the local concentration and orientation might be different from metal ions used in spiking experiments. It is also known that metal ions (e.g. Fe) have low solubility in presence of phosphate ion (P043-). Therefore the actual concentration of soluble metal ions and contribution of metals present as metal-phosphate complex on JEV degradation is unknown.

30

4.4 Experiment 20110819: Preparation of DP samples with different Alum lots

Spiking studies as described earlier can give first evidence of possible instability of the formulated antigen in the presence of some metal ions, but might not be

50

completely representative of the real conditions where these metal ions present in Aluminium hydroxide are expected to be incorporated in the three-dimensional structure of the Alumini um-hydroxid e gel resulting in different local concentration and orientation/accessibility. To overcome these limitations an initial study was 5 started to simulate the real conditions. A single NIV (11A74) was formulated with various Alum lots obtained by Brenntag covering a broad range of residual metals. Formulated vaccine samples were stored at 2-8°C and under accelerated conditions at 22°C. All of these Alum lots contained residual metal, ions at different levels. Lot 4230 had the highest level for Fe, Cu, Ni and V (see Table 19). Note that the actual content 10 of metal ions present in the formulated product is only 1/20 of the concentration in Alum (2%) stock solution. Note that metal ion valences cannot be specified by ICP-MS. Analysis of desorbed antigen by ELISA of samples stored at 22°C for 6 weeks is shown in Table 20.

Pooled standard deviation was calculated from all samples (spooled ~ 0.075) as a 15 measurement of experimental uncertainty. Mean values for ratio and 95% confidence intervals (calculated based on spooled) were plotted against the individual formulations (see Figure 10). Samples formulated with Alum 4230 showed a trend to lower ratio compared to the other samples. However, differences were not as large to show statistical significant differences between the various formulations.

20

5 Summary

It was shown that metal ions contribute to the degradation of the inactivated JEV under accelerated storage conditions (22°). In spiking studies higher concentration of residual metal ions (range 100 -1000 ng/mL) were used than present in FVL 25 formulated with Alum lot 4230 (e.g. Fe~310 ng/mL; Cr~64ng/mL; Ni~52ng/mL). Cu content in FVL can be only estimated as 3ng/mL based on ICP-MS data of 2% Alum stock solution since LOD is 25 ng/mL. Higher metal concentration and storage temperature were chosen to increase the rate of any potential degradation reaction. In fact, for FVL JEV09L37, the potency loss occurred after 11 month stored at 2-8°C. It 30 was also shown that metals can form insoluble complexes with phosphate ions making estimation of actual levels of metals present difficult.

In spiking experiments ELISA results showed statistically significant structural changes of virus surface in as little as 4 weeks at 22°C in presence of metal ions. It was shown that the ELISA ratios for formulations containing Cu(II) and metal ion

51

mix (containing Fe(II), Fe(III), Co(II), Cu(II) and Zn(II)) were statistically significant lower compared to the non-spiked control formulation. Cu(II) was also found in Alum lot 4230 (2% stock solution) at 64ng/mL, corresponding to ~3ng/mL in FVL. In all other Alum (2%) lots Cu(II) content was < 25ng/mL (below limit of detection). 5 For formulation experiments of the antigen using different Alum lots, longer storage time (>6 weeks at 22°C) at accelerated conditions is required. There is a trend that formulations prepared with Alum 4230 showed lower ELISA ratios compared to other lots. Slower degradation rate compared to spiked formulation might be contributing to lower metal ion content in commercial Alum lots. It was also observed that the 10 antigen shows higher stability at pH 7.5-8 compared to pH 7 and that PS fragments do not contribute to any degradation reaction. These results are in good agreement with DOE results described in Example 1.

Example 3

15

As part of the out-of-specification investigation concerning FVL JEV09L37, the used Aluminum hydroxide lot (lot 4230) was determined to be the most probable root cause for the observed loss of potency. Aluminum hydroxide (referred to as Alum during the manufacturing process of JE-PIV) is purchased from Brenntag Biosector as 20 autoclaved suspension termed "Alhydrogel® Aluminium Hydroxide Gel Adjuvant", Each batch is sterilized by radiation prior to use in the JEV production process. A number of different Alhydrogel® batches were analyzed for appearance, metal ion content and physical properties.

25 1 Introduction

1.1 Aluminum hydroxide

Brenntag Biosector's Alhydrogel® has a specified Aluminum content of 10 mg/mL which translates to 2% A1203 and 3% Al(OH)3, respectively. Further specifications are Nitrogen (max 0.005%), free Sulphate (max 0.05%), total Sulphate (max 0.1%) and 30 pH (6.5 ± 0.5). it has a shelf life of 26 months when stored at room temperature.

1.2 Generation of Aluminum hydroxide

Alhydrogel® 2% (referred to as Aluminum hydroxide) is manufactured by Brenntag (CAS no. 21645-51-2).

35

52

It is noteworthy, that after the initial precipitation by Ammonia and Ammonium sulphate, all remaining downstream processing steps can only remove water soluble impurities, and will consequently not remove any metal ions either adsorbed to or precipitated with the Aluminum hydroxide particles. No specific purification step to 5 removal of contaminating metal ions is included in the manufacturing process. The Ammonium sulphate solution is recycled following the precipitation and concentration of the Aluminum hydroxide gel.

1.3 Use of Aluminum hydroxide lots in JEV manufacturing 10 For the production of commercial JEV vaccine batches a total of 5 different Brenntag Alhydrogel® 2% lots have been used so far. These lots differed in the origin of the raw material Ammonia alum source and in the number of JEV batches they were used to formulate. An overview is given in Table 21.

15 2 Definitions & Abbreviations

Alhydrogel 2% Aluminum hydroxide solution (also referred to as alum) DS/DPDrug Substance / Drug Product ESG Environmental Scientifics Group F-AAS Flame Atomic Absorption Spectrometry 20 FVL Final Vaccine Lot

GF-AAS Graphite Furnace Atomic Absorption Spectrometry ICP-MS Inductively-Coupled-Plasma Mass Spectrometry JEV Japan Encephalitis Virus

JE-PIV Japan Encephalitis Purified Inactivated Virus 25 LOQ Limit Of Quantification

P&TD Patch & Technical Development PSD Particle Size Distribution PZC Point of Zero Charge QCI Quality Control Immunology

30

3 Materials and Methods

3.1 Alhydrogel® batches and raw materials

Alhydrogel® 2% lots: 3877, 4074, 4187, 4230, 4414, 4470, 4539, 4563, 4587, 4621 (not 35 all Alum lots listed were used in the formulation of JE-PIV; see Table 21)

53

Alhydrogel® 2% 7x washed lots: 4577, 4580, 4596 (sourced from Brenntag, not typical of the 2% Alum received for formulation)

Ammonia Alum: AA0427, 6-80578-28, 10094, 10095, 91480 (represents ammonia alum used to produce different Alum lots. AA0427 was used to produce Alum lot #4230 used 5 to formulate JEV09L37)

Ammonia 24% technical grade solution: 4140716; sample only and not related to any of the alum batches tested.

3.2 Alhydrogel® PSD Measurements 10 Aluminum hydroxide particle size distribution (PSD) was analyzed on a Malvern

Mastersizer 2000pP system with a 20mL sample cell. Alhydrogel® 2% bulk substance was diluted 1:20 in water and lmL was added to the sample cell. Final dilution of sample in sample cell was therefore 400 fold (0.005% Aluminum hydroxide).

15 3.3 Alhydrogel® Zeta-Potential Measurements

Zetapotential and point of zero charge (PZC) was measured on a Malvern Zetasizer ZS system equipped with a MPT-2 autotitrator. Alhydrogel® 2% bulk substance was diluted 1:20 in PBS and equilibrated over night at room temperature. For recording of the charge titration curve the pH was adjusted using lOOmM HC1 and lOOmM NaOH 20 solutions. PZC was determined by extrapolation of the zero charge in the titration plot (intersection of titration curve and x-axis). Point of zero charge corresponds to the pH value where the surface of the sample has no net charge.

3.4 Analysis of metal ions in Aluminum hydroxide and raw materials

25 Selected metal ions were analyzed either by inductively-coupled-plasma mass spectrometry (ICP-MS), flame atomic absorption spectrometry (F-AAS) and graphite furnace atomic absorption spectrometry (GF-AAS) at the Medical Laboratory Bremen (Germany). In short, samples containing Aluminum hydroxide were treated with conc. HN03 under heat until a clear solution is obtained. The clear solution is then further 30 diluted and analyzed. The presence and content of following metal ions were determined: Pb, Cd, Cr, Co, Fe, Cu, Ni, Ag, W and Al. Depending on the sample dilution, the limit of quantification (LOQ) was 5 to 25 ng/mL. Crystalline Ammonia Alum samples were dissolved in HN03 before analysis. Ammonia had to be adjusted to acid pH before analysis.

54

In addition a semi-quantitative 70-Element scan was performed by ESG (UK) using a combination of ICP-MS (Agilent 7500ce) and ICP-AES (Perkin Elmer Optima 4300DV), which were calibrated using certified standards. The element scan is a 5 screening method and not as sensitive as trace metal analysis for selected metals as performed by Medical Laboratory Bremen. However, such a screening gives a good overview about the presence and levels of certain metals.

4 Results

10 4.1 Determination of Alhydrogel® particle size distribution

Table 22 summarizes PSD data of two sublots each of Alhydrogel© lots 4230 and 4740. Distribution results are shown in Figure 11. Mean particle was ~2-4jim with populations of smaller (<lpm) and larger (>20pm) particles being present in all four samples. The four tested Alhydrogel® samples showed no significant difference in 15 mean particle size distribution.

4.2 Zeta-Potential Measurements

Two sublots each of Alhydrogel® lots 4230 and 4740 (2% stock solution diluted 20 fold in PBS and equilibrated overnight at RT before analysis) were analyzed for point of 20 zero charge. Table 23 summarizes results of PZC for the four samples showing very similar PZC in PBS buffer. Titration curves are shown in Figure 12. No difference in the titration curves and PZC could be observed between the four analyzed samples.

4.3 Determination of residual metal ion content in Alhydrogel® batches 25 The current limits for Fe in 2% Aluminum hydroxide solutions according to the Ph. Eur. are 15 ppm (= 15 pg/mL) and a total maximum of 20 ppm (= 20 ]ig/mL) for other heavy metals (such as Pb). However, a concentration of 15 ppm Fe would correspond to 0.27 mM Fe in solution. Taking into consideration that even trace amounts of residual metal ions can catalyze a variety of degradation reactions for proteins (e.g. 30 oxidation, activation of proteases etc.) and that metals remain stable in solution, differences in metal ion content between Aluminum hydroxide lots might cause differences in antigen stability over time.

The concentrations of a number of metal ions in commercially available aluminum hydroxide lots were analyzed using ICP-MS. The results of these analyses are 35 summarized in Table 24. Lots 4074, 4230, 4470, 4414 and 4539 were used in the

55

10

production of commercial JEV batches. As a 2% Alhydrogel® stock solution equals an Al concentration of 10 mg/mL the quantification of Al content in the different samples can be used as reference for the results obtained for the other metal ions. Indeed an average Aluminum content of 10.3 mg/mL could be measured showing the accuracy and reproducibility of the method.

When comparing the different Alhydrogel® lots large variations in the amount of contaminating metal ions could be observed. Most notable contaminating metal ions are Fe, Cr and Ni which were detected in all batches. In addition lot 4230 contained detectable amounts of Cu which was below LOQ in all other batches.

However, it has to be noted that none of these metals were detected in quantities near the specifications of Alhydrogel® mentioned above. For example the highest concentration of iron found in lot 4230 was 5.6 pg/mL or roughly 40% of the permitted concentration.

15 As the precipitation of Aluminum hydroxide during the process results in significant amounts of residual Ammonia in the final product, Brenntag Biosector offers a second grade of Alhydrogel®. In contrast to the standard product this "improved" Alhydrogel® is washed 7 times with water during the purification step instead of only 4 times for standard Alhydrogel®. To test if this additional washing steps would 20 result in reduced metal ion contamination three different lots (4580, 4596 and 4577) were analyzed. Results are included in Table 24. No difference in metal ions comparing to the standard grade Alhydrogel® could be observed suggesting that the metal ions are either strongly bound to the surface of the Aluminum hydroxide particles or actually co-precipitate during the production process.

25

Figure 13 shows a comparison of the different Alhydrogel® lots analyzed. The total contaminating metal ion content for all tested contaminating elements are shown with the absolute shares for the three major metals Fe, Cr and Ni depicted in different colors. As can be seen lot 4074 has very little contaminating metal ions 30 compared to the majority of the other analyzed batches. Only lot 3877 showed a similar lot contamination whereas lot 4230 shows by far the highest contamination of all batches analyzed during this investigation.

56

During the investigational testing a large variation in the metal ion content was observed between different batches of Alhydrogel® (see Figure 13). To test if these contaminating metal ions are located in the Aluminum hydroxide fraction or in the supernatant Lot 4230 was separated into a supernatant and a sediment fraction (see 5 Table 25). As can be seen less than 2% of the metal ions could be detected in the supernatant indicating that all contaminating metal ions are either bound to the Aluminum hydroxide particle surface or inside the particle structures. The local metal ion concentration can therefore be estimated to be at least 50-100 times higher since the solid volume fraction (volume of Alum-pellet after centrifugation) of 0.1% A1203 10 (corresponding 0.5 mg/mL Al) used in JEV vaccine formulation is approx. 10-20pL per lOOOpL of FVL.

4.4 Determination of residual metal ion content in Alhydrogel® raw materials

15 A total of 5 batches of Ammonia Alum (raw material for Alhydrogel®) were received from Brenntag Biosector. These batches were used for the production of 4 Alhydrogel® batches (see Table 26; note that 2 different batches of ammonia alum were used to produce Alum lot # 4563), three of them used for JEV vaccine lots. Upon delivery a marked difference in the visual appearance between the batches was 20 observed. Whereas four samples showed a crystalline structure similar to large salt crystals one lot was more sand grain like. Furthermore whereas the crystals were completely colorless the Ammonia Alum AA0427 material was zinc-white. This finding is of importance as lot AA0427 was used for the production of Alhydrogel® batch 4230 which was subsequently used in the production of JEV09L37 (see Table 25 26). However, according to Ph. Eur. the appearance of Alum can be "granular powder or colourless, transparent, crystalline masses" (European Pharmacopoeia 7th edition. Monograph 1357p). Consequently, all five raw materials comply with this definition.

ICP-MS analysis of these five Ammonia Alum lots and one lot of 24% Ammonia 30 solution (technical grade) showed that the lot 6-80578-28 has very low amounts of contaminating metal ions whereas lot AA0427 has very high amounts (see Table 27). This correlates with the analysis of the corresponding Alhydrogel® batches (see Table 24) showing that the origin of these metal ions is in the raw material and not process derived.

57

Figure 14 shows a comparison of the different Ammonia Alum lots analyzed. The total metal ion content for all tested contaminating elements are shown with the absolute shares for the three major metals Fe, Cr and Ni depicted in different colors. Ammonia 5 Alum lot 6-80578-28 has ~ 40 times less contaminating metal ions compared to lot AA0427 at identical Al concentration resulting in the observed much higher purity of the corresponding final Alhydrogel® batches 4074 and 4230.

When comparing the results for the three Ammonia Alum lots with their 10 corresponding Alhydrogel® lots a similar ratio for the three major contaminations Fe, Ni and Cr can be observed (Table 28).

This shows that there is basically no change in the metal ion composition during the whole Aluminum hydroxide production process. Consequently, a secondary 15 contamination during the production process (e.g. other chemicals or the storage container) can be excluded.

Table 29 summarizes results of semi-quantitative element scan analysis performed by ESG.

20 5 Summary

Alhydrogel® is used in a 0.1% final concentration as adjuvant in the current JEV vaccine formulation. During the investigation of an out-of-specification (OOS) potency result for production FVL JEV09L37 an evaluation of the Alhydrogel® production process was initiated. A total of 13 different Alhydrogel® lots were analyzed for the 25 presence of contaminating metal ions that could reduce protein stability.

Large variations in the concentration of a number of metal ions were observed for different Alhydrogel® lots. When analyzing the raw materials it was shown that these contaminations were present at the same concentration as found within the Alhydrogel.

30 Higher levels of Fe, Ni and Cu ions were noted in Alhydrogel® lot 4230 when compared to the other investigated lots. Lot 4230 was the only one where residual Cu ions were detected. This lot 4230 was used for the formulation of FVL JEV09L37. When analyzing supernatant and insoluble fraction of an Alhydrogel® batch these contaminating metal ions could only be found in the precipitate indicating that these

58

ions are either bound to the Aluminum hydroxide particle surface or actually part of the particle.

The raw material used for batch 4230 showed an optical difference to the other batches used in JEV vaccine production. Although macroscopic and in composition 5 different from other Alhydrogel® lots used for JEV production, lot 4230 fulfilled all requirements detailed by the Ph. Eur. Also physical characterization (particle size distribution and point of zero charge) showed no differences between lot 4230 and other Alhydrogel® lots not showing these high metal ion contaminations.

10 Example 4

1.1. Materials, Equipment and Methods

1.2. Equipment

Analytical balance (readability of O.lmg; e.g. Mettler Toledo XP205DR/M) 15 Precision balance (readability of O.lg; e.g. Mettler Toledo, Model N° XS6002S Delta Range)

Filter Units 0.22pm (e.g. Stericup Cat N° SCGPV01RE) or 0.2pm filter system (e.g. 50mL Millipore Steriflip)

Freezer (- 20°C) and Ultra-Freezer (-80°C)

20 Fridge (+ 2 to 8°C)

Magnetic stirrer (e.g.KIKA Labortechnik RCT basic) and magnetic stir bars Microplate Washer: e.g. BioTek ELx405

Microplate Reader: e.g. BioTek Synergy 2 and Gen5 Secure software Microplate Incubator (37°C)

25 Microtiter Sealing tape (e.g. Thermo Electron 9503130)

Multichannel pipettes and tips (e.g. Eppendorf Research Pro 50-1200pL, Eppendorf Research, lO-lOOpL)

pH meter (e.g. WTW Series ino Lab, Terminal 740 and pH/Cond. 740)

Pipettes and tips (e.g. Eppendorf Research, 0.5-10pL, 2-20pL, 20-200pL, lOO-lOOOpL, 30 500-5000pL)

Pipettor (e.g. IBS Biosciences Pipetboy)

PP Tubes 15mL (e.g. Sarstedt 62.515.006) or PP tubes 50mL (e.g. Greiner 227261) Reagent Reservoir 50mL (e.g. Corning Incorporated 4870)

Serological pipettes (e.g. Falcon, 2mL, 5mL, lOmL, 25mL, 50mL)

59

Titertube Micro Tubes - Bulk (BioRad 223-9391)

Vortex mixer (e.g. VWR Analog Vortex Mixer, Model N°945304 )

1.5mL or 2.0mL Eppendorf LoBind tubes (Cat N° 0030 108.116, CatN° 0030 108.132,

respectively)

5 96 well Microplate (F96 Cert. Maxisorp Nunc-Immunoplates)

For analysis of DP samples in addition:

Bench top centrifuge (e.g. Beckman coulter, Microfuge 16 Centrifuge, Cat N°A46473) Orbital shaker (e.g. Eppendorfer Thermomixer compact)

10 50mL PP tubes (e.g.Greiner 227261)

1.3. Reagents

PBS lOx (e.g. Gibco, Cat.N° 14200-083)

Tween 20 (e.g. Sigma Cat. N° P7949)

15 2M Sulphuric Acid (Volumetric solution, e.g. Fisher, Cat, No. J/8410/17)

De-ionised water, e.g. (Milli-Q, 18.2Q)

Sodium carbonate - bicarbonate capsules (e.g. Sigma, Cat. No. C3041)

Hydrochloric acid (HC1) lmol/L (e.g. Merck, Cat.N0 1.09057.1000)

Sodium hydroxide (NaOH) lmol/L (e.g. Merck, Cat. N°l.09132.1000)

20 Glycerol (e.g. Sigma)

For analysis of DP samples in addition:

Di-potassium hydrogen phosphate trihydrate (e.g. Sigma, Cat No. P5504)

Potassium di-hydrogen phosphate (e.g. VWR, AnalaR Normapur, Cat No. 26936.260) 25 Albumin, Bovine Serum (BSA), ELISA grade (e.g. Sigma, Cat. No. A3059) TMB Substrate (e.g. BioFX, TMBW-1000-01)

Donkey anti rabbit IgG HRP Conjugate (Jackson Immuno Research, Cat N° 711-035-152)

Reconstitution:

30 The content of 1 vial (0.4mg) is reconstituted in 0.5mL of de-ionised water and thoroughly mixed until total dissolution. Add 0.5mL of Glycerol and mix it further until homogeneity. Aliquots are stored at -20°C until use.

Inactivated JEV Reference Standard (Intercell Biomedical Ltd.)

Purified sheep anti-JEV (Intercell Biomedical Ltd.)

60

Purified rabbit anti-JEV (Intercell Biomedical Ltd.)

1.4. Solutions a) 0.05M carbonate buffer at pH 9.6 (used for coating of ELISA plates)

5 For lOOmL buffer, dissolve one bi-carbonate /carbonate buffer capsule in lOOmL de-ionised water. Check the pH and adjust to 9.6 ± 0.1 with HC1 or NaOH if required. Use on the day of preparation only. Keep ELISA coating buffer at RT during the day of use, then discard.

b) ELISA wash buffer and part of block / sample diluent (PBS-T)

10 Prepare approximately 1 litre for every plate used. Dilute lOx PBS stock 1+9 in de-ionised water, mix well and check pH (7.4 +/- 0.1), adjust with 1M HC1 or 1M NaOH as required. Add 0.05% (v/v) Tween20, mix well.

e.g. ELISA wash buffer (PBS-T) [1L]:

lOOmL lOx PBS 15 900mL de-ionised water

Mix well, check/adjust pll (7.4 +/- 0.1).

0.5mL Tween20 Mix well.

Use on the day of preparation only; keep ELISA wash buffer at RT during the day of 20 use, then discard.

c) Blocking solution: 5% BSA in PBS-T

Prepare approximately 25mL for every plate. Measure required quantity of PBS-T into a clean glass bottle using a serological pipette. Add a clean magnetic stir bar. Weigh the required amount of BSA, add to the surface of the PBS-T and mix gently on 25 a magnetic stirrer until all the BSA has gone into solution. Filter solution using a 0.2pm filter (either Steriflip filter system or syringe filter).

e.g. Blocking solution [lOOmL]

5g BSA

lOOmL PBS-T

30 Use on the day of preparation only; keep blocking solution at RT during the day of use then discard.

d) Sample diluent: 1% BSA in PBS-T

Prepare as above but using Ig of BSA per lOOmL PBS-T, approximately 25mL is required per plate.

61

e.g. Sample diluent [lOOmL]

lg BSA

lOOmL PBS-T

Use on the day of preparation only; keep sample diluent at RT during the day of use, 5 then discard.

For analysis of DP samples in addition:

e) lx PBS

Prepare 1 part lOx PBS with 9 parts de-ionised water 10 e.g. lxPBS [lOOmL]

lOmL lOx PBS 90mL de-ionised water

Use on the day of preparation only; keep lx PBS at RT during the day of use, then discard.

15 f) 20k ELISA buffer

Weigh an appropriate amount of BSA into a suitable container to make a 20x solution. Add the appropriate volume of lx PBS. Add Tween20 to a final concentration of 0.05% Mix on the magnetic stirrer until the BSA is fully dissolved. Filter the solution through a 0.2p.m filter (using either Steriflip filter system or syringe filter) 20 into a sterile container (and aliquote as needed).

e.g. 2Ox ELISA Buffer (25mL|

5g BSA

25mL lxPBS 12.5pLTween20 25 The solution can be stored at +2-8°C for 1 week.

g) 2x ELISA buffer

It is prepared by dilution of the 2Ox ELISA Buffer with lx PBS (1 part 20x ELISA Buffer and 9 part lx PBS).

e.g. 2x ELISA Buffer [20mL]

30 2mL 20x ELISA Buffer 18mL lx PBS

Use on the day of preparation only; keep 2x ELISA buffer at RT during the day of use, then discard.

h) Desorption buffer

62

o Potassium phosphate stock solution: Make a 3x stock solution of potassium phosphate (2.4M) by dissolving the appropriate volume of di-potassium phosphate trihydrate and of potassium dihydrogen phosphate in de-ionised water. Place on a magnetic stirrer and once dissolved make up the required volume, check that the pH 5 of the solution is 8.0+/- 0.1. Filter through a 0.2pm filter.

e.g. 3x stock solution of Potassium phosphate (2.4M) [50mL]

23.963g Di-potassium phosphate trihydrate

2.041g Potassium dihydrogen phosphate

Make up to 50mL De-ionised water 10 Store at +2°-8°C for up to 1 month.

o Make working strength desorption buffer (0.8M potassium phosphate buffer containing 1% BSA and 0.05% Tween20) by adding the appropriate volume of potassium phosphate stock (2.4M), of Tween20 and of BSA to the required volume of de-ionised water. Mix thoroughly and use on the day of preparation. 15 e.g. working strength desorption buffer [15mL]

5mL Potassium Phosphate (2.4M)

7.5pL Tween20

0.15g BSA

lOmL De-ionised water

20 Keep working strength desorption buffer at RT during the day of use.

1.5. Test samples and antibodies

Test samples:

o Drug Substance and / or NIV (various batches)

25 o JEV Vaccine samples (final bulk vaccine and final vaccine lot)

Inactivated JEV Reference Standard (Neutralised Inactivated Virus - NIV) (Intercell Biomedical Ltd.)

Polyclonal Antibodies:

o Coating antibody: Purified Sheep anti-JEV (Intercell Biomedical Ltd.) 30 o Primary detection antibody: Purified Rabbit anti-JEV (Intercell Biomedical Ltd.)

Secondary conjugated antibody: Donkey anti-Rabbit HRP Conjugate (Jackson Immuno Research Cat. N° 711-035-152)

63

2 Procedure

2.1. Plate coating o Label the plate with plate number, date and analyst.

o Prepare fresh 0.05M carbonate buffer (pH 9.6) on the day of plate coating. 5 Allow approximately 12mL for each plate coated.

o Remove the required number of aliquots of the coating antibody from the freezer and allow thawing at RT. Prepare a dilution of Purified Sheep anti-JEV in carbonate buffer. Mix well by inversion of the tube.

o Using the multichannel pipette, apply lOOpL/well to a 96-well Maxisorp plate 10 within 15min of preparation of the antibody dilution.

o Cover with microtiter sealing tape and incubate 17 to 72hrs at + 2-8°C.

2.2. Washing o Remove plate from the refrigerator and allow warming to room temperature, o Wash the plate/s with the Microtiter plate washer 3 times using the respective 15 wash program (300pL per well, three times, final dispense). After that: remove any remaining wash buffer by decanting. Invert the plate and blot it against a clean paper towel. Do not allow microtiter plate to dry between wash steps and reagent addition.

2.3. Blocking o Prepare a Blocking Solution 5% (w/v) of BSA in PBS-T as above. 20 o Apply 200pL Blocking Solution per well, cover the plate(s) with a cover plate and incubate at 37°C for 1 hour +/- 10 min.

2.4. Preparation of Standard Curve Dilutions o Remove the NIV reference standard from the freezer, allow thawing at RT, mix well. Prepare a lAU/mL stock dilution of the current reference standard; use at least 25 20pL of NI V reference standard for dilution.

e.g. NIV reference standard Pre-dilution:

Concentration: 235AU/mL (lot N° 03/2009)

To prepare a lAU/mL working standard solution dilute it 1 to 235 in sample diluent: 4680 pL sample diluent 30 20 pL NIV reference standard o Prepare then the following working standard solutions from the lAU/mL pre-dilution:

0.8 AU/mL, 0.6 AU/mL, 0.4 AU/mL, 0.2 AU/mL, 0.1 AU/mL and 0.05 AU/mL in sample diluent.

64

2.5. Quality Control Samples a) Quality control (QC) samples (for example at 0.75, 0.30 and 0.18 AU/mL)

should be made from the NIV reference standard pre-dilution freshly at the time of the assay then discarded once used.

5 b) These controls are part of the system suitability criteria and allow the performance of the assay to be monitored over time.

2.6. Preparation of Test Samples

Drug Substance Preparation 10 Drug substance test samples are received for testing at unknown concentrations. These will be tested at six dilutions in triplicate. The dilutions will be made independently into the range of the standard curve, e.g. pre-dilution of 1 in 15 or other suitable dilution then six dilutions with sample buffer.

15 NIV Sample Preparation

NIV samples will be received for testing at unknown concentrations and pre-diluted in the range of the standard curve (e.g. 1 in 30 or other suitable dilution) then diluted six times in the same manner as the DS samples.

20 Drug Product Supernatant Preparation a) For the analysis of Bulk-DP samples mix well sample by vortexing. Transfer exactly lmL into a 1.5mL LoBind Eppendorf tube.

For the analysis of final product container samples transfer the content of 2 syringes (0.6mL per syringe) of the same lot into a 1.5mL LoBind Eppendorf Tube. Mix content 25 of tube thoroughly by inversion to ensure homogeneity of the DP and transfer exactly lmL into fresh 1.5mL LoBind Eppendorf tube.

b) Centrifuge tubes containing lmL DP each at 3300xg for 5 minutes.

c) For each sample, pipette 25pL of 20x ELISA buffer into fresh LoBind Eppendorf tube.

30 d) Carefully remove 475pL of the supernatant without disturbing the alum pellet and transfer into the tube containing the 20x ELISA buffer. Mix gently by inversion. Re-spin 2 min at 16,000xg. Store sample at +2-8°C prior to analysis.

NOTE: DP supernatant samples prepared in this way should be measured neat in triplicate in the inactivated JEV ELISA.

65

e) Carefully remove as much of the residual supernatant from the centrifuged tube using a 10-200pL pipette without disturbing the alum pellet and discard the supernatant.

f) The pellet obtained is subjected to the Desorption procedure as described 5 below.

Drug Product Desorption Procedure a) Add 158ptL of working strength desorption buffer to each pellet left in the LoBind tube.

10 b) Resuspend the pellet by pipetting up and down several times to ensure complete re-suspension of the pellet and homogenisation of the sample.

c) Incubate samples for 10 min at RT on an orbital shaker at 500rpm.

d) After incubation centrifuge samples at 3300xg for 5minutes.

e) For each sample pipette 250pL of 2x ELISA buffer into a fresh LoBind 15 Eppendorf tube.

f) Carefully remove 83.3pL from the upper part of the supernatant containing the desorbed product without disturbing the pellet and transfer into the tube containing the 2x ELISA buffer. Remove remaining supernatant using a 20-200pL pipette without disturbing the pellet and discard the supernatant.

20 g) Add another 158]iL of working strength desorption buffer to each pellet.

h) Carry out 2 more desorption cycles (3 in total) pooling the 3x 83.3pL of the desorbed material + 250pL ELISA buffer into the appropriate tube. After the last step the remaining pellet can be discarded.

i) The final concentration of the viral antigen in the desorbed pool(s) should now 25 be the same as the original lmL of DP from which it was desorbed. Therefore, the concentration of inactivated JEV antigen content measured in the desorbed pool can be directly related to the original DP.

Note: Analyse the desorbed samples by ELISA on the day of desorption. j) Dilution of desorbed DP samples:

30 These will be tested at six dilutions in triplicate. An appropriate pre-dilution will be performed in the range of the standard curve e.g. 1 in 15 (lOOpL to 1400pL diluent) or other suitable dilution, then six dilutions of 1 in 15 pre-dilution will be made independently using sample diluent.

66

2.7. Sample Loading and Plate Plan

Prepare samples and standards before analysis.

After blocking wash the plate using the plate washer employing the JEV ELISA program. After that, remove any remaining wash buffer by decanting. Invert the plate 5 and blot it against a clean paper towel. Do not allow microtiter plate to dry between wash steps and reagent addition.

Add lOOpL/well of standards / controls / samples and cover with cover plate and incubate for 1 hour-)-/-10 min at 37°C.

Add lOOpL of sample diluent to all wells not required for testing.

10

2.8. Preparation of Primary Antibody

Remove the required number of aliquots of the primary antibody from the freezer and allow to thaw at RT. Prepare max 15 min before use Rabbit anti-JEV in sample diluent at a suitable dilution. Following sample incubation, wash the plate using the 15 plate washer employing the JEV ELISA program. After that, remove any remaining wash buffer by decanting. Invert the plate and blot it against a clean paper towel. Do not allow microtiter plate to dry between wash steps and reagent addition. Add lOOpL/well of diluted primary antibody, cover with cover plate and incubate for 1 hour +/- 10 min at 37°C.

20

2.9. Preparation of Secondary Antibody Conjugate

Remove the required number of aliquots of the secondary antibody conjugate from the freezer and allow to thaw at RT. Prepare max 15 min before use a dilution of Donkey anti-Rabbit-HRP in sample diluent; e.g. for a 1 in 10,000 dilution for make a 1 in 100 25 pre-dilution then make a second dilution of 1 in 100.

Following primary antibody incubation, wash the plate using the plate washer employing the JEV ELISA program. After that, remove any remaining wash buffer by decanting. Invert the plate and blot it against a clean paper towel. Do not allow microtiter plate to dry between wash steps and reagent addition. 30 Add 100pL/well of diluted secondary antibody conjugate, cover with cover plate and incubate for 1 hour +/- 10 min at 37°C.

2.10. Substrate Incubation

When the conjugate has been added, remove TMB from the 2-8°C refrigerator. Pipette 35 the required volume (12mL of TMB per plate) into a 50mL centrifuge tube, using a

67

serological pipette. Allow the TMB to reach room temperature in the dark. Following conjugate incubation wash the plate 3 times with the plate washer employing the JEV ELISA program. After that: remove any remaining wash buffer by decanting. Invert the plate and blot it against a clean paper towel. Do not allow microtiter plate to dry 5 between wash steps and reagent addition. Add lOOpL/well of TMB and develop the plate in the dark at Room Temperature for 10 minutes.

2.11. Stopping and Reading

After 10 minutes of TMB incubation, stop the development by adding lOOpL/well 2M 10 sulphuric acid. Read the plate at 450nm (reference filter 630nm) within 10 minutes of stopping using the BioTek reader and Gen5 Secure software.

2.12. Data analysis

NTV/DS data analysis:

15 Gen5Secure software will be used to calculate the % Recovery of the QCs,

concentration x dilutions, mean concentration of the samples corrected for dilutions and this value multiplied by 1.05 to correct for the addition of ELISA buffer.

DP data analysis:

20 GenSSecure software will be used to calculate the % Recovery of the QCs,

concentration x dilutions and mean concentration of the dilutions for the samples.

For DP Supernatant samples:

If the concentration of the supernatant sample is below the LLOQ of the assay (i.e. 25 0.05AU/ml), then the supernatant sample should be recorded as <0.05 AU/ml

If the concentration of the supernatant sample is within the LOQs of the assay (i.e. 0.05AU/ml to 1.25AU/ml), then the concentration value is recorded for the supernatant sample.

30

If the concentration of the supernatant sample is above the ULOQ of the assay (i.e. 1.25AU/ml), then the preparation of the drug product supernatant should be repeated. The supernatant sample will be re-tested by performing a suitable pre-dilution into the range of the standard curve followed by 6 sample dilutions. (The desorbed Drug

68

Product sample does not need to be repeated.) The mean concentration for the dilutions that are within the LOQs of the assay (LLOQ 0.05AU/ml to 1.25AU/ml) will be the recorded concentration value for the supernatant sample, provided that the system suitability are met and at least 4 out of the 6 sample dilutions are within the 5 LOQs.

2.13. Assay Acceptance Criteria a) The correlation coefficient for the calibration curve must be > 0.980.

b) %CVs <15% for standards and samples (except DP supernatant) for the four 10 highest concentrations of the dilutions, %CV <15% for controls c) Individual blank ODs must be < 0.2.

d) Assay controls must be within specified defined limits (for freshly prepared controls 2 out of 3 QCs should have observed concentrations within ± 30% of the nominal values; OR the levels set during QC qualification) for the assay to pass.

15 e) Assay validity will be recorded on the Gen5-print-out. If the plate fails to meet the defined acceptance criteria the assay is deemed invalid.

2.14. Reporting of data

NIV

20 a) Antigen content

The reported value for inactivated AU/mL is the mean of the concentrations (which are within the LOQs of 0.04 to 1.25AU/mL) calculated for the single sample dilutions corrected by the respective dilution factors, and the mean multiplied by a correction factor of 1.05 to account for the 5% volume of 2Ox ELISA buffer that was added to 25 each sample when it was taken. Antigen content will be recorded on Gen5 print-out. DS:

a) Identity

If the absorbances of the samples at lowest dilution (highest concentration) are higher than 3 standard deviations above the mean value of the blank the result will be 30 reported as positive b) Antigen content

The reported value for inactivated AU/mL is the mean of the concentrations (which are within the LOQs of 0.04 to 1.25AU/mL) calculated for the single sample dilutions corrected by the respective dilution factors, and the mean multiplied by a correction

69

factor of 1.05 to account for the 5% volume of 20x ELISA buffer that was added to each sample when it was taken. Antigen content will be recorded on Gen5 print-out. Desorbed DP:

a) Identity:

5 If the absorbances of the samples at lowest dilution (highest concentration) are higher than 3 standard deviations above the mean value of the blank the result will be reported as positive.

b) Antigen Content

The reported value for inactivated AU/mL is the mean of the concentrations (which 10 are within the LOQs of 0.05 to 0.8AU/mL) calculated for the single sample dilutions corrected by the respective dilution factors. Antigen content will be recorded on Gen5 print-out.

DP supernatant (degree of adsorption / degree of non-adsorption):

Degree of adsorption is reported in relationship to aluminium hydroxide formulated 15 drug substance post filtration.

a) For calculation of the reported value, the reported antigen content (AU/mL) for the respective DS sample (post filtration) corrected for the dilution with aluminium hydroxide (5%) will be set at 100% and the percentage of the concentration measured in the supernatant (corrected for the addition of 5% ELISA buffer) calculated in 20 relation to that. The reportable value will be the difference between 100% and the percentage calculated for the supernatant. Results will be reported to 2 decimal places.

DP supernatant (AU/mL) * 1.05

25 Degree of adsorption (%) = 100% - * 100%

DS (AU/mL) * 0.95

The degree of non-adsorption will be calculated as detailed below and results will be reported to 2 decimal places:

30

DP supernatant (AU/mL) * 1.05

Degree of non-adsorption (%) = * 100%

DS (AU/mL) * 0.95

70

b) In case the neat supernatant does not contain any measurable antigen (ie. observed supernatant concentration less than LLOQ, where LLOQ = 0.05 AU/mL), the LLOQ will be used for the calculation of the result. The result in this case is 5 reported as "greater than x%". For example if the DS sample is measured as

12.00AU/mL and no signal was measured in the supernatant; with the LLOQ of 0.05AU/mL then amount in supernatant is <0.05*1.05 = < 0.0525 AU/mL. The amount of DS after buffer correction is 12.00*0.95 = 11.40AU/mL, and the reported result for degree of adsorption is < 100-0.0525/11.4*100 = > 99.54%. The degree of 10 non-adsorption will also be reported (i.e. 100 - the % degree of adsorption).

Example 5 Introduction:

15 In order to further investigate the mode of action that leads to product instability/potency loss of the JEV vaccine, Ala-(His)6-OprF190-342-OprI21-83 (SEQ ID NO: i; Figure 15) - herein also referred to as "protein A" was used in a preeliminary screening assay incorporating alhydrogel lots with different metal content and spiking with copper ions and sulfite.

20

Material:

• Copper(II)chloride dihydrate (Sigma, Order no. 807483)

• lOxPBS (Gibco, Order No. 14200-091)

• 15ml Falcon tubes(Greiner, Cat. No. 188724)

25 • Incubator Infors HT Incubator Multitron Standard (InforsAG)

• Aqua bidest. (Presenilis Kabi, Art no. 0712221/01 A)

Preparation of stock solutions:

• Copper(II) stock solution t 30 20 mM Copper(II) stock solution was prepared by dissolving 341 mg of

Copper(II)chloride dihydrate in 100 mL of aqua bidest..

71

• Sodiummetabisulfite stock solution

200 mM Sodiummetabisulfite stock solution was prepared by dissolving 1.52 g of Sodiummetabisulfite in 35 mL PBS. This solution was adjusted the 5 pH to 7.3 with NaOH and filled up to a volume of 40 mL with PBS. The solution was them filtered via 0.2 p syringe filter.

Preparation of working solutions

Working solutions were prepared by dilution of metal stock solutions with 10 aqua bidest. and sterile filtration via 0.2p syringe filter. (Mini Kleenpak

25mm- Pall)

Preparation of buffer solutions

• % PBS + 0.9% NaCl

15 1 x PBS buffer solution was prepared by 1:10 dilution of 10 x PBS with aqua bidest.. The pH of this buffer solution was 7.5. PBS buffer solutions adjusted to pH 7.3 and 8.0 were prepared by adjusting the pH with HC1 or NaOH respectively.

9 g of NaCl were dissolved in 333mL of either pH 7.3 of pH 8.0 buffer 20 solution and then brought to 1000 mL with aqua bidest.. followed by filtration via 0.2p bottletop filter.

Sample preparation:

Formulations of Protein A and different lots of alhydrogel (Lot 4230 & Lot 4074) 25 were prepared in V3 PBS + 0.9% NaCl at two different pH values and were spiked with sulfite according to the following scheme:

Sample Spike

72

No.

Name pH

Alum batch

Cu(II) [ng/mL]

Sulfit [mM]

1

1711201 l_PROTEIN A_4074_ref_4°_pH7.3

7.3

4074

2

1711201 l_PROTEIN A_4074_ref_37°_pH7.3

7.3

4074

3

1711201 l_PROTEIN A_4074_Sulfit_37o_pH7.3

7.3

4074

1

4

17112011JPROTEIN A_4230_ref_4°_pH7.3

7.3

4230

5

17112011_PROTEINA_4230_ref_37°„pH7.3

7.3

4230

6

1711201 1JPROTE1N A_4230_Sulfit_3 7°_pH7.3

7.3

4230

1

7

1711201 l_PROT EIN A_4074_ref_4°_pH8

8

4074

8

17112.01 l_PROTE!N A_4074_ref_37°_pH8 171120 11JPROTEIN A_4074_Sulfit_37°_pH8

8

4074

9

8

4074

1

10

1711201 l_PROTEIN A_4230_ref_4°_pH8

8

4230

11

17112011 PROTEIN A_4230_ref_37°_pH8

8

4230

12

17112011_PROTEIN A_4230_Sulfit_37°_pH8

8

4230

1

Samples 1, 6, 11 and 16 were stored at 4°C (reference samples). All other samples were incubated at 37°C for 96 hours.

After 96 hours all samples were subjected to a desorption procedure to separate 5 the antigen from alhydrogel. The desorbed antigen was analyzed by RPC.

Results:

Results showed severe degradation of the antigen Protein A in the presence of sulfite. The degradation was more pronounced in the samples formulated with 10 alhydrogel of higher metal impurity content.

73

Table 1. Metal ion content in Aluminium hydroxide lot 4230. and 4074 analyzed by ICP-MS.

Alum Lot (2% solution)

Al

Cr

Fe

Co

Ni

Cu

Ag

Cd

W

Pb

V

Rb

Mo

Hg/mT,

ng/mL

ng/mL

ng/mL

ng/mL

ng/mL

ng/mL

ng/mL

ng/mL

ng/mL

ng/ml,

ng/mL

ng/mL

Alum Lot 4230 RQCS 0890

9570

1139

5640

7

816

64

<5

<5

<25

24

13

<5

11

Alum Lot 4074 RQCS0013

9130

20

266

<5

15

<25

<5

<5

<25

19

<5

<5

<5

Mix 50/50% of both Lots*

9130

579

2952

<5.8

415

<45

<5

<5

<25

21.6

<9

<5

<8

* Calculated residual metal content

74

Table 2: Pipetting scheme of DOE

Sample

Hg/mL

ppm

NIV dil

Alum 423C

Ai um 4074

PS Frag. Stock

Extractables Stock

CH20

2011Q819_D 0 E_s p i _1

rby^l:

100

20110819_DO E_s p l_2

: H- :::.!:-":iiob rv"'^

-'v-: ;20110819^DGE Jpl^: ^

100

:V::':;u;::i;v;'

2011081B_DQE_spI_4

■4:-v0

■ : ■■■ J.;

2011G81S_DOE_spl_5

2:

'50 r: -:;

-6.

20110819_DOE_spl_6

■ z •

■^^i^IOO' '" ■■ -

^':^0

'.v.;7

20110819_DOE_spl_7

■ ■ 7': .

; 2;.

;^.;ioo:--

20HOS19_DOE_spS_8

JS-

100

20110819_DO E_sp! 9

0

-V'iO1

20110819 DOE spl_10

20110819_DOE_spi_ll

■i1:;--0:L\':100^-

:;.;,-:40 /:

20110819_DOE_spS_12

[}^\ 7:

: : 2-;-,:

!,' :: 4.2 i';!' V ■

■ ' id: :': ■

20110819_DOE_spl_13

r 8 ;

;2> '■

40

2G110819_DOE_spl_14

7 '

" ■■■■'■

c

!;V' ' 0 '

40

20110819_DOE_spl 15 1 8

100 | 0

OV'V

75

; Ot?

oox

1?£~"]ds~3oa~6X80XX0Z j

Wy:'y:::^y-'\

^"''■yrZ^'^y.-yy':

:-v-p;:'1

;00X

y::rt .

ee_|ds~3oa""6i8oxxoz j

.-ee

OCX

7£~jds~3oa_6I80XX0Z j

"!;-v;:ps;:-:;.-:j.

y.^.y/y.

S'Z

X£~ids"3oa~6X80XX0Z j

::. ;

: :

: OOX : ;

X]\Zyy

0£~jds~3oa~6I80XX0Z j

;yyy:\yMXmXy

I ■: i ■ ■

.y;.: ";::0S.=--'=:...

OOX

; : 8..,

6Z~ids~3GCf6X80XX0Z j

6Z'.■■■■■,■ ;:\-yy p^Ofcfr 'V

001

y-'yZ.. ,

r8Z4 d s^O(3^T80TIQZ f ; j

-OS---

oox r

. Z .

8

/Z~|ds_3Oa_6X80XT0Z

vMyp^xtyy"'''::

hi- yOy^y

'r;\ i:_0 :

, r;

; y'-Z-yyy.

: H-.k :

9Z |ds~3OG~6T80XI0Z j

'y[y \9Z'y -.y^y j,:o;;

; 0.'

i ' ' ■ ' . 0 ' " -

V . 'Z'.:;-":'"

SZ~ids~3OG-6X80XT0Z j y'^y^Xy

.y'yy.'yyOSy s;;::0;':-

-;.'POT;.;.:;,:-.:

i.:/y'Z-:: y

frZ~|ds~3Oa"6I80XI0Z j

Z't?"

oox

ez~!ds_3oa_6180110z jj

fc:;r';

'■■XZV: :■*■:V:;;:

;\0

Q

go*;

XXmXyy ry-y-y^

ZZ~ids~3oa"6X80XI0Z i

XyXh^ZZ:yyty::y

I 0i7

oox

X yjy'/yy

■yr:% f/"k

XZ_|ds~3oa~6X80X:0Z j

1; ;■0t7 ;'

V :i;

oox

\yyyZyS,,

..si;:

0Z~|ds~3oa~6X80XX0Z f

0t7

OOX;

'^S}Z:\\yy

6X~|ds~3Oa~6X80XI0Z

5X

OP

l-ji'vi^-.-s-pcinc

XpZ^i:[\

8X_|ds_3OG_6X80XT0Z

8X

0

: os oox

Z'\:ry yyb ':L';yyy

Z.T |d s-3 OG~6X80I I0Z

LI

017

:,POT !:■;.•!■

\^\2r'yy

9X~|ds~3oa"6X80IX0Z |

31

panui^uoD z aiqnx

Table 2 continued

Sample

Volume [pi]

No.

Name pH

NIV

Buffer

Alum 4230

Alum 4074

PS Frag. Stock

Extractables Stock

CH20

Total

1

20110819_DOE_spl_1

7

2500

1812

250

0

125

313

0

5000

2

20110819_DOE_spl_2

7

2500

1937

0

250

0

313

0

5000

3

20110819_DOE_spl_3

7

2500

2228

0

250

0

0

22

5000

4

20110819_DOE_spl_4

8

2500

1812

0

250

125

313

0

5000

5

20110819_DOE_spl_5

7.5

2500

2125

125

125

125

0

0

5000

6

20110819_DOE_spl_6

7

2500

2103

0

250

125

0

22

5000

7

20110819_DOE_spl_7

7

2500

2125

250

0

125

0

0

5000

8

20110819_DOE_spl_8

8

2500

2125

0

250

125

0

0

5000

9

20110819_DOE_spl_9

8

2500

2250

250

0

0

0

0

5000

10

20110819_DOE_spMO

8

2500

2125

250

0

125

0

0

5000

11

20110819_DOE_spl_11

8

2500

1790

0

250

125

313

22

5000

12

20110819_DOE_spl_12

7

2500

1937

250

0

0

313

0

5000

13

20110819_DOE_spl_13

8

2500

2228

0

250

0

0

22

5000

14

20110819_DOE_spl_14

7

2500

2228

250

0

0

0

22

5000

15

20110819_DOE_spl_15

8

2500

1937

0

250

0

313

0

5000

16

20110819_DOE_spl_16

8

2500

2103

250

0

125

0

22

5000

77

Table 2 continued

17

20110819_DOE_spl_17

7

2500

1812

0

250

125

313

0

5000

18

20110819_DO E_s pi 18

7

2500

1790

0

250

125

313

22

5000

19

20110819_DO E_s pi 19

8

2500

1915

250

0

0

313

22

5000

20

20110819_DOE_spl_20

8

2500

1915

0

250

0

313

22

5000

21

20110819_DOE_spI_21

8

2500

2103

0

250

125

0

22

5000

22

20110819_DOE_spl_22

8

2500

1937

250

0

0

313

0

5000

23

20110819_DOE_spl_23

7

2500

1915

0

250

0

313

22

5000

24

20110819_DOE_spl_24

7

2500

2103

250

0

125

0

22

5000

25

20110819_DOE_spl_25

7

2500

2125

0

250

125

0

0

5000

26

20110819_DOE_spl_26

7

2500

2250

0

250

0

0

0

5000

27

20110819_DOE_spl_27

8

2500

1812

250

0

125

313

0

5000

28

20110819_DOE_spl_28

8

2500

2228

250

0

0

0

22

5000

29

20110819_DOE_spl_29

8

2500

1790

250

0

125

313

22

5000

30

20110819_DOE_spl_30

7

2500

2250

250

0

0

0

0

5000

31

20110819_DOE_spl_31

7.5

2500

2250

125

125

0

0

0

5000

32

20110819_DOE_spI_32

8

2500

2250

0

250

0

0

0

5000

33

20110819_DOE_spI_33

7

2500

1790

250

0

125

313

22

5000

34

20110819_DOE_spl_34

7

2500

1915

250

0

0

313

22

5000

sum:

85000

69014

4250

4250

2125

5015

346

170000

78

Table 3: Comparison of leachables from stopper extract and JEV09L37 SN. All peaks with an area of > O.lmAU.min are included in this table.

Retention time

Stopper extract concentrate

Stopper extract concentrate 1:16 diluted in formulation

FVL

L37

SN

Relative concentration compared to FVL

(min)

Peak area (mAU.min)

(%)

12.40

0.79

0.05

0.10

47

13.14

1.91

0.12

n.d.

additional peak compared to FVL

13.48

0.81

0.05

n.d.

additional peak compared to FVL

13.74

0.43

0.03

n.d.

additional peak compared to FVL

14.10

0.75

0.05

n.d.

additional peak compared to FVL

14.41

0.49

0.03

0.25

12

16.12

29.64

1.85

0.45

414

16.71

2.75

0.17

n.d.

additional peak compared to FVL

17.15

14.68

0.92

0.11

815

18.68

0.90

0.06

0.57

10

20.08

0.70

0.04

0.25

18

20.75

0.42

0.03

n.d.

additional peak compared to FVL

21.27

0.54

0.03

n.d.

additional peak compared to FVL

22.13

2.84

0.18

0.15

122

22.74

0.55

0.03

0.17

20

23.74

1.74

0.11

0.27

41

24.88

0.98

0.06

0.12

51

27.63

0.84

0.05

0.16

32

29.46

0.72

0.04

n.d.

additional peak compared to FVL

31.40

0.39

0.02

n.d.

additional peak compared to FVL

35.23

4.70

0.29

0.41

72

SUM

67.59

4.2

3.0

140

79

Table 4: DOE results obtained after 4 and 8 weeks at 22°C. Antigen was desorbed from Alum and analysed by ELISA (monoclonal and polyclonal). * (x-fold increase compared to FVL)

v weeks at 22°C

4: :8:^eeks: at 22-C -4-■

Sample

P H

Alum 4 2 3 0 (°/9

Spiked PS Frag.

(ng/mL)

S p ike d leachables *

Spiked Fct rm&lii:

(ppm)

Mono.;-/ i:.ELISA^:' ..(AU/mL).

ELISA.:::;. (AU/mL)

Ratio

-ELISA. ■

XAfflmL):

-:;P.o1y4-tAU/iiiL)

Ratio i

7

100

50

1.4

0

4-9.7254:

::412.814>:

0.76

4i:3:;649;i::-

m.575X:

2

7

0

0

1.4

0

:15.254 v

16.331

0.93

9.865

i.^:;8v039::.

40.815-

3

7

0

0

0

40

: 12.457

-12.614

0.99

9.037

:-..'0.:904::

4

8

0

50

1.4

0

13.513

12.971 •

1.04

4o;3284

:;;a923;:'

5

7.5

50

50

0

0

: 13.592

14.924

0.94-

:: 40.358-

8.032":.v

40.775::

6

7

0

50

0

40

: 12.942:

13.878

4;:;o.93 4:

10.008

-:;0i:989-.::.

7

7

100

50

0

0

-9.649

12.608

■ :-6.0894::

4;. :4,092:>:-

8

8

0

50

0

0

11.184

11.902

0.94

:'49;113::4"

:-^048'-::

9

8

100

0

0

0

11.436

12.883

: -0,89 .

7.409

0.836

10

8

100

50

0

0

13.361

15.516

0.86

-■10.525-

8.158

11

8

0

50

1.4

40

13.209

13.608

0.97

9.201

0.984

12

7

100

0

1.4

0

8.4

11.913

0.71-

:ii::5?b0i:;4

2.95!

0.590

13

8

0

0

0

40

10.294

10.483

0.98

8.284

1.033

14

7

100

0

0

40

9.972

12.015

0.83

5.802

0.780

15

8

0

0

1.4

0

14.096

15.61 §

0.94-r

"':1®;192":

9.531

0.935

80

Table 4 continued

16

8

100

50

0

40

9,784

'"':::13.514::7

:>40,724;

11.011

>5>:;B:947':":;:'

40.8134

17

7

0

50

1.4

0

..11:213^

: 414.285:,-:

440:784v

410246: ^

::J:3774::

0.818

18

7

0

50

1.4

40

■;iEi834

0.97

10.539

:-0i875":

19

8

100

0

1.4

40

10.536

■ 10.935;-

0.96

10.151

8.341

0 822

20

8

0

0

1.4

40

;: 10.026 /

9.654:, ;•

1.04

■4;ii:3d64::

4i::-9;281,,4

21

8

0

50

0

40

10.149

9.986

1.02

/,11.213:4

4- - 9.205

::-&;82l4

22

8

100

0

1.4

0

10.051

12.193

0.82

■ ;:lO.:918:Ji:

6.841

0.627

23

7

0

0

1.4

40

10.024

10,74

" ■ 0*93^4

-10:711;:

:-0:7434:

24

7

100

50

0

40

9.535

10.254

:;v0.93: :

4:10:642: :

rMSlp

25

7

0

50

0

0

11.143

11,765

4 0.95

4 13,054-;

xoms-

26

7

0

0

0

0

•••■'11.431:''

11.796

0.97

13.753

:; 110.3914

mm:-

27

8

100

50

1.4

0

10.09

11.953

0.84 :

11.506

mm'rr

28

8

100

0

0

40

: 10.137

411.223 ■

,::.;4O.9..:

4:4:i:i:4;:'

:: 48:131^

4©,7394

29

8

100

50

1.4

40

9.605

10,535

0.91

10.848

:: ! 7-939

4o.;7324:

30

7

100

0

0

0

6.485

$379

0.77

:.;-;7:957;4:

3.523

40.4434

31

7.5

50

0

0

0

11.081

12,229

0.91

4:12.377''::

4 9;5:354:

:'0.77O;

32

8

0

0

0

0

11.421

12.002

0.95

:;;4II,36,:'::

9.859": i;.::

0.868

33

7

100

50

1.4

40

9.03

10.583

0.85

9.753

6.571

0.674

34

7

100

0

1.4

40

9.05

4 10.927

0.83

>11,7154"

6.872

4oy587-

81

Table 5: Analysis of Variance for ratio Monoclonal/Polyclonal ELISA after storage 4 weeks at 22°C

Analysis of Variance for Ratio 4 weeks

Source

Sum of Squares

Df

Mean Square

F-Ratio

P-Value

A: pH

0.02205

1

0 . 02205

5 .

.48

0

. 0326

B:Alum

0.117612

1

0.117612

29.

.20

0

. 0001

C:PS Fragments

0.0008

1

0.0008

0 ,

.20

0

.6618

D:Extractables

0 . 0008

1

0.0008

0 .

.20

0

.6618

E:Formaline

0.0242

1

0.0242

6 .

. 01

0

. 0261

AB

0.0001125

1

0

.0001125

0 ,

. 03

0

. 8694

AC

0.00045

1

0.00045

0 ,

. 11

0

. 7425

AD

0.01125

1

0.01125

2 .

. 79

0

. 1141

AE

0.00405

1

0.00405

1,

. 01

0

.3309

BC

0.0000125

1

0

.0000125

0 .

.00

0

. 9563

BD

0.0010125

1

0

.0010125

0 .

.25

0

. 6229

BE

0.0006125

1

0

.0006125

0 .

. 15

0

. 7017

CD

0.0008

1

0.0008

0 .

.20

0

. 6618

CE

0.0008

1

0.0008

0.

,20

0

. 6618

DE

0.0072

1

0.0072

1.

. 79

0

. 1999

Total error

0.0644375

16

0 .

00402734

Total (corr.) 0.2562 31

R-squared = 74.8488 percent

R-squared (adjusted for d.f.) = 51.2695 percent Standard Error of Est. = 0.0634614 Mean absolute error = 0.0352734 Durbin-Watson statistic = 1.47556

82

Table 6: Analysis of Variance for ratio Monoclonal/Polyclonal ELISA after storage at 22°C for 8 weeks.

Analysis of Variance for Ratio 8 weeks

Source

Sum of Squares

Df

Mean Square

F-Ratio

P-Value

A: pH

0 .102378

1

0 .102378

11,

. 19

0

.0041

B:Alum

0.275653

1

0 .275653

30 ,

. 13

0

. 0000

C:PS Fragments

0.00300312

1

0 .00300312

0 ,

.33

0

. 5747

D:Extractables

0.0108781

1

0.0108781

1

.19

0

.2917

E:Formaline

0.0318781

1

0.0318781

3 ,

.48

0

. 0804

AB

0.00137813

1

0.00137813

0 ,

. 15

0

. 7031

AC

0.00382812

1

0.00382812

0 ,

.42

0

. 5269

AD

0.00137813

1

0.00137813

0 ,

. 15

0

. 7031

AE

0.0166531

1

0.0166531

1.

. 82

0

. 1961

BC

0.000253125

1

0.000253125

0 .

. 03

0

. 8700

BD

0.00382813

1

0.00382813

0 .

.42

0

.5269

BE

0.00195313

1

0.00195313

0 .

.21

0

. 6503

CD

0.00195313

1

0.00195313

0 .

.21

0

. 6503

CE

0.000153125

1

0.000153125

0 .

. 02

0

. 8987

DE

0.00525313

1

0.00525313

0 .

. 57

0

.4596

Total error

0 .1464

16

0.00915

Total (corr.) 0.606822 31

R-squared = 75.8743 percent

R-squared (adjusted for d.f.) = 53.2565 percent Standard Error of Est. = 0.0956556 :

Mean absolute error = 0.0571484 Durbin-Watson statistic = 0.888586

83

Table 7: Regression analysis for "Ratio 8 weeks" including pH, Alum, and Formaldehyde.

Regression coeffs. for Ratio 8 weeks constant = 0.0228125

A:pH = 0.113125

B:Alum = -0.00185625

E:Formaline =0.0315625

Table 8: Estimation of results "Ratio 8 weeks" generated using the fitted model.

Estimation Results for Ratio 8 weeks

Observed Fitted Lower 95.0% CL Upper 95.0% CL

Row Value Value for Mean for Mean

1

0.

58

0.5975

0 .

.536766

0 .

658234

2

0.

81

0.783125

0 .

. 722391

0 .

843859

3

0

i. 9

0.84625

0 .

.785516

0 .

906984

4

0 .

92

0.89625

0 .

.835516

0 .

956984

6

0 .

99

0.84625

0 .

.785516

0 .

906984

7

0 .

67

0.5975

0 .

.536766

0 .

658234

8

0 .

99

0.89625

0 .

.835516

0 .

956984

9

0 .

84

0.710625

0.

.649891

0.

771359

10

0 .

78

0.710625

0.

.649891

0.

771359

11

0 .

98

0.959375

0 .

.898641

1

.02011

12

0 .

59

0.5975

0 .

.536766

0 .

658234

13

1.

03

0.959375

0 .

.898641

1

. 02011

14

0.

78

0.660625

0 ,

.599891

0 .

721359

15

0.

94

0.89625

0 ,

.835516

0 .

956984

16

0 .

81

0.77375

0 ,

.713016

0 .

834484

17

0 .

82

0.783125

0 ,

.722391

0 .

843859

18

0 .

88

0.84625

0 .

.785516

0 .

906984

19

0 .

82

0.77375

0 .

.713016

0 .

834484

20

0 .

82

0.959375

0 .

.898641

1

..02011

21

0 .

82

0.959375

0 .

.898641

1

..02011

22

0 .

63

0.710625

0.

. 649891

0.

771359

23

0.

74

0.84625

0.

.785516

0.

906984

24

0 .

66

0.660625

0

.599891

0.

721359

25

0 .

67

0.783125

0.

. 722391

0.

843859

26

0 .

76

0.783125

0 .

.722391

0 .

843859

27

0.

64

0.710625

0

. 649891

0 .

771359

28

0.

74

0.77375

0

.713016

0 .

834484

29

0.

73

0.77375

0

.713016

0 .

834484

30

0 .

44

0.5975

0

.536766

0 .

658234

32

0 .

87

0.89625

0

.835516

0 .

956984

33

0.

67

0.660625

0

.599891

0.

721359

34

0 .

59

0.660625

0

.599891

0 .

721359

84

98

OS

00^

Z

OoZZ /Hd 9>l!dssd OOf(ll)nO AIN

91-

OS

W&0®xmmv

L

QoZ2~ZHd~s>l!dssd_000r(ll)!N~AIN

P i

OS

009

L

0°22 ZHd 8>|!dssd 009_(ll)!N AIN

£1

09,

00 ^

L

0o22~ZHcT8>|!dssd 001-_(ll)!N_AIN

21-

OS

L

0=22 ZHd 8>j!ds Sd AIN

n

.' '..v. : ■ :

L

Oo22 ZHd 0001. (lll)yO AIN

01

00S

.

L

0o22~ZHd 009 (lll)yo AIN

6

001

•

'

L

Oc22_ZHd~OOl- (lll)yO~AIN

8

0001-

L

Oo22 ZHd 0001- (ll)no AIN

Z

009

L

Oc22_ZHd"009lll)no_AIN

9

.

001-

L

0=22 ZHd 00 L (ll)no AIN

9

'

0001

I

Oo22_ZHd—0001- (ll)lN AIN

P

009

L

Oo22 ZHd 009 (ll)lN AIN

£

001

L

Oo22_ZHd~OOI. (ll)lN AIN

Z

L

Oo22 ZHd pa>|!dsun"AIN

I

Sd

P8}U0UJ6BJJ

(into

2IO(ll)nO

1?0S(ll)!N

Hd

0LUB[\J

'ON

"1/Brl

0|duiBS

J- io(iii)jo

I 2IO(ll)nO WV14 :|BU8}BUi AIN

V K)S(ll)!N iri 0002 :ewn|OA |B}oj.

ElAJUj] |os >|oo?s

UEfd Buipady;

(AIN)SI60II0S luaurugdxa JOJ uny :q aiqej,

98

OS

ooo i.

8

0=22~8Hd~a>l!cissd"OOOr(lll)yO_MN

Ofr

OS

'

8

0o23_8Hd~9>i!dssd~00S~(lli)y0"AIN

68

OS

1

:■ ..". ■ :;.. .., ..:■■ ■ , .. ■■ .-. . ■

8

OoZZ~8Hd~a>i!dssd-OOr(lll)idO~AIN

88

0001.

!llilli!il3?Si:f;;?':

8

Oo22~8Hd~9>i!dssd~OOOr(ll)no"AIN

ze

OS

oos

8

0 022"8Hd"a>| |d s s d~00Slll)no~AI N

9e

OS

001

8

0 o32"8Hd~8>i !d s s d~00 nil)no_AI N

se

OS

000^

8

Oo33-8Hd~9>l!dssd_OOOr(ll)!N_MN

1?£

. oos

8

OoS2_8Hd~9^!dssd_OOS~(ll)!N_AIN

se

oot

8

Oo2Z"8Hd~a>l!dssd~OOr(ll)!N_AIN

se

Wm&WM

8

Oo22~8Hd~9>!!ds Sd~AIN

i-e

WBMMWk

8

Oo22_8Hd~000 r(lll)yO~AIN

oe

is?;;® V /X

8

Oo22~8Hd~OOSllll)yO_AIN

62

Si^illl

8

0o22~8Hd~ 00 r(lll)yO~AIN

82

000!.

8

0o22~8Hd~0001- (11)nO AlN

Z2

OOS

8

Oo3Z~8Hd~OOSlll)no~AIN

92

1

001-

8

0o22~8Hd~00r(ll)no_AIN

S2

0001.

8

0o22"8HcT000 (ll)!N_AIN

172

8

Oo22-8Hd~OOS_(ll)!N~AIN

£2

|||||||i;l'M|| |I«

00^

8

OoZZ~8Hd~OOr(ll)!N~AIN

22

llllllill!llli

8

0=22_8Hd_PB>l!dsun_AIN

\Z

OS

0001-

Z

Oo22~/Hd_9>l!dssd"OOOr(lll)yO_AIN

02

OS

00S

L

0o22_ZHd_9>l!dssd—00S~(lll)yo_AIN

61-

OS

001 1

'

L

Oo23~ZHd_e>l!dssd_OOr(lll)yO_AIN

81.

OS

V.C != :-;!i; K = =» ^ ■; J :<: = :="

0001-■

.

L

0=22~ZHd~9>l!dssd 000f(li)no~AIN

11

OS

003

L

0o22~ZHd~9>l!dssd OOS (ll)no~AIN

91

■ Dorini^noo 6 ajqej,

18

oooz

OS

0

g

0

LPi

009 i

L

Oo22~ZHd~8>i!dssd 001- (ll)no AIN

si.

oooz

OS

0

0

P£

9ii

0081

L

0,22 ZHd e>i!dssd 0001. (ll)!N AIN

Pi

OOOZ

OS

0

0

LI

ceu

0081.

L

0=22 ZHd 0>|!dssd OOS (ll)lN AIN

Si oooz

OS

0

0

£

LP i

0081-

L

Oo22 ZHd 9>|!dssd 001- (ll)!N AIN

Zi oooz

OS

0

0

0

OS i

0081-

I

Oo22~ZHd~8>l!ds Sd~AIN

a

0002

0

8£

0

0

Z9i

0081.

L

0„22~ZHd~000 r(lll)yO~AIN

Oi

0002

0

61.

0

0

i8i

0081-

L

Oo22~ZHd_OOS (lll)yo AIN

6

0002

0

P

0

0

96 i

0081-

L

Qo22~ZHd_001. (Ill)yo AIN

8

0002

0

0

l-E

0

69 i

0081-

L

Oo23_ZHd 000 i (ll)no AIN

L

0002

0

0

91-

0

P8i

0081-

L

Oo2Z ZHd-OOS (ll)no AIN

9

0002

0

0

S

0

L6i

0081-

L

O„22~ZHd~00l. (ll)no AIN

S

0002

0

0

0

PZ

991-

0081.

L

0=22 ZHd 0001- (ll)lN AIN

P

0002

0

0

0

Li

£81-

0081-

L

0»22 ZHd OOS (ll)lN AIN

£

oooz

0

0

0

S

Z61-

0081-

L

G„22~ZHd~00 r(ll)!N~AIN

2

OOOZ

0

0

0

0

002

0081-

L

Oo22~ZHd_pe>l!dsun-AIN

i

Sd pajuaiuBej^

(ink)

2IO(ll)nO

K)S(ll)!N

sad

AIN

Hd aiuBN

■ON

[|r1] 8iun|0A

8|daiBS

no !N

[PW/B] Ml/M

-|UI/6LU-JJ ;

Ij Sd Bbj] 1 :ps >{001S

isaAijippe j0Lj}o

079 ,9:sg Z'8S

panui^uoo q afqej,

88

oooz

OS

8£

0

0

Zii

0081-

8

Oo3Z~8Hd~a>i!dssd~OOOnill)y3-AIN

017

OOOZ

OS

61.

0

0

i£i

0081

8

Oo32_8HcT9>l!dssd~OOSllll)yO_AIN

6£

OOOZ

OS

p

0

0

9PI

0081-

8

Oo22~8Hd~s>l!dssd_OOr(lll)yO~AIN

8£

oooz

OS

0

iS

0

6ii

0081

8

3o2S~8Hd~e>l!dssd_OOOr(ll)no_AIN

LS

oooz

OS

0

91

0

PSi

008U

8

Oo2S~8Hd~8>i!dssd_00Slll)n0~AIN

9£

oooz

OS

0

£

0

LP I

0081-

8

3o32~8Hd~a>l!dssd~00f(ll)nD_AIN

S£

oooz

OS

0

0

PS

911

0081-

8

Oo22~8Hd~e>l!d ssd~000 r(ll)!N~AIN

PS

oooz

OS

0

0

Li

SSi

0081-

8

OoZZ-8Hd-e>l!dssd_OOSlll)!N~AIN

ss

oooz

OS

0

0

£

LP I

0081-

8

3o2S"8Hd_a>l!dssd_00r(ll)!N_AIN

zs

oooz

OS

0

0

0

OS1-

0081-

8

3oZZ~8Hd_a>1!ds Sd~AIN

iS

oooz

0

8£

0

0

291.

0081.

8

3o22~8Htf~000r(lll)y3_AIN

OS

oooz

0

61.

0

0

1.81.

0081

8

3o22_8Hd~00Sllll)y3~AIN

6Z

oooz

0

p

0

0

961

0081

8

3<,Z2_8Hd~00 ill ll)y3~AIN

8Z

oooz

0

0

1£

0

691

0081-

8

OoZZ~8Hd~000 U—(ll)n3_AIN

LZ

oooz

0

0

91.

0

P8i

0081.

8

3oZ2~8Hd_00Slll)n3~AIN

9Z

oooz

0

0

£

0

16 i

0081-

8

3oZ2~8Hd_00 111 l)nO~AIN

SZ

oooz

0

0

0

PS

991

0081-

8

3oZZ_8Hd~000 r(ll)!N~AIN

PZ

oooz

0

0

0

LI

esi-

0081

8

OoZZ-8Hd_00S—(ll)!N~AIN

ez

oooz

0

0

0

£

Z61

0081-

8

3oZZ_8Hd_00r(ll)!N~AIN

zz

oooz

0

0

0

0

OOZ

0081-

8

0,,ZZ~8Hd~ pa>|!dsun_AIN

l-Z

oooz

OS

8£

0

0

Zii

0081-

L

OoZZ_ZHd_8>l!dssd_OOOnill)yO-AIN

oz

oooz

OS

61.

0

0

iSi

0081

L

OoZZ_ZH(r3>l!dssd_OOSllll)yO_AIN

61.

oooz

OS

p

0

0

9Pi

0081

L

OoZZ~ZHd_e>l!dssd_OOr(lll)yO_AIN

81-

oooz

OS

0

IS

0

61. i

0081-

L

OoZZ ZHd a>|!dssd 0001- (ll)no AIN

Li

oooz

OS

0

9i

0

PSi

0081

L

OoZZ ZHd a>|!dssd OOS (ll)no~AIN

91-

panui^uoo 6

68

0001-

8

8Hd 0001- (lll)JQ da

0Z

OOS

.* , ;

8

8Hd oos (into da

6i

001.

8

8Hd~oot (into-da

Qi

.

y^j£. O O: O

8

8Hd_000i (ll)no~da

Li

OOS

8

8Hd~00S_(ll)no~da

91-

001-

8

8Hd_00r(ll)no~da si.

0001-

8

8Hd 0001- (ll)!N da n

OOS

8

8Hd"00Slll)!N~da

£i

001.

8

8Hd~00r(ll)!N~da

Zi

'

8

8Hd pe>l!dsun da a

0001,

Z

zHd~ooor(nito_da

OS-

^ 009

,

z

ZHd 00S (lllto~da

6

OOi-

z

ZHd~ooninto~da

8

; '^.<r::r:': -i ^ f!-S^

O •O

;/0

I

ZHd~000r(ll)no_da z

■ OOS

L

ZHd OOS (ll)no da

9

001-

L

ZHd 001- (ll)no da s

0001.

L

ZHd-0001- (ll)!N da p

'

OOS

L

ZHd OOS (ll)!N da

£

00 f.

L

ZHd 001. (ll)!N da

Z

L

ZHd- p8>|!dsun_da

I

Sd P8}U9LUBB4

(ink)

SIO(li)no tt>S(ll)!N

Hd eiiiBN

'ON

l/Brl

a|dujBS

G!0(ill)J0 ZIO(ll)nO t-OS(i!)!N [1AILU] |0S >|00}S

>||nq Z8QK-Iri 0000!-

_ -da

:8Lun|0A leioj.

\

ue|d Buspedftj>

(da)ei60II0S ^uerauadxa JOJ UBIJ :ox aiqux

06

00001-

0

261-

0

0

80S

00S6

8

8Hd~000r(lll)JCrda

02

00001-

0

96

0

0

POP

00S6

8

8Hd~00S (into da

61-

00001-

0

61-

0

0

18P

00S6

8

8Hd~oonmkTda

81-

00001-

0

0

151

0

£P£

00S6

8

8Hd_000r(ll)no"da

LV

0000 s-

0

0

6L

0

IZP

00S6

8

8Hd~00S_(ll)no~da

91

00001-

0

0

91

0

P2P

00S6

8

8Hd~00l- (ll)nQ da

9V

00001-

0

0

0

0Z1-

oee

00S6

8

8Hd~"0001- (ll)!N da

PI

00001

0

0

0

S8

91P

00S6

8

8Hd~~00Slll)lN~da

U

00001-

0

0

0

LI

£8P

00S6

8

8Hd~001- (ll)!N da

Zl

00001-

0

0

0

0

OOS

00S6

8

8Hd~p9>j!dsun da

W

00001-

0

261-

0

0

8oe

00S6

Z

ZHd~OOOr(lll)JO~da

01-

00001-

0

96

0

0

POP

00S6

Z

ZHd~OOSllll)-iO~da

6

00001-

0

61-

0

0

IQP

00S6

Z

ZHd 001- (into da

8

00001

0

0

ZSI-

0

£P£

0096

Z

ZHd 0001 (ll)nD da

Z

00001-

0

0

61

0

UP

00S6

z

ZHd OOS (ll)no da

9

0000I

0

0

91

0

P2P

00S6

z

ZHd 001- (ll)no dO

S

00001

0

0

0

0ZI-

ose

00S6

z

ZHd 0001- (ll)!N da

P

00001.

0

0

0

S8

SIP

00S6

z

ZHd"00S (ii)in da

£

00001-

0

0

0

LI

£817

00S6

z

ZHd 001- (ii)in da

Z

00001-

0

0

0

0

OOS

00S6

z

/Hd_ps>|!dsun da

I

16101

Sd ps}u8uj6e4

(lll)JO

210(ll)n0

tOS(ll)lN

jsijng da

Hd

01UB[S|

'ON

Qri] aoinpA

3|dlilBS

!•

pemii^uoo oi

Table 11: Plan for experiment 20110812-Metal Ion Spiked DP

Pipetting plan i

Total Volume: DP:

30000 Ml 11D8T bulk

Stock Sol [mM] Fe(ll)CI3 Fe(lll)CI3 Ni(ll)S04 Co(ll)C!2 . Cu(ll)CI2

Zn(ll)S04

Cr(lll)CI3

MW [g/Mol]

Fe 55.9

Fe 55.9

Ni 58.7

Co 58.9

Cu 63.6

Zn 65.4

Cr 52.0

! FVL ng/mL ||jM

1 2B2 5 049

282 5.045

] 41 0.698

0.33 0.006 j 3 0.04/

Sample

Mg/L

No.

Name

PH

DP dil

Fe(ll)CI3

Fe(lll)CI3

Ni(ll)S04

Co(ll)CI2

Cu(ll)CI2

Zn(ll)S04

1

DP_Fe(ll)_pH7

7

1

500

2

DP_Fe(!li) pH7

7

1

500

'

3

DP_Ni(ll)_pH7

7

1

.

500

4

DP_Co(ll)_pH7

7

1

'

500

5

DP_Cu(ll)_pH7

7

1

500

'

6

DP_Zn_pH7

7

1

.

500

7

DP_metalmix_pH7

7

1

500

500

500

500

500

8

DP_unspiked_pH7

7

1

-

9

DP_Fe(ll)_pH7.4

7.4

1

500

10

DP Fe(lll)_pH7.4

7.4

■ 1

.

500

11

DP Ni(II)_pH7.4

7.4

1

500

12

DP_Co(ll)_pH7.4

7.4

1

500

13

DP_Cu(ll)_pH7.4

7.4

1

500

14

DP_Zn_pH7.4

7.4

500

15

DP_metalmix_pH7.4

7.4

■ ■ 1

500

500

500

'500

500

500

16

DP_unspiked_pH7.4

7.4

1

■: : : ■ ■ "

17

DP_Fe(ll)_pH7.8

7.8

1

soo

18

□P_Fe(lll)_pH7.8

7.8

1

500

19

DP_Ni(ll)_pH7.8

7.8

1

,

500

20

DP_Co(ll)_pH7.8

7.8

1

'

500 |

21

DP_Cu(ll)_pH7.8

7.8

'

500

22

DP_Zn_pH7.8

7.8

1

'

500

23

DP_metalmix_pH7.8

7.8

1

500 ■

500

500

500

. 5W

500

24

DP_unspiked_pH7.8

7.8

1

91

Table 11 continued

Sample

M3/L

No.

Name pH

DP dil

Cr(lll)

25

DP_CR(lll)_pH7

7

1

500

26

DP_Cr(III)_pH7.4

7.4

1

500

27

DP_Cr(lll)_pH7.8

7.8

1

500

; : i !

Sample

Volume [pi]

No.

Name

PH

NIV dil

Buffer

Fe(il)CI3

Fe(lfl)CI3

Ni(!l)S04

Co(ll)CI2

Cu(l!)CI2

Zn(ll)S04

Total

1

DP_Fe(ll)_pH7

7

27750

1981

269

0

0

0

0

0

30000

2

DP_Fe(lll)_pH7

7

27750

1982

0

268

0

0

0

0

30000

3

DP_Ni(II)_pH7

7

27750

1981

0

0

269

0

0

0

30000

4

DP_Co(ll)_pH7

7

27750

1994

0

i1.

0

0

256

0

0

30000

5

DP_Cu(ll)_pH7

7

27750

1995

0

r

0

0

0

255

0

30000

6

DP_Zn_pH7

7

27750

2014

0

0

0

0

0

236

30000

7

DP_metalmix_pH7

7

27750

698

269

" 268

269

256

255

236

30000

8

DP_unspiked_pH7

7

27750

2250

0

0

0

0

0

0

30000

9

DP_Fe(ll)_pH7.4

7.4

27750

1981

269

0

0

0

0

0

30000

10

DP_Fe(lll)_pH7.4

7.4

27750

1982

0

268

0

0

0

0

30000

11

DP_Ni(ll)_pH7.4

7.4

27750

1981

0

0

269

0

0

0

30000

12

DP_Co(ll)_pH7.4

7.4

27750

1994

0

0

0

256

0

0

30000

13

DP_Cu(ll)_pH7.4

7.4

27750

1995

0

0

0

0

255

0

30000

14

DP_Zn_pH7.4

7.4

27750

2014

0

0

0

0

0

236

30000

15

DP_mefalmix_pH7.4

7.4

27750

698

269

268

269

256

255

236

30000

16

DP_unspiked_pH7.4

7.4

27750

2250

0

0

0

0

0

0

30000

17

DP_Fe(ll)_pH7.8

7.8

27750

1981

269

0

0

0

0

0

30000

18

DP_Fe(lll)_pH7.8

7.8

27750

1982

0

268

0

0

0

0

30000

19

DP_Ni(ll)_pH7.B

7.8

27750

1981

0

0

269

0

0

0

30000

20

DP_Co(ll)_pH7.8

7.8

27750

1994

0

*

0

0

256

0

0

30000

21

DP_Cu(ll)_pH7.8

7.8

27750

1995

0

0

0

0

255

0

30000

22

DP_Zn_pH7.8

7.8

27750

2014

0

0

0

0

0

236

30000

23

DP_metalmix_pH7.8

7.8

27750

698

269

268

269

256

255

236

30000

24

DP_unspiked_pH7.8

7.8

27750

2250

0

0

0

0

0

0

30000

92

Table 11 continued

Sample

Volume [fjl]

No.

Name

PH

^ j Cr(lll)CI3

-

8

DP_CR(lll)_pH7

7

~ "— - - r 48

—

--

16

DP_Cr(lll)_pH7.4

7.4

" ——. . .

- - - . [ 48

—— ■ -. . .

—-— .

24

DP_Cr(lll)_pH7.8

7.8

—- .

- - . . f 48

"

"

"

-

93

Table 12: Antigen recovery determined by SEC-HPLC of exp. 20110913(NIV). For pH 7 and pH 8 the recoveries are based on non-spiked NIV control samples #1 and #21, respectively. Samples were stored at 22°C for 3 weeks. Samples marked with "n.a" were not analyzed due to sample prioritization

Results

Sample

Area mAU*s Recovery (%)

1

NIV_unspiked_pH7_22°C

4.054

100.0

2

NIVJNi(II)„10Q_.pH7_22°C';; :.:

3.972

98.0

S

NIV_Ni(H). 500.. pH7. 22°C

4.454

109.9

4

NI^M(IILl0Q0_j>H7i_22pG; ■; : =

>4.185r;

103.2:

5

NIV_Cu(II)_100_pH7_22°C

3.913

96.5

6

NIV_Cu(II)_500_pH7_22°C

3.733

92.1

7

NIV_Cu(Il). 1000 pH7_.22°C

3.115

76.8

8

NIV_CE(IH)..100_pH7 22°C

97.6

9

Nr^Cp(mL5P0_jpH7„22°C

4.000 "J;

^98.7

10

NrV_CR(III)_1000__pH7_22oC

3.611 :

89.1 .

11

NIV PS spike_pH7_22° C

4.068

100.3

NI V_N i (II )__10 0__PSsp i ke_.pII7_22 °C

3.884

95.8

::13::.v;;:|

;mV^i(II)_500JPSspike_pH7^2:2^G ;:;

3.688 ::

91.0

d^0.

NIV_Ni(II)_10:00_PSspike_pII7^22oC; : i

3.971 i

98.0

15

NIV..Cu(II) .100..PSspike_pH7_22°C

3.568

88.0

16

NIV_Cu(II)_500_PSspikc„pH7. 22°C

3.325

82.0

17

NIV_Cu(II)_1000_PSspike__pH7_22°C

3.486

86.0

18

NI¥k:GsaiIL100jPSspike^pH7^22°G ::

::;3.747;

19

NI\^GE(III)_f5p0JSspike^H7i_j2°G

-3.904

96.3

20

? NI^GE(in)_1000J'SspiKe^pHT^a^C:

3.685

90.9

21

NIV...U nspiked._pH8_22°C

4.213

100.0

22

NIV_Ni(II)_100_pH8_22°C

4.181

99.2

23

NIV_Ni(II)_500_pH8_22°C

4.150

98.5

24

NIV_Ni(II)_1000_pII8_22oC

3.772

89.5

2 ">

NIV_Gii(IIkl0(^pH8^22,:,C:

4.146

j 98.4;;

26

NIF^G^II^QO^HS^S^G; /

100.0 ; ^':W

27

NIV_Cu(II)_1000_pH8...22°C

4.152

"38.6 ^ ;:

28

NIV CR(IIlU00.pH8_22°G

4.213

100.0

29

NIV_CR(III)J500_pH8_22°C

3.997

94.9

30

NIV CR(III)_1000_pH8__22°C

4.231

100.4

94

31

NIV PS spike_pII8_22°C

4.150

98.5

32;

NrV_Ni(n^lO(LPSspikejpHS^0 C i

3.(523

8(5.0

33

NIV_Ni(II)i\.500„PSspike„pH8„22°G: ;

■ 3.725

88.4

34,;

; NIV^i(ntl000_PSspil^H8_22aG;;

4.079 ;

96,8

35

NIV__Cu(II)_100_PSspike_pH8_22cC

3.473

82.4

36

NIV_Cu(II)_500_PSspike_pH8_22°C

3.180

75.5

37

NIV_Cu(II)_1000_PSspike_pH8„22°C

4.056

96.3

38

NW_GE(in)_100_PSspike_pH8_22o(3;:

3.042

^2-2

;S9 ;:;

NIV_GR(I1D_500. ..PSspike.. pil8_._22 °G

4.113

97.B

40

NIV. GR(III) .1000. PSspiku_pH8_2 2 "C

n.a.

'n.a;

Table 13: ELISA results of exp. 20110913(N1V) obtained after 3 weeks at pH 8 (sample 21-40) and 7 weeks at pH 7(sample 1-20) at 22°C. Samples marked with "n.a" were not analyzed due to sample prioritization.

20110913_metal spiked_NIV_22°C 7 weeks at pli 7

No.

Name poly mono ratio

1

NIV_unspiked_pH7_220C CONTROL

20.699

17.512

0.846

2 \

NI^Ni(II)_100_pH7_22°G

17.243!

14.787

0.858

3

N]^^i(II)_500_pH7_22°C : ;

18.877 i

16.215

0.859

4

NIV_Ni(II)_1000_pH7_22°C

16.9

14.522:

0.859

5

NIV_Cu(II)_100_pH7_22°C

16.718

13.278

0.794

6

NIV_Cu(II) 500_pH7_22°C

14.664

5.459

0.372

7

NI V_C u(II)_1000_pH 7_2 2° C

7.112

0.421

0.059

BX:

NI V_CR(III)_10 0_pH7_2 2 °C

19.207

16.602

0.864

9

NIV_CR(III)_500_pH7_22°C

20.313

17.84; n

0.878

10

NIV_CR(III)_1000_pH7_22°C

16.762

13.128

0.783

11

NIV_PS spike_pH7_22°C CONTROL

19.907

16.994

0.854

12

NIV_Ni(II)_100_PSspike_pH7_22°C

19.546

16.534

0.846

13

NIV_Ni(II)_500_PSspike_pH7_22°C

18.337

15.2

0.829

14

NIV_Ni(II)_1000_PSspike_pH7_22oC

20.759

16.934

0.816

15

NIV_Cu(II)_100_PSspike_pH7_22oC

18,249

13.348

0.731

16

:NI^Cu(IlE500ijPSspikeLpH7_22oC ;;

19.201

77:733 ;

0.403

17

NIV_Cu(II)_1000_PSspike._pH7_22°C

:8.5i5;:

0.711

0.083

18

NIV_CR(III)_100JPSspike_pH7_22°C

18.377

16.521

0.899

19

NIV_CR(IIIL500_PSspike_pH7_22°C

19.678

17.119

0.870

20

NIV_CR(III)_1000_PSspike_pH7_22°C

20.505

18.219

0.889

95

20110913.. metal spiked,N1V__22°C 3 weeks at. pH 8

1st anal;

Vrsis

2nd ana ysis

No

poly mono ratio poly mono ratio

21

NIV_unspiked_pH8_22°C CONTROL

22.729

20.679

0.910

24.687

24.806

1.005

22

NIV. Nidi).. 100_pH8_22°C

23.732

23,572;

0.993

24.179

22.148 r

0.916

23

NI V_N i (IT )_500_pH 8_22° C

20.086 j

19.793:

0.985

22.411

22.207

0.991

:24::;;:

;NIVxNi(II)J1000jpH8^22°C :;

16.553

15.402

0.930

23,841;

21.645 ;

0.908

25

NIV Cu(II)_ 100_pH8_22° C

18.736

18.175

0.970

28.024

24.7.14

0.882

26

NIV_Cu(II)_500_pH8_22°C

21.173

19.109

0.903

25.774

23.983

0.931

27

NIV_Cu(II)„1000_pH8_22°C

19.709

16.406

0.832

24.799

21.580

0.870

28 ■ ;

NIVHGR(III)i,100IpH8122°lG :

22:464

20.687 \

0.921

22;782>

21.166!:

0.929

29

;NIV7CR(III)iS(WjH8^22°C V;

20.527 7

;:;20247:;:::

0:986

23.040

21 JoM-

0.940

30

K1V CR(lIl) 1000 pi!8 22 0

; 20.838:

19,094;:

0,916

25.676

r24.047:;

0.937

31

NIV. PS spikc;_pTT8_22"C CONTROL

19.051

20.112

1.056

25.413

25.991

1.023

32

NIV_Ni(II)_100_PSspike_pH8_22°C

18.729

19.250

1.028

n.a.

n.a.

n.a.

33

NIV Ni(II) 500_PSspike_pH8. 22°C

20.923

20.516

0.981

n.a.

n.a. ' :

n.a.

34

NIV_Ni(II).J.000_PSspike_pIT8_22°C

21.734

19.794

0.911

25.164

24.89,9

0.989

;jr>

.NIVLCu(II)_100^PSspike^H8l22°G ;:

21.526

19.852

0.-922

n.a.

■■n.a,;;;.:''-

n.a.

36

NIV_Gu(II)„5007PSspik6£_pH8_22°C

21.914

19.449

0.888

n.a.

n.a.

;37

NlfcCTi(n)„i000„PSlspifcejpH8^22oC\

18.646

16.259

0.872

25.371

24.275 ;

0.957

38

NI V_CR( III)_ 100. .PSspike _pH8_22 ° C

20.292

18.667

0.920

n.a.

n.a. '

n.a.

39

NIV CR(IlI)_500_PSspike. pH8_22°C

22.835

21.558

0.944

25.561

24.480

0.958

40

NIV_CR(III)_1000_PSspike_pH8_22°C

27.862

25.291

0.908

n.a.

n.a.

n.a.

Table 14: Antigen recoveries after 5 weeks at 22°C of desorbed JEV obtained by SEC-HPLC. Recoveries were based on non-spiked DP control samples stored at either pH 7 or pH 8

JEV area

No

Sample mAU.min recovery

1

DP ..imspiked_pH7 22°C

3.710

100%

2

DP_Ni(II)_100_pH7 22°C

3.656

99%

3

DP Ni(II)_500_pII7 22°C

3.705

100%

4

DP_Ni(II)_1000_pH7 22°C

3.444

93%

5

DP_Cu(II)_100__pH7 22°C

3.565

96%

6

DP_Cu(II)_500_pII7 22°C

3.313

89%

7

DP_Cu(II)_1000jpH7 22°C

3.367

91%

8

DP Cr(III)_100. pH7 22°C

3.562

96%

9

DP_Cr(III). 500 pH7 22°C

3.422

92%

10

DP_Cr(IH) J-000_.pH7

3.148

85%

96

22°C

11

DP_unspiked_pH8 22°C

4.297

100%

12

DP_Ni(n)_100_pH8 22°G

4.029

94%

13

DP_Ni(II)_500_pH8 22°C

4.306

100%

14

DP„Ni(II)_1000_pH8 22°C

4.065

95%

15

DP,. Cu (I IL 100...pH8 22°C

3.751

87%

16

DP_Cu(ID_500_pII8 22°C

3.698

86%

17

DP_Cu(II)_1000_pH8 22°C

3.511

82%

18

DP_Cr(III)_100_pH8 22°C

3.805

89%

19

DP_.Cr(llI). .500. pH8 22°C

3.843

89%

20

DP_Cr(III)_1000_pH8 22°C

4.212

98%

Table 15: ELISA results of desorbed JEV antigen after 5 weeks at 22°C

No.

Name pH

Poly. ELISA

Mono. ELISA

Ratio

% Ratio compared to non-spiked Control

1

DP_.unspiked._pH7

7

14.203

13.13

0.924

100

2

DP_Ni(II)_100_pH7

7

13.089

12.623

0.964

104

3

DP_Ni(II)_500„pH7

7

12.640

12.572

0.995

108

4

DP_Ni(II)_1000_pH7

7

15.051

11.757

0.781

84

5

DP_Cu(II)_100j>H7

7

13.420

10.792

0.804

87

6

DP_Cu(II)_500_4>H7

7

13.247

10.079

0.761

82

7

DP_Cu(II)_1000_pH7

7

12.981

9.654

0.744

80

8

DP_Cr(III)_100_pH7

7

16.936

11.886

0.702

76

9

DP_Cr(III)_500_pH7

7

13.991

11.219

0.802

87

10

DP_Cr(III)_1000_pH7

7

13.061

10.438

0.799

86

11

DP.unspiked _pH8

8

12.647

11.287

0.892

100

12

DP_Ni(II)_100_pH8

8

12.308

10.689

0.868

97

13

DP_Ni(II)_500_pH8

8

14.300

12.623

0.883

99

14

DP_Ni(II)_1000_pH8

8

12.082

10.930

0.905

101

15

DP_Cu(II)_100_pH8

8

11.041

9.937

0.900

101

16

DP_Cu(II)_500_pH8

8

9.869

9.176

0.930

104

17

DP_Cu(II)_1000_pH8

8

9.379

8.802

0.938

105

18

DP_Cr(III) .100, pH8

8

10.164

9.545

0.939

105

97

19

DP_Cr(III)_500_pH8

8

11.241

10.057

0.895

100

20

DP_Cr(III)_1000_pH8

8

12.400

11.183

0.902

101

Table 16: ELISA results of desorbed JEV antigen after 4 weeks and 7 weeks stored at 22°C. Samples marked with "n.a" were not analyzed due to sample prioritization

4 weeks @ 22°C

7 weeks @ 22°C

No

pH

poly mono ratio poly mono ratio m/p

1

DP Fe(II) pH7

7

14.285

11.213

0.785

14.181

11.378

0.802

2

DP Fe(III) pH7

7

14.879

11.552

0.776

14.323

11.765

0.821

3

DP Ni(II) pH7

7

16.572

11.862

0.716

14.231

11.666

0.820

4

DP Co(II) pH7

7

12.81

12.629

0.986

14.246

11.244

0.789

5

DP Cu(II) pH7

7

12.474

.9.747

0.781

11.464

7.654

0.668

6

DP Zn(II) pH7

7

13.131

11.186

0.852

15.122

11.514

0.761

25

DP Cr(III) pH7

7

12.144

11.598

0.955

13.543

10.73

0.792

7

DP metalmix pH7

7

11.078

8.366

0.755

8.617

5.392

0.626

8

DP iraspiked pII7

7

16.159

13.224

0.818

14.158

11.559

0.816

9

DP Fe(IDpH7.4

7.4

15.096

12.649

0.838

14.417

11.239

0.780

10

DP Fe(III) pH7.4

7.4

13.509

12.029

0.890

16.948

12.344

0.728

11

DP Ni(II) pH7.4

7.4

13.036

11.306

0.867

13.293

12.571

0.946

12

DP Co(II) pH7.4

7.4

13.715

11.714

0.854

13.079

11.96

0.914

13

DP Cu(II) pH7.4

7.4

14.235

11.748

0.825

10.843

9.276

0.855

14

DP Zn(II) pH7.4

7.4

13.815

12.882

0.932

13.28

12.619

0.950

26

DP Cr(III) pH7.4

7.4

12.324

11.659

0.946

13.312

10.804

0.812

15

DP_metalmix_pH7.4

7.4

n.a.

9.74

6.551

0.673

16

DP unspiked pII7.4

7.4

13.951

13.225

0.948

11.97

12.672

1.059

17

DP Fe(II) pH7.8

7.8

14.356

13.102

0.913

12.625

13.304

1.054

18

DP Fe(III) pII7.8

7.8

13.554

12.388

0.914

13.003

10.811

0.831

19

DP Ni(II) pH7.8

7.8

13.949

12.496

0.896

13.106

11.757

0.897

20

DP Co(II) pH7.8

7.8

12.826

11.931

0.930

13.333

11.059

0.829

21

DP Cu(II) pH7.8

7.8

12.593

11.268

0.895

12.538

9.872

0.787

22

DPZn(II) pH7.8

7.8

15.217

14.204

0.933

15.55

13.217

0.850

27

DP Cr(III) pi 17.8

7.8

12.977

13.228

1.019

14.642

11.975

0.818

23

DP metalmix pH7.8

7.8

11.196

9.811

0.876

10.771

7.539

0.700

98

DP unspiked

24

pH7.8

7.8

11.906

11.819

0.993

12.472

11.034

0.885

Table 17: ANOVA for stability samples stored at 22°C for 7 weeks.

Analysis of Variance for Ratio - Type III Sums of Squares

Source

Sum of Squares

Df

Mean Square

F-Ratio

P-Value

MAIN EFFECTS A:Metal Type B: pH

0.139643 0 . 0462028

8 2

0 . 0174554 0. 0231014

3.00 3 . 97

0.0293 0 . 0398

RESIDUAL

0.0931046

16

0 . 00581904

TOTAL (CORRECTED)

0 .27895

26

All F-ratios are based on the residual mean square error.

99

Table 18: Multiple range test for ratio by metal ion type. l=Fe(II); 2=Fe(III); 3=Ni(II); 4=Co(II); 5=Cu(II); 6=Zn(II); 7=Cr(HI); 8=Mix[l-6]; 9=non-spiked control

Multiple Range Tests for Ratio by Metal Type

Method: 95.0 percent LSD

Metal Type Count LS Mean Homogeneous Groups

8

3

0 .

.665087

X

5

3

0 ,

.770168

XX

2

3

0 .

.793725

XXX

7

3

0 ,

.807248

XX

4

3

0.

.844388

XX

6

3

0 .

.853867

XX

1

3

0.

.878563

XX

3

3

0 .

.887505

XX

9

3

0 .

.919926

X

Contrast Difference +/- Limits

1 -

- 2

0.084838

0 .

132037

1 -

- 3

-0.0089421

0 .

132037

1 -

- 4

0.0341754

0 .

132037

1 -

■ 5

0.108395

0 .

132037

1 -

- 6

0.0246961

0 .

.132037

1 -

- 7

0.0713155

0.

. 132037

1 -

- 8

*0.212476

0.

. 132037

1 •

- 9

-0.0413627

0 .

. 132037

2 -

- 3

-0.0937801

0 .

. 132037

2 -

- 4

-0.0506626

0 .

.132037

2 ■

- 5

0.0235569

0.

.132037

2 -

- 6

-0.0601419

0 .

. 132037

2 ■

- 7

-0.0135225

0 .

. 132037

2 -

- 8

0.127638

0 ,

. 132037

2 -

- 9

-0.126201

0 ,

.132037

3 -

- 4

0.0431175

0 .

.132037

3 -

- 5

0.117337

0 .

.132037

3 ■

- 6

0.0336382

0 .

.132037

3 •

- 7

0.0802576

0 .

.132037

3 ■

- 8

*0.221418

0 .

.132037

3 ■

- 9

-0.0324206

0 .

.132037

4 ■

- 5

0.0742195

0.

.132037

4 ■

- 6

-0.00947935

0 .

.132037

4 ■

- 7

0.0371401

0.

.132037

4 ■

- 8

*0.1783

0

.132037

4 •

- 9

-0.0755381

0

.132037

5 ■

- 6

-0 . 0836988

0

.132037

5 •

- 7

-0.0370794

0

.132037

5 ■

- 8

0.104081

0

.132037

5 •

- 9

*-0.149758

0

. 132037

6 ■

- 7

0.0466195

0

.132037

6 ■

- 8

*0.18778

0

.132037

6 ■

- 9

-0.0660588

0

.132037

7 •

- 8

*0.14116

0

.132037

7 ■

- 9

-0.112678

0

.132037

8 ■

- 9

*-0.253838

0

.132037

* denotes a statistically significant difference.

100

Table 19: ICP-MS results of residual metal ion impurities present in various

Alum (2%) lots

Residual metal content (ng/mL)

Alum (2%)Lot

Cr

Fe

Ni

Cu

V

Co

4074

19.8

266

14.8

<25

<5

<5

4470

1637

1179

17.5

<25

<5

<5

4563

1874

2485

8.9

<25

<5

<5

4621

1333

1183

7.6

<25

<5

<5

3877

48.2

183

12.2

<25

<5

<5

4230 (nonGI**, GI***)

1139

5640

816

64

12.6

7

Mix* 4074/4230

579.4

2953

415.4

<44.5

<6

<6

''Calculated content of residual metals based on Alum lot 4230 and 4074 **nonGI: non gamma irradiated ***GI: gamma irradiated

101

Table 20: Summary of metal ion content and analysis of DP samples formulated with various Alum lots. Samples were analysed in duplicate by ELISA and the ratio of monoclonal/polyclonal ELISA is reported. Formulations were stored at 22°C for 6 weeks.

#

Alum Lot (2% Stock solution)

Monoclonal 1st analysis

Polyclonal 1st analysis

Monoclonal 2nd analysis

Polyclonal 2nd analysis

Ratio

1st analysis

Ratio

2nd analysis

Mean ratio (Mono/poly)

Range* Ratio

1

4470

23.584

26.397

21.666

22.948

0.893

0.944

0.918

0.025

2

4563

23.027

23.397

19.051

18.862

0.984

1.010

0.997

0.013

3

4621

22.758

24.056

19.041

19.196

0.946

0.991

0.968

0.023

4

3877

23.5

23.85

21.186

21.24

0.985

0.997

0.991

0.006

5

4230

(non gamma irradiated)

21.85

25.155

20.682

23.792

0.868

0.869

0.869

0.000

6

4230

(gamma irradiated)

20.509

22,904

18.002

23.512

0.895

0.765

0.830

0.065

7

4074

23.022

24.047

16.695

19.866

0.957

0.840

0.898

0.058

8

Mixture (50%/5Q%) of 4074 and 4230

20.833

22.954

22.473

21.217

0.908

1.059

0.983

0.076

*Range is the abso ute difference between 1st and 2nd analysis.

102

Table 21: overview of Alhydrogel® lots used in JEV production lot #

Brenntag ID

Ammonia alum source used for # of JEV batches

1

4074

BK Giulini

17

2

4230

Holland (USA)

24

3

4414

Canton India

7

4

4470

Canton India

19

5

4539

Holland (USA)

4

Table 22: Results of PSD analysis of Alhydrogel® (2%) stock solution in water.

No

Sample Name a (o.i)

d (0.5)

d (0.9)

Obscuration (%)

1

Non-irradiated AlOH RQCS0890 Lot 4230

0.70

2.13

46.53

1.51

2

GI AlOH RQCS1200 Lot 4230

0.71

4.14

69.64

1.98

3

GI AlOH RQCS1342 Lot 4740

0.78

2.23

53.44

2.11

4

GI AlOH RQCS0448 Lot 4074

0.73

4.49

78.58

1.96

(0.1): 10% of all measured particles have a diameter aelow t lis value d(0.5): 50% of all measured particles have a diameter below this value d(0.9): 90% of all measured particles have a diameter below this value Obscuration: amount of laser light reduction by sample; corresponds to concentration of sample in measurement chamber

103

Table 23: Results of Alhydrogel® titration curves for determination of POZ. Samples were analyzed in PBS (1:20 dilution).

Sample ID

PZC (pH)

Non-irradiated AlOH RQCS0890 Lot 4230

4.58

GI AlOH RQCS 1200 Lot 4230

4.62

GI AlOH RQCS 1342 Lot 4740

4.49

GI AlOH RQCS0448 Lot 4074

4.48

104

Table 24: Summary of metal ion analysis for various Aluminum hydroxide stock solutions; Note: A 2% stock solution equals lOmg/mL

of Al

Alhydrogel® (2% solution)

Al

Fe

Ni

Cu

Co

Cr

Ag

Cd

W

Pb

V

Rb

Mo

jjg/mL

ng/mL

Lot 4074 (RQCS0013)

9130

266

15

<25

<5

20

<5

<5

<25

19

<5

<5

<5

Lot 4230 (RQCS 0890)

9570

5640

816

64

7

1139

<5

<5

<25

24

13

<5

11

Lot 4470 (RQCS1254)

9560

1179

18

<25

<5

1637

<5

<5

<25

20

<5

<5

22

Lot 4414 (RQCS1220)

10272

2790

36

<25

<5

1710

<5

<5

35

<25

<7

n.a.

n.a.

Lot 4539

n.a.

943

119

<25

<5

276

<5

<5

<25

27

<5

<5

<5

Lot 3877

10766

183

12

<25

<5

48

<5

<5

<25

10

<5

<5

<5

Lot 4187

14100

3617

172

<25

<5

2333

<5

<5

<25

18

<5

<5

<5

Lot 4287

9800

2047

296

<25

<5

620

<5

<5

<25

30

10

<5

<5

Lot 4563

9360

2485

9

<25

<5

1874

<5

<5

<25

8

<5

<5

<5

Lot 4621

9760

1183

8

<25

<5

1333

<5

<5

<25

8

<5

<5

<5

Lot 4580 (7xwashed)

10497

2610

27

<25

<5

1470

<5

<5

<25

<25

<5

n.a.

n.a.

Lot 4596 (7xwashed)

10776

3530

27

<25

<5

1710

<5

<5

<25

<25

<5

n.a.

n.a.

Lot 4577 (7xwashed)

10720

3060

24

<25

<5

1650

<5

<5

<25

<25

<5

n.a.

n.a.

average

10359

2272

121

n.a.**

<5

1217 >

• <5-

<5 '•••

18*

11*

<5

16*

stdev

1312

1538

225:

745 ;

8

2

8

min :"

9130 :

183

s

<25

<5

20

• <5;

<5

<25

8

10 ..

<5

<5

14100

5640

816

64

.7'.-'

2333

<5:>

: <5--

35 /

30

13 -

-.<57 '

22

BSD (%).

13

68

"186-

61

:44 -

14

47

^results below LOQ were not used for average calculation; **no average calculation possible due to results below LOQ

105

Table 25: Analysis of supernatant and Aluminum hydroxide (Lot 4230) gel fraction for contaminating metal ions shows metal ions are located in the gel, not the supernatant

Sample

Fe

Ni

Cu

Co

Cr

Ag

Cd

W

Pb

V

ng/mL

Lot 4230 Supernatant

82

12

<25

<5

7

<5

<5

<25

70

<5

Lot 4230 Sediment

6200

920

<25

<5

1200

<5

<5

<25

45

13

% supernatant compared to sediment

1.3

1.3

n.a.

n.a.

0.6

n.a.

n.a.

n.a.

155.6

n.a.

Table 26: Overview on raw materials used for Alhydrogel© production

Ammonia Alum lot

Origin

Resulting

Alhydrogel batch

# JEV production batches

6-80578-28

BK Giulini

4074

17 (e.g. JEV08A02)

AA0427

4230

24 (e.g. JEV09L37)

91480

Canton India

4470

19 (e.g. JEV10P54)

10094

unknown

4563

Not used

10095

106

Table 27: Results of ICP-MS analysis for Ammonia Alum raw materials; lots in bold were used in JEV vaccine production

Ammonia

Alum

Used for

Alhydrogel®

Lot

Al

Fe

Ni

Cu

Co •

Cr

Ag

Cd w

Pb

V

mg/kg

Pg/kg

6-80578-28

4074

60400

770

16

<25

<10

60

<10

<10

<25

<25

<10

AA0427

4230

59800

25700

3560

210

25

6470

<10

15

<25

40

70

91480

4470

59800

8200

50

<25

<10

11700

<10

<10

<25

<25

16

10094

4563

59800

16000

20

<25

<10

11200

<10

<10

<25

<25

20

10095

4563

60100

19400

20

<25

<10

13100

<10

<10

<25

<25

25

ng/mL

24% Ammonia solution # 4140716

<2,0

1,5

28

<0,2

0,9

<0,2

<0,2

<1,0

<0,2

<0,2

Table 28: Comparison of Ammonia Alum and corresponding Alhydrogel® samples

Fe

Cr

Ni

Cu rel %

Ammonia Alum Lot

6-80578

88

7

2

3

Alum Lot

4074

82

6

5

8

Ammonia Alum Lot

AA0427

72

18

10

1

Alum Lot

4230

74

15

11

1

Ammonia. Alum Lot

91480

41

59

0

0

Alum Lot

4470

41

57

1

1

Table 29: Summary of 70 element scan performed at ESG. Only those elements where differences were observed between individual lots are shown.

Alum 4074

Ammonia Alum B-805Y9-2C

Alum 4563

Ammonu-Alum 10094

Alum 4563

Amnionic Alum 1009S

Laboratory Reference

ASCI 2666.002

Laboratory Reference

ASC / 2666.003

Laboratory

Reference

ASC/ 2666.004

LOD

mg/kg

LOD

mg/kg

LOD

mg/kg

Al

3

59000

Al

3

59000

Al

3

59000

Ca

0.8

<0.8

Ca

0.8

<0.8

Ca

0.8

<0.8

Cr

0.2

<0.2

Cr

0.2

4.9

Cr

0.2

6.5

Fe

0.3

0.4

Fe

0.3

18

Fe

0.3

25

K

3

5

K

3

200

K

3

200

Li

0.003

<0.003

Li

0.003

<0.003

Li

0.003

<0.003

Mg

0.05

<0.05

Mg

0.05

1.6

Mg

0.05

2.6

Mo

0.03

<0.03

Mo

0.03

<0.03

Mo

0.03

<0.03

Na

1

3

Na

1

<1

Na

1

<1

Ni

0.3

<0.3

Ni

0.3

<0.3

Ni

0.3

<0.3

Rb

0.02

<0.02

Rb

0.02

1

Rb

0.02

1.1

Zr

0.02

<0.02

Zr

0.02

<0.02

Zr

0.02

<0.02

108

Table 29 continued

Alum 4470

Amnionic; Aliirn ?148U

Alum 4230

r-'.rllllKJlifc-

Allllil rN,W>l

Laboratory Reference

ASCI 2666.005

Laboratory Reference

2666.007

LOD

mg/kg

LOD

mg/kg

Al

3

58000 47

Al

3

59000

Ca

0.8

Ca

2

<2

Cr

0.2

4.4

Cr

0.8

<0.8

Fe

0.3

10

Fe

0.8

32

K

3

180

K

5

5600

Li

0.003

<0.003

Li

0.02

0.2

Mg

0.05

0.3

Mg

0.5

42

Mo

0.03

<0.03

Mo

0.02

0.08

Na

1

<1

Na

1

110

Ni

0.3

<0.3

Ni

0.2

•1 1

Rb

0.02

0.61

Rb

0.05

1.2

Zr

0.02

<0.02

Zr

0.01

0.14

The following terms used in this specification are acknowledged as Registered Trade Marks:

"Triton", "Falcon", "UltiMate", "Infers", "Multitron", "LoBind", "Kleenpak", "Eppendorf', "Superose", "Millipore", "Steriflip", "BioTek" and "Tween".

109

SEQUENCE LISTING

<110> Intercell AG

<120> Aluminum compounds for use in therapeutics and vaccines

<130> ICP104/GB-PR-1

<160> 1

<170> Patentln version 3.5

<210> 1

<211> 223

<212> PRT

<213> Artificial Sequence <220>

<223> OprF-OprI fusion protein from pseudomonas ae.

<400> 1

Ala His His His His His His Ala Pro Ala Pro Glu Pro Val Ala Asp 15 10 15

Val Cys Ser Asp Ser Asp Asn Asp Gly Val Cys Asp Asn Val Asp Lys 20 25 30

Cys Pro Asp Thr Pro Ala Asn Val Thr Val Asp Ala Asn Gly Cys Pro 35 40 45

Ala Val Ala Glu Val Val Arg Val Gin Leu Asp Val Lys Phe Asp Phe 50 55 60

Asp Lys Ser Lys Val Lys Glu Asn Ser Tyr Ala Asp lie Lys Asn Leu 65 70 75 80

Ala Asp Phe Met Lys Gin Tyr Pro Ser Thr Ser Thr Thr Val Glu Gly 85 90 95

His Thr Asp Ser Val Gly Thr Asp Ala Tyr Asn Gin Lys Leu Ser Glu 100 105 110

Arg Arg Ala Asn Ala Val Arg Asp Val Leu Val Asn Glu Tyr Gly Val 115 120 125

Glu Gly Gly Arg Val Asn Ala Val Gly Tyr Gly Glu Ser Arg Pro Val 130 135 140

110

Ala Asp Asn Ala Thr Ala Glu Gly Arg Ala lie Asn Arg Arg Val Glu 145 150 155 160

Ser Ser His Ser Lys Glu Thr Glu Ala Arg Leu Thr Ala Thr Glu Asp 165 170 175

Ala Ala Ala Arg Ala Gin Ala Arg Ala Asp Glu Ala Tyr Arg Lys Ala 180 185 190

Asp Glu Ala Leu Gly Ala Ala Gin Lys Ala Gin Gin Thr Ala Asp Glu 195 200 205

Ala Asn Glu Arg Ala Leu Arg Met Leu Glu Lys Ala Ser Arg Lys 210 215 220

111

112

Claims (22)

Claims

1. A method for preparing an aqueous pharmaceutical or vaccine composition comprising aluminium, a reactive compound and a protein said method comprising

- selecting an aluminium-salt that is able to provide an aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition and

- combining said aluminium salt, said reactive compound, said protein and water to produce said aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition;

wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof.

2. A method according to claim 1, further comprising buffering said aqueous composition at a pH of between 6.5 and 8.5.

3. A method according to claim 1-2, further comprising packaging aliquots of said aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition in separate air-tight storage containers.

4. A method for preparing a clinical grade aluminium-salt precipitate for incorporation into a medicament and/or vaccine, said method comprising preparing an aqueous solution of aluminium ions and precipitating said aluminium-ions from said solution, and determining the level of a heavy metal in the solution and/or the aluminium-salt precipitate, wherein the precipitate is selected that is able to provide an aqueous composition comprising less than 350 ppb heavy metal based on the weight of the aqueous composition.

113

5. An aqueous pharmaceutical or vaccine composition comprising a protein, a reactive compound and an aluminium-salt, said composition comprising less than 350 ppb heavy metal based on the weight of the aqueous composition, wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof.

6. An aqueous pharmaceutical or vaccine composition according to claim 5, wherein said heavy metal is selected from Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V, Cr, Pb, Rb and Mo.

7. An aqueous pharmaceutical or vaccine composition according to claim 5-6, wherein said heavy metal is selected from Cu, Ni, W, Co, Os, Ru, Cd,

Ag, Fe, V.

8. An aqueous pharmaceutical or vaccine composition according to claim 5-7, wherein said heavy metal is selected from Cu or Ni.

9. An aqueous pharmaceutical or vaccine composition according to claim 5-8, wherein said heavy metal is present in ionic form.

10. An aqueous pharmaceutical or vaccine composition according to claim 5-9, wherein the aluminium-salt is al u mi ni umhy droxid e (Al(OH)3) or aluminiumphosphate (A1P04).

11. An aqueous pharmaceutical or vaccine composition according to claim 5-10, wherein the aluminium-salt is aluminiumhydroxide (Al(OH)3).

12. An aqueous pharmaceutical or vaccine composition according to claim 5-11, wherein the reactive compound is selected from the group

114

consisting of formaldehyde, ethanol, chlorophorm, trichloroethylene, acetone, 4-( 1,1,3,3-tetramethylbutyDphenyl-polyethylene glycol, deoxycholate, diethylpyrocarbonate, sulphite, Na2S20s, beta-proprio-lacton, polysorbate such as polyethylene glycol sorbitan monolaurate, polyethylene glycol sorbitan monooleate, O2, phenol, pluronic type copolymers, and a combination of any thereof.

13. An aqueous pharmaceutical or vaccine composition according to claim 5-12, comprising between 5 pg/ml and 50 mg/ml aluminium.

14. An aqueous pharmaceutical or vaccine composition according to claim 5-13, comprising between 50 pg/ml and 5 mg/ml aluminium.

15. An aqueous pharmaceutical or vaccine composition according to claim 5-14, comprising between 5 ppb and 250 ppb Fe based on the weight of the aqueous composition.

16. An aqueous pharmaceutical or vaccine composition according to claim 5-15, comprising less than 3 ppb Cu based on the weight of the aqueous composition.

17. An aqueous pharmaceutical or vaccine composition according to claim 5-16, comprising less than 40 ppb Ni based on the weight of the aqueous composition.

18. An aqueous pharmaceutical or vaccine composition according to claim 5-17, wherein said protein is a therapeutic and/or a vaccine.

19. An aqueous pharmaceutical or vaccine composition according to claim 5-18, wherein said protein is a viral or bacterial protein.

115

20. An aqueous pharmaceutical or vaccine composition according to claim 5-19, wherein said viral protein is a protein of the Japanese encephalitis virus or a protein of the Pseudomonas aeruginosa bacterium.

21. An aqueous pharmaceutical or vaccine composition according to claim 5-20, wherein said protein is protein within a formaldehyde inactivated virus particles.

22. A vaccine comprising an aqueous vaccine composition according to claim 5-21.

22. An aqueous pharmaceutical or vaccine composition according to claim 5-21, further comprising sulphite.

23. A vaccine comprising an aqueous vaccine composition according to claim 5-22.

116

Claims

1. A method for preparing an aqueous pharmaceutical or vaccine composition comprising aluminium, a reactive compound and a protein said method comprising

- selecting an aluminium-salt that is able to provide an aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition and

- combining said aluminium salt, said reactive compound, said protein and water to produce said aqueous composition (i) having less than 350 ppb heavy metal based on the weight of the aqueous composition and (ii) comprising between 5 pg/ml and 50 mg/ml aluminium;

wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof.

2. A method according to claim 1, further comprising buffering said aqueous composition at a pH of between 6.5 and 8.5.

3. A method according to claim 1-2, further comprising packaging aliquots of said aqueous composition having less than 350 ppb heavy metal based on the weight of the aqueous composition in separate air-tight storage containers.

4. A method for selecting a clinical grade aluminium-salt precipitate for incorporation into a medicament and/or vaccine, said method comprising preparing an aqueous solution of aluminium ions and precipitating said aluminium-ions from said solution, and determining the level of a heavy metal in the solution and/or the aluminium-salt precipitate, wherein the precipitate is selected that is able to provide an aqueous composition comprising (i) less than 350 ppb heavy metal based on the weight of the aqueous composition and (ii) between 5 pg/ml and 50 mg/ml aluminium.

117

5. An aqueous pharmaceutical or vaccine composition comprising a protein, a reactive compound and an aluminium-salt, said composition comprising (i) less than 350 ppb heavy metal based on the weight of the aqueous composition and (ii) between 5 pg/ml and 50 mg/ml aluminium, wherein the reactive compound is selected from the group consisting of a redox active compound, a radical building compound, a stabilizing compound and a combination of any thereof.

6. An aqueous pharmaceutical or vaccine composition according to claim 5, wherein said heavy metal is selected from Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V, Cr, Pb, Rb and Mo.

7. An aqueous pharmaceutical or vaccine composition according to claim 5-6, wherein said heavy metal is selected from Cu, Ni, W, Co, Os, Ru, Cd, Ag, Fe, V.

8. An aqueous pharmaceutical or vaccine composition according to claim 5-7, wherein said heavy metal is selected from Cu or Ni.

9. An aqueous pharmaceutical or vaccine composition according to claim 5-8, wherein said heavy metal is present in ionic form.

10. An aqueous pharmaceutical or vaccine composition according to claim 5-9, wherein the aluminium-salt is aluminiumhydroxide (Al(OH)3) or aluminiumphosphate (A1P04).

11. An aqueous pharmaceutical or vaccine composition according to claim 5-10, wherein the aluminium-salt is aluminiumhydroxide (Al(OH)3).

12. An aqueous pharmaceutical or vaccine composition according to claim 5-11, wherein the reactive compound is selected from the group consisting

118

of formaldehyde, ethanol, chlorophorm, trichloroethylene, acetone, 4-(l, 1,3,3-tetramethylbutyl)phenyl-polyethylene glycol, deoxycholate, diethylpyrocarbonate, sulphite, Na2S20s, beta-proprio-lacton, polysorbate such as polyethylene glycol sorbitan monolaurate, polyethylene glycol sorbitan monooleate, Oa, phenol, pluronic type copolymers, and a combination of any thereof.

13. An aqueous pharmaceutical or vaccine composition according to claim 5-12, comprising between 50 pg/ml and 5 mg/ml aluminium.

14. An aqueous pharmaceutical or vaccine composition according to claim 5-13, comprising between 5 ppb and 250 ppb Fe based on the weight of the aqueous composition.

15. An aqueous pharmaceutical or vaccine composition according to claim 5-14, comprising less than 3 ppb Cu based on the weight of the aqueous composition.

16. An aqueous pharmaceutical or vaccine composition according to claim 5-15, comprising less than 40 ppb Ni based on the weight of the aqueous composition.

17. An aqueous pharmaceutical or vaccine composition according to claim 5-16, wherein said protein is a therapeutic and/or a vaccine.

18. An aqueous pharmaceutical or vaccine composition according to claim 5-17, wherein said protein is a viral or bacterial protein.

19. An aqueous pharmaceutical or vaccine composition according to claim 5-18, wherein said viral protein is a protein of the Japanese encephalitis virus or a protein of the Pseudomonas aeruginosa bacterium.

119

20. An aqueous pharmaceutical or vaccine composition according to claim 5-19, wherein said protein is protein within a formaldehyde inactivated virus particles.

21. An aqueous pharmaceutical or vaccine composition according to claim 5-21, further comprising sulphite.