WO2017167768A1 - Novel vaccine composition - Google Patents

Abstract

The present invention relates to novel vaccine compositions, vials comprising such compositions, as well as to uses of and to methods for producing said compositions. In particular, the invention relates to vaccine compositions comprising certain excipient components and their use for the vaccination against influenza virus disease.

Description

NOVEL VACCINE COMPOSITION

Technical field

The present invention relates to novel vaccine compositions, as well as processes for preparing such compositions and to uses of the compositions. In particular, the invention relates to influenza vaccine compositions that comprise of a excipient for stabilising the vaccine compositions.

Background of the invention

Flu vaccines are typically designed to protect against three different influenza viruses (trivalent influenza vaccine or TIV). These vaccines include two influenza A viruses with a B virus chosen either from the Yamagata or Victoria lineage, despite circulation of both Yamagata and Victoria viruses during most influenza seasons. The more recent quadrivalent influenza vaccine (QIV) is designed to protect against four different influenza viruses; two influenza A viruses and two influenza B viruses, one from each lineage. The inclusion of a second B virus to the vaccine gives a broader protection against circulating influenza viruses.

Concerns over the safety of thimerosal in vaccines have led to its removal from most vaccine formulations in recent years. Three single-dose presentations of thimerosal-free formulations of quadrivalent seasonal split virion influenza vaccines (QIV) have become available in the US market (FluLaval®, Fluarix® and Fluzone®).

The stabiliser systems used in the GlaxoSmithKline (GSK) FluLaval® Quadrivalent and Fluarix® Quadrivalent formulations consist of a combination of Tween® 80 and a-tocopheryl succinate or vitamin E succinate (VES) with or without Triton X-100® in the presence of phosphate buffered saline solution. In these formulations VES together with non-ionic surfactants, Tween® 80 and Triton X-100® are useful to form stabilising micelles (WO02/097072). Fluzone® Quadrivalent is formulated in sodium phosphate saline with Triton X-100®, formaldehyde and gelatin. The stabiliser performs the critical role of preventing the bioactive haemagglutinnin antigens (HA) from undergoing aggregation resulting in a significant decline in their activities over the designated shelf-life of 12 months when stored under refrigerated conditions. Previously, WO02/097072 established the usefulness of VES in the stabilisation of HA. However, VES is a poorly soluble waxy solid at room temperature. To deal with this, existing vaccines manufacturing processes have used a large amount of Tween® 80 in order to dissolve the VES in the final vaccine formulation.

Such high amounts of stabilisers are not always acceptable under local regulatory provisions. There is a need therefore for provision of alternative methods for stabilising influenza vaccines.

Summary of the invention

Studies on stabilisation of HA in split-virion seasonal influenza vaccines by the inventors have demonstrated the need to include a stabiliser system in order to assure adequate HA content over their designated shelf-life. Stabilisation of the haemagglutinin antigen (HA) is believed to be brought about with the formation of complex micelles made up of proteins, lipids, VES, Tween® 80 and Triton X-100®. Stabilisation of HA is achieved through physical separation of HA within the complex micelles. This helps to maintain the HA Content at a reasonably constant level over a period of 12 months or more when stored under refrigerated conditions of 5±3 °C.

Surprisingly, the present inventor has found that it is possible to produce a vaccine that significantly reduces the amount of Tween® 80 present. This was achieved by substituting Tween® 80 in the formulation with Triton X-100® in order to solubilise adequate amount of vitamin E succinate (VES) which led to comparable or better stability profiles for haemagglutinnin antigens over the designated 1 year product shelf-life at refrigerated conditions. Stability data under accelerated (30±2 °C, 4 weeks) and real time (5±3 °C, 12 months) storage conditions were generated for monovalent vaccine bulks and filled syringes and demonstrated improved product stability for reduced stabiliser formulae for monovalent bulk and final QIV filled syringes over formulas using higher amounts of Tween® 80. Maintaining a comparable or better product stability profiles ensures vaccine efficacy and minimises production costs. Furthermore, an overall reduction in the amount of additives present in the vaccines may address any regulatory requirements (e.g. country/region specific) for lower amounts of stabilisers present in influenza vaccines.

Accordingly, the invention provides: - a vaccine composition comprising a split influenza virus preparation or subunit influenza virus preparation and pharmaceutically acceptable excipient, said excipient comprising polyoxyethylene sorbitan monooleate (Tween® 80 or Polysorbate 80), alpha-tocopherol or a derivative thereof and t- octylphenoxypolyethoxyethanol (Triton X-100®), wherein the amount of polyoxyethylene sorbitan monooleate per human dose is <200Mg and the amount of t-octylphenoxypolyethoxyethanol per human dose is > 100Mg.

- a multi-dose or single dose vial comprising the vaccine composition described above.

- a process for producing the above vaccine composition comprising the steps of:

(i) preparing a split influenza virus preparation or subunit influenza preparation,

(ii) optionally inactivating the split influenza virus preparation before or after the splitting step, (iii) purifying and filtering said preparation and (iv) adding a pharmaceutically acceptable excipient to the preparation, said excipient comprising the following stabilisers: polyoxyethylene sorbitan monooleate, alpha- tocopherol and t-octylphenoxypolyethoxyethanol, wherein the concentration of polyoxyethylene sorbitan monooleate is <200 g/ml and the concentration of t- octylphenoxypolyethoxyethanol is from 300 to 700 g/ml.

- the vaccine composition described above for use in medicine

- the vaccine composition described above for use in the treatment and/or prevention of disease caused by influenza.

- a method of inducing an immune response in a human subject, said method comprising administering to the subject the vaccine composition described above.

Brief description of the figures

Figure 1: Monovalent bulk manufacturing process flow chart

Figure 2: Stabilizer addition procedure

Figure 3: Formulation flow chart

Figure 4: Filling flow chart (batch size at least 700 syringes, filling volume: 0.5ml)

Figure 5: Solubilisation of VES in Tween® 80

Figure 6: Solubilisation of VES in Triton-X100®

Figure 7: HA Content (A/Texas H3N2) in MVBs (Mg/ml) at 5±3 °C Figure 8: HA Content (A/Texas H3N2) in MVBs (Mg/ml) at 30±2 °C

Figure 9: HA Content (A/California H1N1-179A) in MVBs (Mg/ml) at 5±3 °C

Figure 10: HA Content (A/California H1N1-179A) in MVBs (Mg/ml) at 5±3 °C

Figure 11: HA Content (A/California H1N1-179A) in MVBs (Mg/ml) at 30±2 °C Figure 12: HA Content (A/California H1N1-179A) in MVBs (Mg/ml) at 30±2 °C Figure 13: HA Content (A/Christchurch HlNl-74xp) in MVBs (Mg/ml) at 5±3 °C Figure 14: HA Content (A/Christchurch HlNl-74xp) in MVBs (Mg/ml) at 30±2 °C Figure 15: HA Content (B/Massachusetts - YMG) in MVBs (Mg/ml) at 5±3 °C

Figure 16: HA Content (B/Massachusetts -YMG) in MVBs (Mg/ml) at 30±2 °C Figure 17: HA Content (B/Brisbane-VIC) in MVBs (Mg/ml) at 5±3 °C

Figure 18: HA Content (B/Brisbane -VIC) in MVBs (Mg/ml) at 30±2 °C

Figure 19: VES Content in MVBs (Mg/ml) at 5±3 °C (A/Texas H3N2)

Figure 20: VES Content for MVBs (Mg/ml) at 30±2 °C (A/Texas H3N2)

Figure 21: VES Content in MVBs (Mg/ml) at 5±3 °C (A/California H1N1-179A) Figure 22: VES Content in MVBs (Mg/ml) at 5±3 °C (A/California H1N1-179A) Figure 23: VES Content in MVBs (Mg/ml) at 30±2 °C (A/California H1N1-179A) Figure 24: VES Content in MVBs (Mg/ml) at 30±2 °C (A/California H1N1-179A) Figure 25: VES Content in MVBs (Mg/ml) at 5±3 °C (A/Christchurch HlNl-74xp) Figure 26: VES Content in MVBs (Mg/ml) at 30±2 °C (A/Christchurch HlNl-74xp) Figure 27: VES Content in MVB (Mg/ml) at 5±3 °C (B/Massachusetts- YMG)

Figure 28: VES Content (Mg/ml) in MVBs at 30±2 °C (B/Massachusetts -YMG) Figure 29: VES Content (Mg/ml) in MVBs at 5±3 °C (B/Brisbane-VIC)

Figure 30: VES Content (Mg/ml) in MVBs at 30±2 °C (B/Brisbane-VIC)

Figure 31: pH Results for QIV Filled Syringes at 5±3 °C

Figure 32: pH Results for QIV Filled Syringes at 5±3 °C

Figure 33: pH Results for QIV Filled Syringes at 30±2 °C

Figure 34: pH Results for QIV Filled Syringes at 30±2 °C

Figure 35: HA Content (A/California H1N1-179A) for QIV Filled Syringes at 5±3 °C Figure 36: HA Content (A/Christchurch HlNl-74xp) for QIV Filled Syringes at 5±3 °C Figure 37: HA Content (A/California H1N1-179A) for QIV Filled Syringes at 30±2 °C Figure 38: HA Content ( A/Christchurch H1N1-74) for QIV Filled Syringes at 30±2 °C Figure 39: HA Content (A/Texas H3N2) for QIV Filled Syringes (5±3 °C)

Figure 40: HA Content (A/Texas H3N2) for QIV Filled Syringes (5±3 °C)

Figure 41: HA Content (A/Texas H3N2) for QIV Filled Syringes (30±2 °C)

Figure 42: HA Content (A/Texas H3N2) for QIV Filled Syringes (30±2 °C)

Figure 43: HA Content (B/Massachusetts B-YMG) for QIV Filled Syringes at 5±3 °C Figure 44: HA Content (B/Massachusetts B-YMG) for QIV Filled Syringes at 5±3 °C Figure 45: HA Content (B/Massachusetts B-YMG) for QIV Filled Syringes (30±2 °C) Figure 46: HA Content (B/Massachusetts B-YMG) for QIV Filled Syringes (30±2 °C) Figure 47: HA Content (B/Brisbane B-VIC) for QIV Filled Syringes at 5±3 °C

Figure 48: HA Content (B/Brisbane B-VIC) for QIV Filled Syringes at 5±3 °C

Figure 49: HA Content (B/Brisbane B-VIC) for QIV Filled Syringes at 30±2 °C

Figure 50: HA Content (B/Brisbane B-VIC) for QIV Filled Syringes at 30±2 °C

Figure 51: VES Content (Mg/ml) for QIV Filled Syringes at 5±3 °C

Figure 52: VES Content (Mg/ml) for QIV Filled Syringes at 5±3 °C

Figure 53: VES Content (Mg/ml) for QIV Filled Syringes at 30±2 °C

Figure 54: VES Content (Mg/ml) for QIV Filled Syringes at 30±2 °C

Detailed description

Provided herein is a vaccine composition comprising or consisting of a split influenza virus preparation or subunit influenza virus preparation and a pharmaceutically acceptable excipient said excipient comprising polyoxyethylene sorbitan monooleate (Tween® 80 or Polysorbate 80), alpha-tocopherol or a derivative thereof and t-octylphenoxypolyethoxyethanol (Triton X-100®), wherein the amount of polyoxyethylene sorbitan monooleate per human dose is <200 pg and the amount of t-octylphenoxypolyethoxyethanol per human dose is >100 M9-

Excipients and stabilisers

Many vaccines that are currently available require a preservative or stabiliser to prevent deterioration during processing and/or in the final vaccine. Such preservatives or stabilisers may be added for example in the final vaccine as part of a pharmaceutically acceptable excipient. As described in WO02/097072, alternatives to the commonly used mercury based stabilisers such as thiomersal have been developed due to public concerns about the effects of mercury containing compounds.

As well as the function of thiomersal as a preservative to prevent microorganism growth, it also had a stabilisation effect on hemagglutinin (HA), particularly for the HA of B strain influenza, but also certain A strain HA such as H3. Stabilisers such as alpha-tocopherol have been successfully used however instead of thiomersal in influenza vaccines (WO02/097072).

The stabiliser systems used in the GlaxoSmithKline (GSK) FluLaval® Quadrivalent and Fluarix® Quadrivalent formulations consist of a combination of Tween® 80 and a-tocopheryl succinate or vitamin E succinate (VES) with or without Triton X-100® in the presence of phosphate buffered saline solution. In these formulations VES together with non-ionic surfactants, Tween® 80 and Triton X-100® are useful to form stabilising micelles (WO02/097072). Stabilisation of the haemagglutinin antigen (HA) is believed to be brought about with the formation of complex micelles made up of proteins, lipids, VES, Tween® 80, Triton X-100®. Stabilisation of HA is achieved through physical separation of HA within the complex micelles. This helps to maintain the HA content at a reasonably constant level over a period of 12 months or more when stored under refrigerated conditions of 5±3 °C. Importantly, stabilisation of HA antigens prevents the bioactive HA antigens from undergoing aggregation resulting in a significant decline in HA activity during storage.

The combination of stabilisers used in the vaccine composition of the present invention comprises Tween® 80, alpha-tocopherol or a derivative thereof and t- octylphenoxypolyethoxyethanol (Triton X-100®).

Polyoxyethylene sorbitan monooleate (Tween 80® or Polysorbate 80)

Despite high concentrations of Tween® 80 being used previously as part of certain stabilisation systems for split influenza vaccines, the present inventors have surprisingly found that as part of the stabilisation system of the invention, a lower concentration of Tween® 80 is sufficient to maintain stability of HA. Such higher concentrations of Tween® 80 were believed to be necessary for Tween® 80 to solubilise alpha-tocopherol succinate (or VES) in order to maintain HA stability in split influenza vaccines. Such higher concentrations of Tween® 80 are not required to maintain HA stability in the vaccine compositions of the present invention.

Thus, in the vaccine composition of the invention, the amount of Tween® 80 per human dose is <200 M9, for example, <150 M9, <100 M9, <90 M9 or <80 M9- In particular, the amount of Tween 80® per human dose (for example, an adult dose) may be from 10 to 200 Mg, from 10 to 150 Mg, from 20 to 100 M9, from 50 to 150 M9, from 50 to 100 ^g, or from 50 to 80 ^g, such as around 70 M9- Preferably, for an adult dose the amount per dose is from 20 to 100 M9-

Smaller doses are contemplated, for example for some paediatric human doses. Some paediatric doses may contain half the human adult dose amount of Tween® 80 as well as other vaccine ingredients such as other excipient or stabiliser components. In particular, the amount of Tween® 80 per paediatric human dose may be from 10 to 60 ^g, from 10 to 50 M9, or from 20 to 45 M9, such as around 35 Mg. Preferably, the paediatric human dose is from 10 to 50 M9- t-octylphenoxypolyethoxyethanol

Synonyms for t-octylphenoxypolyethoxyethanol (Triton X-100®) include but are not limited to polyethylene glycol p-(l,l,3,3-tetramethylbutyl)-phenyl ether, octyl phenol ethoxylate, polyoxyethylene octyl phenyl ether, 4-octyl phenol polyethoxylate, Mono 30, TX-100, Octoxynol-9, octoxynol 10, X-100, and octylphenol ethylene oxide condensate.

In the vaccine composition of the invention, the amount of Triton X-100® per human dose is >100 ^g, for example, >150 M9 or >200 M9- In particular, the amount of Triton X-100® per human dose (for example, an adult dose) may be from 100 to 500 μς, from 100 to 400 g, from 100 to 300 g, from 110 to 500 g, from 120 to 300 M9, from 190 to 500 M9, or from 190 to 300 M9, such as around 250 M9- Preferably, the adult human dose amount is from 120 to 300 M9-

Smaller doses are contemplated, for example for some paediatric human doses. Some paediatric doses may contain half the human adult dose amount of Triton X-100® as well as other vaccine ingredients such as other excipient or stabiliser components. In particular, the amount of Triton X-100® per paediatric human dose may be from 100 to 175 μς, from 100 to 150 μς, from 100 to 140 μς, or from 115 to 135 μς, such as around 125 μς. Preferably, the paediatric dose amount is from 100 to 140 μς.

In one embodiment, the vaccine composition of the invention contains an amount of Tween® 80 of from 10 to 200 μς and an amount of Triton X-100® of from 100 to 500 μς per human dose. In another embodiment, the amount of Tween 80® is from 50 to 150 μς and the amount of Triton X-100® is from 190 to 500 μς per human dose.

In a further embodiment, the vaccine composition of the invention contains an amount of Tween® 80 of from 10 to 60 μς and an amount of Triton X-100® of from 100 to 175 μς per paediatric human dose. In another embodiment, the amount of Tween® 80 is from 50 to 150μς and the amount of Triton X-100® is from 190 to 500 μς per paediatric human dose.

Alpha tocopherol

Alpha tocopherol (a-tocopherol) and derivatives of a-tocopherol such as a- tocopherol succinate are included in the vaccine composition of the invention. Other preferred tocopherol derivatives for use in the invention include D-a tocopherol, D-δ tocopherol, D-γ tocopherol and DL-a-tocopherol. Preferred derivatives of tocopherols that may be used include acetates, succinates, phosphoric acid esters, formiates, propionates, butyrates, sulfates and gluconates. Alpha-tocopherol succinate is particularly preferred.

The α-tocopherol or derivative is present in an amount sufficient to stabilise the haemagglutinin (for example using the SRD method or any other method described herein for stability analysis). See for example WO2002/097072. Preferred concentrations for the α-tocopherol or derivative are between 1 μg/ml - 10 mg/ml, more preferably between 10 μg/ml - 500 μg/ml.

Pharmaceutically acceptable excipient

The influenza virus preparation in the vaccine composition of the invention is present along with a pharmaceutically acceptable excipient. This excipient comprises Tween® 80, alpha-tocopherol or a derivative thereof and Triton X-100®. Typically, these and optionally any other standard excipient ingredients will be present in liquid form. For example, the stabiliser components, antigen and any adjuvant present may be diluted in a volume of aqueous buffer, mineral oil, water for injection or other suitable diluent well known to those skilled in the art, together with any additional excipient components, such as salts that are commonly used in such diluents in a vaccine context. Thus, in one aspect there is provided a vaccine composition of the invention in which each specified component is diluted in an aqueous buffer or water for injection.

The vaccine composition of the invention comprises Tween® 80 and Triton X- 100® in specified human doses. By the term "human dose" is meant the amount of the specified ingredient in the vaccine composition which is delivered along with any other vaccine composition ingredients (other excipient components, antigen and any adjuvant) to a human subject, for example in a volume suitable for human use. Generally the dose volume is from 0.25 to 1.5 ml. In one embodiment, a human dose is about 0.5 ml. In a further embodiment, a human dose is higher than 0.5 ml, for example about 0.6, 0.7, 0.8, 0.9 or about 1 ml. In a further embodiment, a human dose is from 1 ml to 1.5 ml. In another embodiment, in particular when the vaccine composition is for the paediatric population, a human dose may be less than 0.6 ml, for example from 0.25 to 0.5 ml or exactly 0.1 ml, 0.2 ml, 0.25 ml, 0.3 ml or 0.4 ml. Preferably, the dose volume for an adult is from 0.4 to 1.2 ml and the dose volume for the paediatric population is from 0.2 to 0.6 ml. Slight adaptation of the dose volume will be made routinely depending on the HA concentration in the original bulk sample and depending also on the delivery route with smaller doses being given by the intranasal or intradermal route.

Typically, the stabiliser components, antigen and any adjuvant present will be diluted in a volume of aqueous buffer, water for injection or other suitable diluent well known to those skilled in the art, together with any additional excipient components, such as salts that are commonly used in such diluents in a vaccine context. Thus, in one aspect there is provided a vaccine composition of the invention in which each specified component is diluted in an aqueous buffer or water for injection. Influenza virus preparation

Preferably the influenza virus preparation of the invention is either a split virus preparation, or a subunit antigen prepared from whole virus (e.g. particularly by a splitting process followed by purification of the surface antigen) or by expression of recombinant protein. Most preferred are split virus preparations.

The split or subunit influenza virus preparation according to the invention may be derived from the conventional embryonated egg method, or they may be derived from any of the new generation methods using tissue culture to grow the virus or express recombinant influenza virus surface antigens. Suitable cell substrates for growing the virus include for example dog kidney cells such as MDCK or cells from a clone of MDCK, MDCK-like cells, monkey kidney cells such as AGMK cells including Vero cells, suitable pig cell lines, or any other mammalian cell type suitable for the production of influenza virus for vaccine purposes. Suitable cell substrates also include human cells e.g. MRC-5 cells. Suitable cell substrates are not limited to cell lines; for example primary cells such as chicken embryo fibroblasts are also included.

Techniques for splitting and/or preparation of subunit antigens are well known in the art. Traditionally split flu was produced using a solvent/detergent treatment, such as tri- 7-butyl phosphate, or diethylether in combination with Tween® (known as "Tween-ether" splitting) and this process is still used in some production facilities. Other splitting agents now employed include detergents or proteolytic enzymes or bile salts, for example sodium deoxycholate. Detergents that can be used as splitting agents include cationic detergents e.g. cetyl trimethyl ammonium bromide (CTAB), other ionic detergents e.g. laurylsulfate, taurodeoxycholate, or non-ionic detergents such as the ones described above including Triton X-100® (for example in a process described in Una et. al., 2000, Biologicals 28, 95-103) and Triton N-101®, or combinations of any two or more detergents.

The preparation process for a split vaccine will include a number of different filtration and/or other separation steps such as ultracentrifugation, ultrafiltration, zonal centrifugation and chromatography (e.g. ion exchange) steps in a variety of combinations, and optionally an inactivation step e.g. with heat, formaldehyde or β- propiolactone or U.V. which may be carried out before or after splitting. The splitting process may be carried out as a batch, continuous or semi-continuous process. After splitting, inactivation and filtration steps, the influenza virus preparation may be purified before using in the vaccine composition of the invention. For example, the split preparation may be prepared using the techniques described in WO2011/051235 or WO2002/097072.

The subunit influenza virus antigen preparation according to the invention may be derived from a source other than the live influenza virus, for example the haemagglutinin antigen may be produced recombinantly using techniques well known to those skilled in the art.

In one embodiment, the vaccine composition is monovalent, i.e. only comprises one specific type of influenza HA. For example, the vaccine composition may comprise a influenza HA associated with a pandemic or a strain noted and monitored by the World Health Organisation as a potentially pandemic strain of influenza (e.g. a H5, H7, H9, H8 or H10 type influenza strain, preferably an H5 strain). In alternative embodiments, the composition is multivalent, i.e. comprises multiple influenza virus antigens. For example, the composition may be bivalent, trivalent or quadrivalent, e.g. may contain two or three seasonal strains of influenza virus. Preferably, the vaccine composition is a trivalent or quadrivalent vaccine comprising seasonal strains of influenza virus (e.g. as recommended by the World Health Organisation each year for inclusion in a seasonal influenza vaccine). Such strains may include two A strains (e.g. HlNl and H2N3) and one or two strains of B influenza virus (e.g. from Victoria and/or Yamagata lineages).

A pharmaceutically acceptable excipient as described herein may be added to the influenza virus preparation to make up the vaccine composition of the invention. The influenza virus preparation along with the pharmaceutically acceptable excipient may be stored in bulk before transfer to vials ready for administration (e.g. in single- dose prefilled syringes or in single or multi-dose vials).

An adjuvant may be added before or after storage. For example, the adjuvant may be added to a vaccine composition of the invention just prior to administration by mixing the contents of adjuvant and vaccine composition in separate vials.

Thus, in one embodiment, the vaccine composition of the invention further comprises an adjuvant. Preferably, the adjuvant is an oil-in-water emulsion adjuvant. Oil in water emulsion adjuvants, such as MF59 or AS03 are well known in the art and are described below.

Adjuvant

Thus, in one embodiment, the vaccine composition of the invention further comprises an adjuvant. Preferably, the adjuvant is an oil-in-water emulsion adjuvant. Oil in water emulsion adjuvants, such as MF59 or AS03 are well known in the art and are described below.

Tocols (e.g. Vitamin E) are also used in oil emulsions adjuvants (EP0382271B1; US5667784; WO95/17210). In an oil-in-water emulsion, the oil and emulsifier should be in an aqueous carrier. The aqueous carrier may be, for example, phosphate buffered saline or a citrate buffer. One example of a tocol-containing oil- in-water emulsion is AS03.

A preferred oil-in-water emulsion comprises a metabolisable oil, such as squalene, Tween® 80 and optionally alpha tocopherol. Additionally, the oil-in-water emulsion may contain Span® 85 and/or lecithin.

In one instance, the oil-in-water emulsion has one of the following compositions:

- From 0.5 to 11 mg squalene, from 0.05 to 5% polyoxythylene sorbitan monooleate (Tween 80® or Polysorbate 80) and optionally, from 2 to 12% alpha-tocopherol; or

- About 5% squalene, about 0,5% polyoxyethylene sorbitan monooleate (Tween® 80 or Polysorbate 80) and about 0.5% sorbitan trioleate (Span® 85). This adjuvant is called MF59.

An alternative adjuvant that may be used with the vaccine composition according to the present invention, comprises an immunologically active saponin fraction derived from the bark of Quillaja Saponaria Molina (e.g. QS21) presented in the form of a liposome and a lipopolysaccharide (e.g. 3D-MPL), optionally further including a sterol (cholesterol). In one embodiment, the adjuvant comprises or consists of a saponin (e.g. QS21) presented in the form of a liposome, a lipid A derivative such as 3D-MPL and a sterol (e.g. cholesterol). The liposomes suitably contain a neutral lipid, for example, phosphatidylcholine, dioleoyi phosphatidylcholine (DOPC) or dilauryl phosphatidylcholine. The liposomes may also contain a charged lipid which increases the stability of the liposome-QS21 structure for liposomes composed of saturated lipids. An example of such an adjuvant is AS01, which comprises 3D-MPL and QS21 in a quenched form with cholesterol, and can be made as described in W096/33739. Either the AS01B or AS01E forms of this adjuvant may be used. The AS01 B adjuvant comprises liposomes, which in turn comprise dioleoyl phosphatidylcholine (DOPC), cholesterol and 3D-MPL (in an amount of approximately 1000 micrograms DOPC, 250 micrograms cholesterol and 50 micrograms 3D-MPL per vaccine dose), QS21 (50 micrograms/dose), phosphate saline buffer and water to a volume of 0.5 ml.

The AS01E adjuvant comprises the same ingredients than AS01 B but at a lower concentration in an amount of approximately 500 micrograms DOPC, 125 micrograms cholesterol, 25 micrograms 3D-MPL and 25 micrograms QS21, phosphate saline buffer and water to a volume of 0.5 ml.

Stabilisation

Stability of hemagglutinin can be measured by a variety of techniques and would be well known to those skilled in the art. For example, stability can be measured by the appearance of the antigen, such as the appearance of precipitation, or floccules that sediment out of solution. Appearance can be monitored over time under normal storage conditions (e.g. refrigerated at 0±3, 5±3, 10±2, 15±2 etc. °C), or under accelerated stability testing conditions such as room temperature or higher temperatures (e.g. 20±2, 25±2, 30±2, 35±2, 37±2, 37.5±2, 38±2, 40±2, 45±2, 50±2 °C). Samples may be stored for example in a constant temperature chamber or room. The appearance can be measured using techniques well known in the art, for example using a spectrophotometer to obtain a transmission reading, or upon visual inspection.

Another way of measuring stability is to measure the HA content over time either under normal storage conditions for a vaccine (e.g. refrigerated conditions) or accelerated conditions such as room temperature or higher temperatures (e.g. 20±2, 25±2, 30±2, 35±2, 37±2, 37.5±2, 38±2, 40±2, 45±2, 50±2 °C). Preferably, the accelerated conditions are from 20 to 40 °C. Techniques for measuring HA content, for example the commonly used SRD assay, would be well known to those skilled in the art in this field.

Stability may be measured over a period of time e.g. 1, 2, 3 or 4 weeks or 1, 2, 3, 4, 5, 6, or 12 months or more. For example, under accelerated conditions, the time period may be 1, 2, 3 or 4 weeks or more. Under normal storage conditions such as refrigerated conditions, the time period may be 6 or 12 months, or from about 6 months to about 12 months. Samples are withdrawn periodically at designated time points throughout the stability testing in accordance with established methods for stability testing in this way.

Thus, in one instance, stability is assessed after storage of the vaccine composition at from 20 to 40 degrees centigrade (e.g. 30±2 °C) for 4 weeks or more. In another instance, stability is assessed after storage of the vaccine composition at from 2 to 10 degrees centigrade (e.g. 5±3 °C) for about 12 months.

Further dosing and efficacy criteria Typically, the vaccine composition according to the present invention is a standard 0.5 ml injectable dose in most cases, and contains 15 Mg or less, of hemagglutinin antigen component from an/each influenza virus strain, as measured by single radial immunodiffusion (SRD) (J.M. Wood et al. : J. Biol. Stand. 5 (1977) 237-247; J. M. Wood et al., J. Biol. Stand. 9 (1981) 317-330).

In one embodiment, the vaccine composition according to the invention contains a low amount of HA antigen - e.g. any of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 g of HA per influenza virus strain (e.g. less than 10 Mg of HA, or less than 7.5 g of HA) or which does not exceed 15 Mg of HA per strain. Said low amount of HA amount may be as low as practically feasible provided that it allows to formulate a vaccine which meets the international e.g. EU or FDA criteria for efficacy, as detailed below (see Table 1 and the specific parameters as set forth). A suitable low amount of HA is from 1 to 7.5 Mg of HA per influenza virus strain, suitably from 3.5 to 5 Mg, such as 3.75 or 3.8 Mg of HA per influenza virus strain, typically about 5 Mg of HA per influenza virus strain. Another suitable amount of HA is from 0.1 to 5 Mg of HA per influenza virus strain, suitably from 1.0 to 2 of HA per influenza virus strain, such as 1.9 Mg of HA per influenza virus strain.

The vaccine composition of the invention suitably meets certain international criteria for vaccines. Standards are applied internationally to measure the efficacy of influenza vaccines.

Serological variables are assessed according to criteria of the European Agency for the Evaluation of Medicinal Products for human use (CHMP/BWP/214/96, Committee for Proprietary Medicinal Products (CPMP). Note for harmonization of requirements for influenza vaccines, 1997. CHMP/BWP/214/96 circular N°96-0666: l- 22) for clinical trials related to annual licensing procedures of influenza vaccines (Table below).

Table 1: CHMP criteria

Conversion factor** >2.5 >2.0

Protection rate*** >70% >60%

* Seroconversion rate is defined as the proportion of subjects in each group having a protective post-vaccination titre≥ 1:40. The seroconversion rate simply put is the % of subjects who have an HI titre before vaccination of <1:10 and≥1:40 after vaccination. However, if the initial titre is≥1:10 then there needs to be at least a fourfold increase in the amount of antibody after vaccination.

** Conversion factor is defined as the fold increase in serum HI geometric mean titres (GMTs) after vaccination, for each vaccine strain.

*** Protection rate is defined as the proportion of subjects who were either seronegative prior to vaccination and have a (protective) post-vaccination HI titre of ≥ 1:40 or who were seropositive prior to vaccination and have a significant 4- fold increase in titre post-vaccination; it is normally accepted as indicating protection.

The requirements are different for adult populations (18-60 years) and elderly populations (>60 years). For inter-pandemic influenza vaccines, at least one of the assessments (seroconversion factor, seroconversion rate, seroprotection rate) should meet the European requirements, for all strains of influenza included in the vaccine. The proportion of titres equal or greater than 1:40 is regarded most relevant because these titres are expected to be the best correlate of protection (Beyer et al. (1998) Clin Drug Invest 15: 1).

The compositions for use according to the present invention suitably meet at least one such criteria for the influenza virus strain included in the composition (one criteria is enough to obtain approval), suitably at least two, or typically at least all three criteria for protection. Suitably the above response(s) is (are) obtained after one dose, or after two doses.

Methods of treatment A method of inducing an immune response in a human subject, said method comprising administering to the subject a vaccine composition of any one of claims 1 to 6.

In a further aspect, the vaccine composition of the invention is for use in medicine, such as for use in the treatment, prevention of, or vaccination against, disease caused by influenza e.g. administered to a person (e.g. subject) at risk for influenza infection.

In one embodiment, the vaccine composition of the invention is for use in the prevention of or vaccination against, influenza disease caused by a different clade than the clade on which the influenza virus preparation was based. For example, a H5N1 clade 1 HA antigen could be used for protection against influenza caused by a non-clade 1 virus e.g. a H5N1 clade 2 virus.

In a further aspect, there is provided a method of treatment, prevention and/or vaccination against influenza disease, comprising the administration of a vaccine composition as described herein to a person in need thereof, e.g. to a person (e.g. subject) at risk for influenza infection, e.g. an elderly person (age 50 or over, particularly age 65 or over).

Routes of administration

The composition of the invention may be administered by any suitable delivery route, such as intradermal, mucosal (e.g. intranasal), oral, intramuscular or subcutaneous. Other delivery routes are well known in the art.

The intramuscular (IM) delivery route is particularly suitable for the vaccine composition. The composition according to the invention may be presented in a monodose container, or alternatively, a multidose container, particularly suitable for a pandemic vaccine. In this instance an antimicrobial preservative such a thiomersal may be present to prevent contamination during use. A thiomersal concentration of 5 Mg/0.5 ml dose (i.e. 10 Mg/ml) or 10 Mg/0.5 ml dose (i.e. 20 Mg/ml) is suitably present. A suitable IM delivery device could be used such as a needle-free liquid jet injection device, for example the Biojector 2000 (Bioject, Portland, OR). Alternatively a pen-injector device, such as is used for at-home delivery of epinephrine, could be used to allow self-administration of vaccine. The use of such delivery devices may be particularly amenable to large scale immunization campaigns such as would be required during a pandemic.

Intradermal delivery is another suitable route. Any suitable device may be used for intradermal delivery, for example short needle devices. Such devices are well known in the art. Intradermal vaccines may also be administered by devices which limit the effective penetration length of a needle into the skin, such as those described in WO99/34850 and EP1092444, incorporated herein by reference, and functional equivalents thereof. Also suitable are jet injection devices which deliver liquid vaccines to the dermis via a liquid jet injector or via a needle which pierces the stratum corneum and produces a jet which reaches the dermis. Also suitable, are ballistic powder/particle delivery devices which use compressed gas to accelerate vaccine in powder form through the outer layers of the skin to the dermis. Additionally, conventional syringes may be used in the classical mantoux method of intradermal administration.

Another suitable administration route is the subcutaneous route. Any suitable device may be used for subcutaneous delivery, for example classical needle. Suitably, a needle-free jet injector service is used. Such devices are well known in the art. Suitably said device is pre-filled with the liquid vaccine formulation.

Alternatively the vaccine is administered intranasally. Typically, the vaccine is administered locally to the nasopharyngeal area, suitably without being inhaled into the lungs. It is desirable to use an intranasal delivery device which delivers the vaccine formulation to the nasopharyngeal area, without or substantially without it entering the lungs.

Suitable devices for intranasal administration of the vaccines according to the invention are spray devices. Suitable commercially available nasal spray devices include Accuspray® (Becton Dickinson). Nebulisers produce a very fine spray which can be easily inhaled into the lungs and therefore does not efficiently reach the nasal mucosa. Nebulisers are therefore not preferred.

Suitable spray devices for intranasal use are devices for which the performance of the device is not dependent upon the pressure applied by the user. These devices are known as pressure threshold devices. Liquid is released from the nozzle only when a threshold pressure is applied. These devices make it easier to achieve a spray with a regular droplet size. Pressure threshold devices suitable for use with the present invention are known in the art and are described for example in WO 91/13281 and EP 311 863 B and EP 516 636, incorporated herein by reference. Such devices are commercially available from Pfeiffer GmbH and are also described in Bommer, R. Pharmaceutical Technology Europe, Sept 1999.

Alternatively, the epidermal or transdermal vaccination route is also contemplated in the present invention.

Thus in a further aspect, there is provided a device as described above comprising or containing a vaccine composition of the invention.

The teaching of all references in the present application, including patent applications and granted patents, are herein fully incorporated by reference. Any patent application to which this application claims priority is incorporated by reference herein in its entirety in the manner described herein for publications and references. For the avoidance of doubt the terms 'comprising', 'comprise' and 'comprises' herein is intended by the inventors to be optionally substitutable with the terms 'consisting of, 'consist of, and 'consists of, respectively, in every instance.

EXAMPLES

Example 1:

The current study is aimed at:

• determining the extent of VES solubilisation by Tween® 80 and Triton X- 100® individually and in combination,

• optimising the combinations of VES/surfactants at laboratory-scale and realistic production-scale process operability trials,

• adapting the established influenza manufacturing processes for monovalent bulks, final quadrivalent bulks and syringe filling operations,

• verifying the stability profiles of monovalent and QIV filled syringes in reduces stabiliser formulations under accelerated (at 30±2°C for 4 weeks) and real time conditions (at 5±3°C for 12 months). A set of preliminary formulae (Table 2) containing Tween® 80, Triton X-100® and VES is proposed.

Table 2: Proposed Formula for Monovalent Bulk and QIV Final Bulk and Filled Syringe

* q.s. = sufficient quantities

Materials and Methods

Materials Used for Pre-Formulation Studies

The following materials are used in the pre-formulation study.

• a-tocopheryl succinate or vitamin E succinate, VES (B. 2012120574, BASF)

• Tween® 80 or Polysorbate 80 (B. 2013060415, Well Chemicals)

• Triton X-100® (B. L012017843320, Merck)

• Bovine serum albumin (B.0175, Amresco)

• Phosphate buffered saline solution, PBS (B.13126B, B.13142B, GSK Shenzhen)

• Ethanol 95% (B. 20130501, Guangzhou Chemicals)

• Stock solutions (2.654% VES in 95% ethanol, 1% Tween® 80, 2.5% Triton X-100® in PBS, GSK Shenzhen)

• Water for injection (GSK Shenzhen)

Materials Used for Manufacturing of Monovalent Bulks

The following materials are needed to manufacture monovalent bulks with various stabiliser concentrations.

In order to formulate a batch of quadrivalent Split-Virion Seasonal Influenza Vaccines (QIV), 4 individual lots of monovalent bulks, one bulk from each virus strain, are needed.

Five strains of working seeds (Table 3) are used for the manufacturing of monovalent bulks batches including two H1N1 strains (A/California/7/2009 NYMC X- 179A, H1N1-179A and A/Christchurch/16/2010 NIB-74xp, HlNl-74xp), a H3N2 strain (A/Texas/50/2012 NYMC X-223A, H3N2) and two B strains with one from the Yamagata lineage (B/Massachusetts/2/2012 NYMC BX-51) and the other from the Victoria lineage (B/Brisbane/60/2008 NYMC BX-35). Monovalent bulk lots used to manufacture the monovalent bulks in different formulae A-G are given in Table 4.

Table 3: Details of Reference and Working Seeds

Table 4: Monovalent Bulk Batch used for the Manufacture of Monovalent Bulks in Various Formulae

A total of 6 lots of 0-day eggs are sourced from Da Hua Nong Farm, Guangdong Province, P.R. China. Each lot consists of 70,000 eggs.

Triton X-100® (B. Iol2017843320) is supplied by Merck.

VES (B.2013050344) is sourced from Badische Anilin-und-Soda-Fabrik Tween® 80 (B.2013060415) is purchased from Well Chemical, Nanjing. Other items include 2 L glass bottles supplied by Fisher (Item code: FB800-2000), 500 ml glass bottles supplied by Kimble Chase (Item code: 89000-930).

Other ancillary materials needed during production are recorded in detail (Laboratory Books - Primary 02, 08, 09, 10 and 11 and Batch Manufacturing Records for batches 2013FS01, 2013FS02, 2013FS04, 2014FS01 and 2014FS02). Manufacturing of Final Bulk and Filled Syringes

In order to formulate a batch of quadrivalent (QIV) Split Virion Seasonal Influenza Vaccines, one monovalent bulk from each of the 4 virus strains, is needed. The following materials are used in the manufacturing final bulks and finished products with various stabiliser concentrations (Tables 5 - 7). Other sterile materials used during production include 2 L glass bottles, flexible silicon tubing, magnetic stirrer beads, stainless-steel scissors and stainless-steel forceps. These are recorded in manufacturing records (Laboratory Book 07).

Table 5: Monovalents Bulks (MVB) Used for Manufacturing of Final bulks

Table 6: Buffer Solutions Used for Manufacturing of Final bulks

Table 7: Disposable Sterile Materials Used for Manufacturing of Final bulks and Filled Syringes

Example 2: Preliminary studies on VES solution

Dilution of VES Stock Solution with PBS

In this study, a PBS solution is gradually added into a 250 ml conical flask containing 30 ml solution of VES Stock Solution containing 2.654% VES solution in 95% ethanol with constant agitation. The solution is observed while the PBS solution is added and allowed to stand for several hours. Dilution of VES Stock Solution with 80 μς/ml Tween® 80 in PBS

This study is conducted by gradual addition of a solution of Tween® 80 with a concentration of 80 g/ml in PBS into a 250 ml conical flask containing 30 ml solution of 2.654% VES Stock Solution (in 95% ethanol) with constant agitation. The mixture is observed while Tween® 80 solution is added. The final solution is allowed to stand for several hours.

Results and Discussion

Dilution of VES Stock Solution with PBS

The solution develops an opaque white appearance after addition of 9 ml of PBS. The amount of precipitate present increases after further addition of PBS. Aggregates are observed after the solution is left standing for a few hours.

Dilution of VES Stock Solution with 80 μς/ml Tween® 80 in PBS

The solution develops an opaque white appearance after addition of 7.3 ml of 80 g/ml Tween® 80 in PBS. The amount of precipitates increased after a total of addition of 8.0 ml of PBS. When the solution is left to stand for a few hours, the insoluble precipitates remain and some greasy materials are seen at the bottom of the flask.

Example 3:

Effects of Concentrations of Tween® 80 and VES on % Transmittance

A combination of specified quantities of VES Stock Solution containing 2.654% VES solution in 95% ethanol, 1% Tween® 80 solution in PBS and PBS solution, with a combined total volume of 10ml, is mixed within a 14 ml BD tube with agitation. The concentration for VES in this study is between 50 - 200 g/ml while the Tween® 80 concentration is studied at between 30 - 150 g/ml. Details of the study are given in Table 24. The transmittance of the solutions is measured at 580 nm against a blank PBS solution. Results and Discussion

Effects of Concentrations of Tween® 80 and VES on % Transmittance

The % transmittance for the various solutions at 580 nm are recorded in Table 21. The %T results recorded range from 57.19% - 95.61%. The %T for the QIV product is 86.45%. Solutions with VES:Tween® 80 at 100:60, 100:90, 150:90, 150: 120, 150:60, 200:30, 200:60, 200:90, 200: 120, 200: 150. A non-linear relationship exists between the ratios of the VES/Tween® 80 but this is influenced by both the concentrations of the VES and Tween® 80.

Table 21: Transmittance of the Mixtures of VES and Tween® 80

15 0.060 0.150 9.790 150: 150 1.000 86.10

16 0.080 0.030 9.890 200:30 6.667 78.75

17 0.080 0.060 9.860 200:60 3.333 76.53

18 0.080 0.090 9.830 200:90 2.222 57.19

19 0.080 0.120 9.800 200: 120 1.667 80.95

20 0.080 0.150 9.770 200: 150 1.333 80.51

QIV N/A N/A N/A 533: 1360 0.392 86.45

B.20130907

Example 4: Solubilisation of VES in Tween® 80 Solution

In this study, specified quantities of VES Stock Solution are added to diluted Tween® 80 solutions with constant agitation. The concentration for VES in this study is increased to 50 - 400 Mg/ml while the Tween® 80 concentration is studied at between 90 - 150 g/ml. Details of the study are given in Table 25. The appearance of the mixtures is recorded together with their transmittance, measured at 580 nm against a blank PBS solution.

Results and Discussion

Solubilisation of VES in Tween® 80 Solution

The %T for different combinations of Tween® 80 and VES solution are given in Table 22. At Tween® 80 concentration of 90 Mg/ml, the concentration of VES has to be below 150 g/ml in order to achieve a solution clarity equivalent or better than the QIV product whereas for Tween® 80 concentration of 150 Mg/ml, the concentration of VES has to be below 200 Mg/ml in order to achieve a solution clarity equivalent or better than the QIV product. Table 22: Solubilisation of VES in Tween® 80 Solution

Example 5:_Solubilisation of VES in Triton X-100® Solution

In this study, specified quantities of VES Stock Solution are added to a diluted Triton X-100® Solution with constant agitation. Details of the study are given in Table 26. The appearance of the mixture is recorded together with their transmittance (%T) measured at 580 nm against a blank PBS solution.

Results and Discussion

Solubilisation of VES in Triton X-100® Solution

The %T for different combinations of Triton X-100® and VES solution are given in Table 23. At increasing concentration of Triton X-100® from 100 to 600 Mg/ml, there is a corresponding increasing in the concentration of VES solution that can be solubilised achieving a %T result of at least the same or greater than that of the QIV reference sample. At a Triton X-100® concentration of 500 Mg/ml, a VES concentration less than 400 g/ml will result in a %T close to that of the reference sample.

Table 23: Solubilisation of VES in Triton X-100® Solution

0.180 0.040 9.880 450: 100 4.500-10 0.200 0.040 9.760 500: 100 5.000-11 0.220 0.040 9.740 550: 100 5.500-12 0.240 0.040 9.720 600: 100 6.000-1 0.020 0.080 9.900 50:200 0.250-2 0.040 0.080 9.880 100:200 0.500-3 0.060 0.080 9.860 150:200 0.750-4 0.080 0.080 9.840 200:200 1.000-5 0.100 0.080 9.820 250:200 1.250-6 0.120 0.080 9.800 300:200 1.500-7 0.140 0.080 9.780 350:200 1.750-8 0.160 0.080 9.760 400:200 2.000-9 0.180 0.080 9.740 450:200 2.250-10 0.200 0.080 9.720 500:200 2.500-11 0.220 0.080 9.700 550:200 2.750-12 0.240 0.080 9.680 600:200 3.000-1 0.020 0.120 9.860 50:300 0.167-2 0.040 0.120 9.840 100:300 0.333-3 0.060 0.120 9.820 150:300 0.500-4 0.080 0.120 9.800 200:300 0.667-5 0.100 0.120 9.780 250:300 0.833-6 0.120 0.120 9.760 300:300 1.000-7 0.140 0.120 9.740 350:300 1.167-8 0.160 0.120 9.720 400:300 1.333

0.180 0.120 9.700 450:300 1.500-10 0.200 0.120 9.680 500:300 1.667-11 0.220 0.120 9.660 550:300 1.833-12 0.240 0.120 9.640 600:300 2.000-1 0.020 0.160 9.820 50:400 0.125-2 0.040 0.160 9.800 100:400 0.250-3 0.060 0.160 9.780 150:400 0.375-4 0.080 0.160 9.760 200:400 0.500-5 0.100 0.160 9.740 250:400 0.625-6 0.120 0.160 9.720 300:400 0.750-7 0.140 0.160 9.700 350:400 0.875-8 0.160 0.160 9.680 400:400 1.000-9 0.180 0.160 9.660 450:400 1.125-10 0.200 0.160 9.640 500:400 1.250-11 0.220 0.160 9.620 550:400 1.375-12 0.240 0.160 9.600 600:400 1.500-1 0.020 0.200 9.880 50:500 0.100-2 0.040 0.200 9.760 100:500 0.200-3 0.060 0.200 9.740 150:500 0.300-4 0.080 0.200 9.720 200:500 0.400-5 0.100 0.200 9.700 250:500 0.500-6 0.120 0.200 9.680 300:500 0.600-7 0.140 0.200 9.660 350:500 0.700-8 0.160 0.200 9.640 400:500 0.800

0.180 0.200 9.620 450: 500 0.900

5-10 0.200 0.200 9.600 500: 500 1.000

5-11 0.220 0.200 9.580 550: 500 1.100

5-12 0.240 0.200 9.560 600: 500 1.200

6-1 0.020 0.240 9.740 50:600 0.083

6-2 0.040 0.240 9.720 100:600 0.167

6-3 0.060 0.240 9.700 150:600 0.250

6-4 0.080 0.240 9.680 200:600 0.333

6-5 0.100 0.240 9.660 250:600 0.417

6-6 0.120 0.240 9.640 300:600 0.500

6-7 0.140 0.240 9.620 350:600 0.583

6-8 0.160 0.240 9.600 400:600 0.667

6-9 0.180 0.240 9.580 450:600 0.750

6-10 0.200 0.240 9.560 500:600 0.833

6-11 0.220 0.240 9.540 550:600 0.917

6-12 0.240 0.240 9.520 600:600 1.000

QIV N/A N/A N/A 533: 1365 0.390

B.20130915

Example 6: Soiubiiisation of VES in a Mixture of Tween® 80 and Triton X- 100® Solutions (MVB)

In this study, specified volumes of VES Stock Solution are added to a diluted Tween® 80 Solution (with a concentration of 140 Mg/ml) and Triton X-100® Solution (with a concentration of 500 Mg/ml), adjusted to a final volume of 10.0 ml with constant agitation. Details of the study are given in Table 27. Sample M3' is based on a Triton X-100®:Tween® 80:VES ratio of 500: 140:500. This test is performed to assess this combination of stabilisers on the appearance of the solution. The appearance of the mixture is recorded together with their transmittance (%T) measured at 580 nm against a blank PBS solution. Measurement of %T is repeated after 2 days.

Results and Discussion

Soiubiiisation of VES in a mixture of Tween® 80 and Triton X-100® Solutions (Mixing Order 1)

The results of preliminary studies on VES soiubiiisation of VES in Tween® 80 and Triton X-100® are given in Table 24. Samples Ml, Fl and F2 have higher transmittance than that of the comparator while the %T for M2 and F3 are similar. The %T remains essentially unchanged with storage after 2 days. In the presence of 500 Mg/ml of Triton X-100® and 140

Tween® 80, a concentration of VES at 530 MQ/ml will have a %T of < that of the reference. Therefore, the VES concentration has to be set between 440 and 530 Mg/ml. A repeat of M3 (Μ3') using a lower amount of VES of 500 Mg/ml gives rise to a %T of > 80%.

Table 24: Solubilisation of VES in a Mixture of Tween® 80 and Triton X-100(R) Solutions CMVB)

Example 7: Solubilisation of VES in a Mixture of Tween® 80 and Triton X- 100® (QIV Formulation)

In this study, specified volumes of VES Stock Solution are added to a mixture of Tween® 80 Solution (70 Mg/ml) and Triton X-100® Solution (250 Mg/ml) at various concentration ratios, with constant agitation and then adjusted to a final volume of 10.0 ml. Details of the study are given in Table 28. The appearance of the mixture is recorded together with their transmittance (%T) measured at 580 nm against a blank PBS solution.

Results and Discussion

Solubilisation of VES in a Mixture of Tween 80 and Triton X-100 (QIV Formulation)

The results of this solubilisation study (Table 25) showed that all mixtures are above that of the product comparator except for F13 and F14 which are cloudy. Samples F12 are slightly lower %T than the product comparators and has similar appearance.

Table 25: Solubilisation of VES in Tween® 80 (140 uo/m/)/Triton X-100(R) (500 μα/ml) Mixing Order 2.

Results of Solubilisation of VES with Tween® 80 and Triton X-100®

Pre-formulation studies established a non-linear relationship between the concentration and solution clarity as measured with % Transmittance (%T). Transmittance of mixtures of VES with surfactants is influenced by both the VES/Surfactant ratios and the concentrations of the surfactant. A certain amount of the stabiliser is needed to stabilise the HA in a flu vaccine [6]. However, it is important to determine the amount of VES that can be solubilised while ensuring that the clarity of the flu vaccine is unaffected by the stabilisers.

At low VES concentrations (50 Mg/ml), Tween® 80 at concentrations of 30- 150 g/ml are able to readily solubilise the VES present. Further increase in VES to 200 g/ml gradually increases the difficulties in solubilising the VES eventually results in incomplete solubilisation, as indicated by the presence of white particulate matter, when the VES concentration reached 200 g/ml. Refer to Figure 5.

Triton X-100® has a poorer solubilising capacity than Tween® 80. At VES concentrations between 50 and 600 Mg/ml, the solublisation results are satisfactory when the VES:Triton X-100® ratio is equal or below 0.75. At VES:Triton X-100® ratios higher than 0.75, incomplete solubilisation is seen. Refer to Figure 6.

In the absence of bovine serum albumin, simulated monovalent formulae with concentration ratios for Tween® 80:Triton X-100®:VES of 450: 140:500 and simulated QIV formulation with concentration ratios for Tween® 80: Triton X- 100®:VES of 70:250:250 M9/ml are able meet the solution clarity requirements (%T > 80%) and these mixtures are able to be stored for 1 day at 5±3 °C without obvious changes to solution appearance or measured %T. Further increase in VES to 500 MQ/ml for the simulated monovalent bulk solutions results in acceptable %T (> 80%) confirming the stabiliser combination of 70:250:250 g/ml is feasible for finished product. In the presence of Tween® 80 and Triton X-100® at 70 and 250 Mg/ml respectively, any further increase in VES concentration results in %T lower than acceptable limit.

In order to prepare a relatively clear solubilised solution of VES in a combination of Tween® 80 and Triton X-100® in PBS, it is necessary to add the VES solution in a drop-wise manner with continuous agitation.

Based on the maximum VES/Surfactant ratios needed to achieve sufficient clarity (meaning the clarity of vaccine solution is no worse than the FluLaval® QIV formulation that has a high surfactant and VES content). Maximum ratios for VES/Tween® 80 and VES/Triton X-100® are 1.1 and 0.75 respectively. From these, maximum VES concentrations possible are derived for lab studies. PAGE INTENTIONALLY LEFT BLANK

Table 28: Preliminary Stabiliser Formulation for QIV Final Bulk

Table 29: Preliminary Stabiliser Formulation for Mono valent Bulk

* Based on 2x concentration in QIV final bulk.

Further production-scale process feasibility studies (Tables 24 to 26) identified difficulties to solubilise more than 500 μς/ηηΙ VES in a stabiliser mixture of 140 g/ml Tween® 80 and 500 μς/ηηΙ Triton X-100® without reducing %T below 80%. These studies lead to adjustment of the maximum VES content that can be solubilised 450 Mg/ml while keeping the Tween® 80 (140 Mg/ml) and Triton X-100® (500 Mg/ml) content at the same level. Further, fine-tuning of the processing method leads to the best results for %T when the VES concentration is set at 450 g/ml. This can be further diluted to prepare final bulk formulations. The proposed formulations for monovalent bulks and QIV final bulks for process trials are given in Tables 30 & 31 respectively.

Conclusions

Solubilisation of VES in Tween® 80 and Triton X-100® Solutions

Solubilisation studies demonstrate that Tween® 80 has a higher solubilising property when compared with Triton X-100® with a solubilising ratio between VES and surfactant of 1.1 and 0.75 respectively. VES present in concentration higher than this ratio will result in cloudiness in the solution.

Example 8: Effects of Rate of Addition of VES Solution to Tween® 80 Solution

Using the stabiliser concentrations for the original MVB (Table 29), a study is performed to evaluate the impact of adding the VES in one shot or at a gradual rate with intermittent mixing on the final solution transmittance (%T). The results are compared with the reference standard - monovalent bulk (B.12049B) and finished product (B.20130911). PAGE INTENTIONALLY LEFT BLANK

Results and Discussion

Effects of Rate of Addition of VES Solution to Tween 80 Solution

The appearance and %T for different mixtures following drop-wise or a single- shot addition of the VES are recorded in Table 26. The appearance of the mixtures following one-shot addition of VES into Tween 80 solution is consistently more cloudy and higher %T than those following gradual addition of VES.

Table 26: Effects of Rate of Addition of Ethanolic VES Solution to Tween 80 on

Transmittance

benchmark references for appearance and %T, as their appearance is deemed satisfactory.

Example 9: Effects of Bovine Serum Albumin on the Stabiliser Mixture

Using the blank stabiliser mixture for the MVB (VES:Tween® 80 of 533: 1,360), a study is performed to evaluate any potential change to the % Transmittance of the solution with addition of bovine serum albumin (BSA). A concentration of 450 g/ml of BSA is chosen to simulate total protein concentration at 4.5 times the amount of HA in a worst case scenario for total protein concentration. The %T of the stabiliser mixture containing added BSA is compared with that of a standard monovalent bulk solution and a blank stabiliser mixture. Refer to details in Table 30.

Results and Discussion

Effects of Bovine Serum Albumin on the Stabiliser Mixture

The appearance and %T for mixtures of VES/Tween® 80 following addition of Bovine Serum Albumin (BSA) are recorded in Table 27. The %T for BSA-containing and BSA-free stabiliser mixture are 81.1% and 72.91% respectively. There is some apparent increase to the %T with BSA.

Table 27: Effects of Bovine Serum Albumin on Transmittance of Stabiliser Mixture

Example 10: Stabiliser System used for Manufacturing of Monovalent Bulks

Pre-formulation development study described above establishes a series of formulae with reduced Tween® 80 content in the presence of Triton X-100® and VES. Positive (Formula B, FluLaval® monovalent bulk) and negative controls (Formula A, stabiliser-free) are included in this study. Monovalent bulks of each virus strain are manufactured in each of the formulae below (Table 8) against established specifications (Table 9).

Table 8 Taraet Stabiliser Concentrations for Monovalent Bulks

Table 9: Release Specifications for Monovalent Bulks

Manufacturing of Monovalent Bulks in Various Formulations

The process used for the manufacture of monovalent bulks is given in Figure 1.

Buffer Preparation

Solutions containing 1% Tween® 80 (B. RD-B-20131208), 5% Tween® 80 (B. RD-B-20131215) and 2.5% Triton X-100® (B. RD-B-20131210) are prepared using phosphate-buffered saline (PBS) solution (B. RD-B-20131203). The 2.6% VES (B. 13127B) is prepared using 95% ethanol. Comprehensive buffer preparation records are included in Laboratory Book-Secondary 02.

Monovalent Bulk Preparation

Working seeds are prepared in 2012 and 2013. Refer to manufacturing batch records for the following seed passages.

• P/A/TX/P3/13011 (A/Texas/50/2012 NYMC X-223A (H3N2));

• P/A/CL/P3/12026 (A/California/7/2009 NYMC X-179A (H1N1-179A));

• P/B/MC/P3/13012 (B/Massachusetts/2/2012 NYMC BX-51B (B-YMG));

• P/B/BX/P3/13026 (B/Brisbane/60/2008 NYMC BX-35 (B-VIC));

• P/A/CC/P2/13018 (A/Christchurch/16/2010 NIB-74xp (HlNl-74xp)).

The method employed for the production of monovalent bulks used in this formulation study is based on the process for monovalent bulk production as described in manufacturing instructions: FSOP0600B03, FSOP0600B04, FSOP0600B05, FSOP0600B06, FSOP0600B07 and FSOP0600B09. These instructions are based on those developed for GSK FluLaval® Quadrivalent. For the purpose of this formulation study, the addition of stabiliser post-homogenization is omitted. Mixtures of stabilisers are added to a 1 L volume of the various pre-sterile filtered bulks in 2 L bottles in accordance with pre-set formulation components requirements. Refer to Figure 1 for schematic process flow chart.

Addition of Stabilizers

Stabilizer mixtures are prepared according to Figure 2 below. For each stabiliser mixture, twice the amount of needed volumes are prepared in 500 ml glass bottles, in accordance with the required quantities listed in Table 10. This is followed by the addition of VES stock solution into the same bottle. The stabilizer mixtures are continuously stirred in the bottle using a magnetic stirrer. Stabiliser mixtures in required volumes as defined in Table 11 are added into the 1 L monovalent bulks respectively. See Figure 2 for detailed processing steps. Table 10: Stabilizer Mixture Compositions

Table 11: Volume of Stabilizer Mixture Added to 1L Sterile Filtered Monovalent Bulk

Sampling

After the addition of stabilizers to the monovalent bulk, it is continuously stirred for 15 minute before a sample of 310 ml is obtained for release testing. The samples are tested against limits included in Table 9.

Stability Testing Programs for Monovalent Bulks

A total of 31 batches of monovalent bulks are manufactured from the 5 different viral strains (HINl/California, HINl/Christchurch, H3N2/Texas, B/Massachusetts and B/Brisbane) in accordance with the different formulae. The contents for HA, VES, Tween® 80 and Triton X-100® in each monovalent bulk are listed in Table 32.

Each batch of monovalent bulk is filled into a plastic bottle. They are placed under 30±2 °C for up to 4 weeks (accelerated conditions) and at 5±3 °C (real time conditions) for up to 12 months within constant temperature stability chambers or validated stability rooms. Details are included in Tables 12. Samples are withdrawn periodically at the designated time points throughout the stability testing program for the following tests: Product Appearance, HA Content, VES Content and sterility in accordance with established stability testing programs. Stability monitoring of Tween® 80 performed during earlier stability studies on trivalent split-virion seasonal influenza products has confirmed that it is stable in the PBS formulation. The contents of Triton X-100® in the MVB are reported as "quantified by input". Monovalent bulks stored under real time conditions of 5±3 °C are required to meet the End-of Life Specifications (Table 13).

T le 12: Stability Testing Programs for MVBs

Batches 2013FS01-A, B, C, D, E (A/Texas 2014FS01-A, B, C, D, E, D-l

H3N2); (A/California H1N1-179A);

2013FS02-A, B, C, D, E 2014FS02-A, B, C, D, E

(A/California H1N1-179A HlNl); (A/Christchurch HlNl-74xp)

2013FS03-A, B, C, D, E

(B/Massachusetts-Yamagata);

2013FS04-A, B, C, D, E

(B/Brisbane-Victoria).

Temperature 5 ± 3 °C at 0, 1, 2, 3, 6, 9 and 5 ± 3 °C at 0, 1, 2, 3, 6, 9

Conditions/Time 12 months. HA Content for 4 and 12 months. Only HA

Points/Tests strains and VES Contents are Content for A/California

only tested at 0, 1, 2, 3, 6, 9, H1N1-179A or A/Christchurch and 12 months. All tests are H IN l-74xp strain, VES performed at 0 and 12 months. Content are tested at 0, 1, 2,

30 ± 2 °C at 0, 1, 2, 3 and 4 3 and 12 months; All tests weeks. All tests except sterility are performed at 0 and 12 are performed for all test points. months.

30 ± 2 °C at 0, 1, 2, 3 and 4 weeks. All tests except sterility are performed for all test points.

Date of 2013FS01-A, B, C, D, E (A/Texas 2014FS01 (A/California commencement H3N2) - 02Jan.2014 H1N1-179A) - 12.Mar.2014

2013FS02-A, B, C, D, E

(A/California H1N1-179A) - 2014FS02 (A/Christchurch

05Jan.2014 HlNl-74xp) - 19.Mar.2014

2013FS03-A, B, C, D, E

(B/Massachusetts-Yamagata) -

08Jan.2014

2013FS04-A, B, C, D, E

(B/Brisbane-Victoria) - 15Jan.2014

Comment Provides full stability data for Provides full stability data HA

MVBs in different formulae. Content for all formulae for

A/California HlNl-179A and A/Christchurch HlNl-74xp strains.

Table 13: End-of-Life Specifications for Monovalent Bulks

Results and Discussion - Monovalent Bulks in Different Formulae

Manufacturing of Monovalent Bulks

The batch sizes for eggs are typical for normal production scale. The monovalent bulk manufacturing process for formulation study is identical with existing flu vaccine monovalent manufacturing process except that the stabilizer addition is shifted from after homogenization to after sterile filtration.

A total of 31 lots of monovalent bulks are manufactured in different stabiliser formulae as defined in Table 14. Two HlNl seeds, A/California/7/2009 NYMC X- 179A and A/Christchurch/16/2010 NIB-74xp, are used to prepare monovalent bulks in different formulae. Refer to Table 32 for detailed information on manufacturing dates, stabiliser addition times and the quantities of each batch of monovalent bulk produced. Stabiliser addition times for stabiliser addition are different for each monovalent bulk manufactured as these lots are produced in a sequential manner. Time lapse from the end of sterile filtered bulks to addition of stabilisers are monitored. Typically, it is between 110 and 255 minutes. One batch with Formula B (2013FS01-B) has stabilisers added after 410 minutes due to some difficulties encountered during the preparation of the stabilisers as it is the first formula manufactured. The impact of varying time lapses between end of sterile filtration and stabiliser addition is unknown but it is expected to have some negative influence on the stability of the HA antigen resulting in possibly in a slightly lowered final HA content results in the purified bulks. For future full-scale manufacturing process for monovalent bulks, there will be no major delays for stabilizers addition. The stabilizers will be added directly into the sterile filtered monovalent bulks immediately after homogenization as for the stabiliser addition in the current QIV manufacturing process.

Routine IPC (in-process control) checks are performed during the manufacturing process of the monovalent bulks. Refer to Table 33 for IPC data for upstream process (Incubation rates, Egg Surface Bioburden (bacterial, mould and yeasts) at Pre-Incubation, Egg Reject Rates at Candling steps, Bioburden after UV step and Total Protein Content after Inactivation step). All IPC results for incubation rates, bioburden limits and egg reject rate are within acceptable limits. The Total Proteins Content varies from 2,600 to 10,000 Mg/ml for after inactivation with batches 2014FS01 (range of between 2,646 - 2,873 Mg/ml) and 2014FS02 (range of between 2,906 - 3,224 Mg/ml) having lower than the 3,000 g/ml limit set.

IPC data for downstream processes (Table 33) include Sucrose Content at the Isopycnic Centrifugation step, Sucrose Content at the Sucrose Removal step, Transmittance at the Removal of the Splitting Agent, sodium deoxycholate (DOC) and Bioburden after Homogenization step. All the results for downstream IPC tests are within limits and are generally comparable with normal IPC results. These data show that the manufacturing of all batches of monovalent bulks (prior to stabiliser addition) are typical. Table 32: Monovalent Bulk Manufactured

Table 33: IPC results (with control limits)^' for monovalent bulks prior to stabilizer

Monovalent Bulk Batch Analysis Data (with Specification Limits^')

A total of 31 batches of monovalent bulks are manufactured from each of the 5 viral strains. Each monovalent bulk is stabilised with a different combination of stabilisers. Formula A does not contain any stabiliser. Formula B contains the same amount of Tween® 80 and VES as in the FluLaval® Quadrivalent formula.

After addition of the stabilizer mixtures to each monovalent bulk, several samples are taken for monovalent bulk release testing. Batch quantities of monovalent bulks of various formulae produced are between 616 and 771 ml. Variability between batch quantity produced is due to different volumes of stabiliser mixtures added to a fixed quantity of monovalent bulk (1 L) prior to addition of stabilisers. Refer to the Table 34 for all test data reported against respective specifications for each lot of monovalent bulk. Tween® 80 concentrations in these monovalent bulks are quantified by inputs.

Due to presence of interference with the Chinese Pharmacopoeia (Ch.P.) analytical test methods for Tween® 80 in the presence of Triton X-100®, data for Tween® 80 and Triton X-100® are provided as "quantified by inputs" or actual quantities added to the formulae. Since the limits for these stabilisers are 20% around the target stabiliser concentrations, these targets are expected to be achievable given the small-scale trial operations with more accurate controls. Degradation of two surface active agents Tween® 80 and Triton X-100® is not expected to be significant. All the VES Contents for those formulae containing this stabiliser are measured by HPLC against a specification limit of target ± 20%.

Except for the HA content of 2 batches of A/California H1N1-179A containing no stabiliser (Formula A: 2014FS01-A and 2013FS02-A) where their HA contents are below specification of 120 Mg/ml, all lots of monovalent bulks containing different stabilizer combinations show that they are all within acceptable ranges for Appearance, VES Content, HA Content, Triton X-100® Content and Total Protein Content. Sterility tests conducted on all these batches show absence of microbial contaminations.

The batch analysis data for HA Content in monovalent bulks produced from A/Christchurch/16/2010 NIB-74xp (A/Christchurch HlNl-74xp) are all above 300 g/ml when compared with another A/California H1N1-179A strain that gives typically very low contents for HA antigen. Monovalent bulks that are low in HA content are not considered practical for routine production. Further, monovalent bulk lots for A/California H1N1-179A are associated with the formation of precipitate sediments under accelerated conditions. A repeated manufacturing campaign for A/California H1N1-179A viral strain also gives rise to low HA Content. From these results, it can be seen that the low HA Content is consistently seen with the A/California H1N1-179A viral strains but not for the A/Christchurch HlNl-74xp strain. This difference is likely to be due to the nature of different A/H1N1 strains. The A/Christchurch HlNl-74xp viral strain is therefore recommended for future manufacture of A/H1N1 monovalent bulks. It is recommended that further studies be conducted to evaluate the reason for the difference between the H1N1 seeds.

HA Contents for all the negative control batches (monovalent bulks with Formula A) are typically 15-30% lower than the positive control batches (monovalent bulks with Formula B). Based on these comparisons, there is clear evidence that stabilizers are needed for monovalent bulks.

HA Content for monovalent bulks manufactured from A/Texas H3N2, B/Massachusetts YMG and B/Brisbane VIC strains with reduced stabilisers are better than negative control (Formula A) and are comparable to positive controls (Formula B). The HA Contents achieved for these strains are comparable to past experience with monovalent bulks manufactured between 2011 and 2013. These data demonstrate that despite having reduced stabiliser concentrations, comparable HA stabilisation effects of HA antigen in monovalent bulks can be achieved.

All 31 lots are subject to stability testing in accordance with approved stability testing protocols and they are also used for the formulation of QIV Final Bulks and then filled into pre-filled syringes.

Table 34: Monovalent Bulk Batch Analysis Data (with Specification Limits)

2013FS0 132 N/A N/A N/A N/A 231 140 469 546 Complie 2-E s

2013FS0 N/A N/A N/A 332 N/A 1,364 N/A N/A N/A Complie 3-A s

2013FS0 N/A N/A N/A 406 N/A 1,483 1,130 623 N/A Complie 3-B s

2013FS0 N/A N/A N/A 424 N/A 1,536 140 367 552 Complie 3-C s

2013FS0 N/A N/A N/A 426 N/A 1,508 140 428 566 Complie 3-D s

2013FS0 N/A N/A N/A 415 N/A 1,510 140 471 525 Complie 3-E s

2013FS0 N/A N/A N/A N/A 442 1,611 N/A N/A N/A Complie 4-A s

2013FS0 N/A N/A N/A N/A 496 1,689 1,130 616 N/A Complie 4-B s

2013FS0 N/A N/A N/A N/A 502 1,726 140 365 543 Complie 4-C s

2013FS0 N/A N/A N/A N/A 498 1,739 140 415 522 Complie 4-D s

2013FS0 N/A N/A N/A N/A 507 1,733 140 468 512 Complie 4-E s

2014FS0 Complie

102 N/A N/A N/A N/A 239 N/A N/A N/A 1-A s

2014FS0 Complie

120 N/A N/A N/A N/A 297 1,130 653 N/A 1-B s

2014FS0 Complie

123 N/A N/A N/A N/A 294 140 372 500* 1-C s

2014FS0 Complie

122 N/A N/A N/A N/A 292 140 428 500* 1-D s 2014FS0 Complie

122 N/A N/A N/A N/A 291 140 433 500* 1-E s

2014FS0 Complie

122 N/A N/A N/A N/A 291 140 484 500* 1-D-l s

2014FS0 Complie

N/A 282 N/A N/A N/A 752 N/A N/A N/A 2-A s

2014FS0 Complie

N/A 331 N/A N/A N/A 830 1,130 637 N/A 2-B s

2014FS0 Complie

N/A 338 N/A N/A N/A 832 140 366 500* 2-C s

2014FS0 Complie

N/A 337 N/A N/A N/A 823 140 422 500* 2-D s

2014FS0 Complie

N/A 335 N/A N/A N/A 826 140 481 500* 2-E s

* Quantified by input'

Conclusions

Proposed Stabiliser System for Monovalent Bulks and OIV Final Bulks

Separate studies at realistic process scale on solubilisation of VES in both Tween® 80 and Triton X-100® lead to a series of proposed formulations for monovalent bulks with the highest possible VES content of 450 g/ml in the presence of 140 g/ml Tween® 80 and 500 g/ml Triton X-100® without causing any visible increase in cloudiness in the product. Medium and Low VES content levels are set based on 20% from the High VES content and 20% from the Medium VES content respectively. Accordingly, a 1:2 dilution of these formulae is proposed for final QIV bulk. Thus, low, medium and high VES formulae are proposed for manufacturing process trials for both monovalent and final bulks.

Monovalent Bulks Manufactured

A total of 31 batches of monovalent bulks from 5 virus strains (A/California H1N1-179A, A/Christchurch HlNl-74xp, A/Texas H3N2, B/Massachusetts - YMG and B/Brisbane-VIC) have been successful manufactured using existing manufacturing processes for flu vaccines monovalent bulks with minor adaptations for the addition of stabiliser mixtures.

All monovalent bulks manufactured are within specifications for HA content, except for monovalent batches for A/California H1N1-179A strain, total protein, stabiliser contents and sterility. Monovalent bulks manufactured from A/California H1N1-179A have HA Contents that are much lower than others (close to 120 Mg/ml). All monovalent bulks pass sterility tests at release.

Batch analysis data for monovalent bulks confirm that stabilizers are useful in ensuring stability of HA as HA Contents for negative controls batches (Formula A) are lower than all the other batches with added stabilizers. Therefore, stabilizers are necessary excipients for the stabilisation of monovalent bulks. HA Contents for all formulae with reduced amount of Tween® 80 (140 Mg/ml) present in the presence of VES (350-450 Mg/ml) and Triton X-100® (500 Mg/ml) are comparable with the FluLaval® Quadrivalent formula (Formula B, positive control) batches where the amount of Tween® 80 present is 1,130 Mg/ml (with 600 Mg/ml VES).

The monovalent bulks manufactured with reduced Tween® 80 contents are suitable for in QIV formulation and conduct of long term and accelerated stability studies.

Results and Discussion - Stability Testing Program for Monovalent Bulks in Various Formulae

Stability Data for MVBs at 5±3 °C and 30±2 °C

Stability data for Appearance of all MVBs at 5±3 °C and 30±2 °C are included in Tables 38 and 39 respectively. Stability data for HA Content of MBVs (B. 2013FS01- 04, 2014FS01 & 2014FS02) at 5±3 °C and 30±2 °C are included in Tables 40 to 51 and Figures 7 to 18. Stability data for VES Content of MVBs (B. 2013FS01-X to 04- X, 2014FS01-X and 2014FS02-X) at 5±3 °C and 30±2 °C are included in Tables 52 to 63 and Figures 19 to 30. Data for Sterility test on all MVBs at 5±3 °C are included in Table 64. Sterility test is not performed for 30±2 °C. Table 38 : Appearance for MVBs at 5±3 °C

Table 38 (Continued) : Appearance for MVBs at 5±3 °C

Table 39: Appearance for MVBs at 30±2 °C

2014FS01-B Complies Complies Complies Complies Complies

HlNl-74xp) sediment Table 40 : HA Content (A/Texas H3N2) in MVBs fua/ml) at 5±3 °C

Table 41 : HA Content f A/Texas H3N2) in MVBs fua/ml) at 30±2 °C

Table 42 : HA Content f A/California H IN 1-179 A) in MVBs fua/ml) at 5±3 °C

Table 43 : HA Content (A/California H IN 1-179 A) in MVBs (ua/ml) at 5±3 °C

Table 44 : HA Content (A/California H IN 1-179 A) in MVBs (ua/ml) at30±2 °C

Table 45 : HA Content (A/California H IN 1-179 A) in MVBs (ua/ml) at30±2 °C

Table 46 : HA Content (A/Christchurch H1N1-74XD) in MVBs (ua/ml) at 5±3 °C

Table 47 : HA Content (A/Christchurch H1N1-74XD) in MVBs (ua/ml) at 30±2 °C

Table 48 : HA Content (B/Massachusetts - YMG) in MVBs (ua/ml) at 5±3 °C

Table 49 : HA Content (B/Massachusetts -YMG) in MVBs (ua/ml) at30±2 °C

Table 50 : HA Content (B/Brisbane-VIC) in MVBs (ua/ml) at 5±3 °C

Table 51 : HA Content (B/Brisbane -VIC) in MVBs (ua/ml) at30±2 °C

Table 52 : VES Content in MVBs (ua/ml) at 5±3 °C (A/Texas H3N2)

Table 53 : VES Content in MVBs (ua/ml) at 30±2 °C (A/Texas H3N2)

Table 54 : VES Content in MVBs (ua/ml) at 5±3 °C (A/California H1N1-179A )

Table 55 : VES Content in MVBs (ua/ml) at 5±3 °C (A/California H1N1-179A )

Table 56 : VES Content in MVBs (ua/ml) at 30±2 °C (A/California H1N1-179A )

Table 57 : VES Content in MVBs (ua/ml) at 30±2 °C (A/California H1N1-179A )

Table 58 : VES Content in MVBs (ua/ml) at 5±3 °C (A/Christchurch H1N1-74XD)

Table 59 : VES Content in MVBs (ua/ml) at 30±2 °C (A/Christchurch H1N1-74XD)

Table 60 : VES Content in MVB (ua/ml) at 5±3 °C (B/Massachusetts- YMG)

Table 61 : VES Content in MVBs (ua/ml) at30±2 °C(B/Massachusetts -YMG)

Table 62 : VES Content in MVBs (ua/ml) at 5±3 °C (B/Brisbane- VIC)

Table 63 : VES Content in MVBs iua/ml) at 30±2 °C (B/Brisbane- VIC)

Table 64: Stei 'ilitv Results for MVBs at5±3 °C

Batch No. _. Seed sterility at >c sp »ec: » D growth

OM lM 2M 3M 6M 9M 12M 18

2013FS01-A A/Texas H3N2 No growth N/A N/A N/A N/A N/A No growth N/A

2013FS01-B A/Texas H3N2 No growth N/A N/A N/A N/A N/A No growth N/A

2013FS01-C A/Texas H3N2 No growth N/A N/A N/A N/A N/A No growth N/A

2013FS01-D A/Texas H3N2 No growth N/A N/A N/A N/A N/A No growth N/A

2013FS01-E A/Texas H3N2 No growth N/A N/A N/A N/A N/A No growth N/A

2013FS02-A A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2013FS02-B A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2013FS02-C A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2013FS02-D A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2013FS02-E A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2013FS03-A B/Mass. - YMG No growth N/A N/A N/A N/A N/A No growth N/A

2013FS03-B B/Mass. - YMG No growth N/A N/A N/A N/A N/A No growth N/A

2013FS03-C B/Mass. - YMG No growth N/A N/A N/A N/A N/A No growth N/A

2013FS03-D B/Mass. - YMG No growth N/A N/A N/A N/A N/A No growth N/A

2013FS03-E B/Mass. - YMG No growth N/A N/A N/A N/A N/A No growth N/A

2013FS04-A B/Brisbane - VIC No growth N/A N/A N/A N/A N/A No growth N/A

2013FS04-B B/Brisbane - VIC No growth N/A N/A N/A N/A N/A No growth N/A

2013FS04-C B/Brisbane - VIC No growth N/A N/A N/A N/A N/A No growth N/A

2013FS04-D B/Brisbane - VIC No growth N/A N/A N/A N/A N/A No growth N/A

2013FS04-E B/Brisbane - VIC No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-A A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-B A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-C A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-D A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-E A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS01-D- A/Calif. H1N1-179A No growth N/A N/A N/A N/A N/A No growth N/A

2014FS02-A A/Christ. H1N1- No growth N/A N/A N/A N/A N/A No growth N/A

2014FS02-B A/Christ. H1N1- No growth N/A N/A N/A N/A N/A No growth N/A

2014FS02-C A/Christ. H1N1- No growth N/A N/A N/A N/A N/A No growth N/A

2014FS02-D A/Christ. H1N1- No growth N/A N/A N/A N/A N/A No growth N/A

2014FS02-E A/Christ. H1N1- No growth N/A N/A N/A N/A N/A No growth N/A

Stability Data for MVB at 5±3 °C and 30±2 °C

Appearance

The appearance of all batches of MVBs remained unchanged for 12 months when stored at 5±3 °C except for MVB in Formula A. MVBs in Formula A for from both HlNl strains (179A and 74xp) and the two B-strains have small amount of flocculated sediments after 1M while the flocculated sediments are only visible after 2 months for MVB in Formula A for A/Texas H3N2. Under accelerated conditions (30±2 °C), the appearance of all batches of MVB is acceptable initially. The appearance of MVB from A/Texas H3N2, B/Massachusetts-YMG and B/Brisbane-VIC in Formulae B, C, D and E remains unchanged after 4 weeks. MVB from A/Christchurch HlNl-74xp and A/California H1N1-179A strains develop flocculated appearance after 1 and 2 weeks respectively. These floccules appear like protein precipitates. These data indicate that stabilisers are needed for MVBs in order to ensure stable appearance throughout 12 months of storage at 5±3 °C.

HA Content (A/Texas H3N21

The HA Contents for A/Texas H3N2 strain for all batches remain well within the proposed specifications over 12 months. Despite presence of some analytical variability with the HA Content results, Formula A has significantly (about 35%) lower HA Content than Formulae B to E. Formulae B to E have very similar HA stability profiles. HA Content stability data clearly support the proposed 12 month shelf-life for these MVBs when stored at 5±3 °C. At elevated temperature of 30±2 °C, HA Contents for A/Texas H3N2 strain for Formulae B to E decline gradually over the four weeks period when stored at 30±2 °C. Although starting at a lower HA Content, a similar trend is observed for Formula A.

HA Contents for various A/Texas H3N2 MVBs appear to be very stable at 5±3 °C over 12 months. Formula A has a lower starting point at release. All Formulae are within specifications of > 120 g/ml. At the elevated temperature of 30±2 °C, HA is less stable. The results show that A/Texas H3N2 requires inclusion of stabilisers for its stabilisation. HA Content -H1N1

HA Content (A/California H1N1-179A1

HA Contents for A/California H1N1-179A strain for all formulae are close to the specification limit of >120 μς/ηηΙ from the outset and remain close to or below specification limit during the 12 month period when stored at 5 ± 2 °C. HA Content for Formula A is markedly lower than the rest of the Formulae. It is below the specification limit at release and remains below the limit throughout the 12 months storage period. A decline of about 10% is seen during the first month. From the second month onwards the decrease is much more gradual until the 12 month time point. For overall trend point of view, the proposed 12 months shelf-life when stored refrigerated cannot be supported since the starting concentration is very close to specification limit. Linear decline of HA Content is seen for all formulae at the elevated temperature 30±2 °C over the 4 weeks. These results demonstrate A/California H1N1-179A is very temperature sensitive and it needs to be stored at 2-8 °C.

HA Content ί A/Christchurch H1N1-74XD1

HA Contents for A/Christchurch HlNl-74xp achieve much higher concentrations at release for all Formulae and they remain stable over 12 months. Formula A although starts off at a significantly lower point, its overall stability profile remain similar. HA Content stability data clearly support the proposed 12 month shelf- life at 5±3 °C for A/Christchurch HlNl-74xp MVBs.

The HA Contents for A/Christchurch HlNl-74xp Formulae B to E are relatively stable (loss of about 10 %) after 4 weeks. Formula A has a lower HA Content at release and declines much faster than those formulae containing stabilisers. The higher thermal stability is noted for the HA from the A/Christchurch HlNl-74xp strain when compared with the A/California H1N1-179A strain.

HA Content - H IN 1

All A/California H1N1-179A Formulae have HA Content at close to or below 120 Mg/ml at the start of the stability studies. At 30 °C the A/California H1N1-179A HA degrades at a fast rate at 30 °C. HA Content from A/Christchurch HlNl-74xp strain are generally higher than the A/California H1N1-179A strain and are much more stable at elevated temperature of 30±2 °C. However, this difference is not as obvious at the 5±3 °C where the HA Content for A/California H1N1-179A is stable over 12 months after initial loss following 1 month of storage.

HA Content (B/Massachusetts YMO

The HA Contents for B/Massachusetts strain from the Yamagata lineage for all Formulae remain within the proposed specifications over 12 months at 5±2 °C despite a decreasing trend. The HA Content decreases by over 30% in 12 months. Formula A has a lower starting point but displays the same trend. HA stability data clearly support the proposed 12 month shelf-life at 5±3 °C for B/Massachusetts YMG MVB. An overall decrease of 40% is seen after 4 weeks at 30±2 °C. Formula A has a much lower HA Content at release than all the MVBs with stabilisers but displays a stable stability profile.

The HA Content results for the B/Massachusetts (Yamagata Lineage) for all Formula B to E are all within specifications over 12 months at 5±3 °C despite some linear reduction. Without any stabiliser present, HA Content decreases markedly during manufacturing and prior to testing. All stabiliser combinations containing VES, Tween® 80 and Triton X-100® are useful to stabilise the HA.

HA Content (B/Brisbane-VIO

The HA Contents for B/Brisbane strain from the Victoria lineage for Formulae B to E display slight decreasing trends but remain within the propose specifications over 12 months at 5±3 °C. Formula A has a lower HA Content at release but displays the same stability trend. HA stability data clearly support the proposed 12 month shelf-life at 5±3 °C for B-VIC MVB. The HA Contents for B/Brisbane strain from the Victoria lineage for Formulae B to E decrease almost linearly from 500 to 280 g/ml after 4 weeks at 30±2 °C. Formula A starts off at a 10% lower HA Content at 450 g/ml than all the MVB with stabilisers but maintains a similar profile. All the HA Contents for B/Brisbane (Victoria lineage) are within specifications. A gentle decline in HA Content is observed following a faster decline over the first month of storage at 5±3 °C. At 30±2 °C, the HA contents are declining faster. The results show that in MVBs HA Content for B/Brisbane-VIC required an addition of a combination of stabilisers in order to meet the proposed shelf-life.

VES Content

VES Contents for all Formulae B to E for all strains remain within the limits of ± 20% of targets over 12 months whens stored at 5±3 °C. VES Contents for Formulae B to E for all strains decline about 10% after 4 weeks at 30±2 °C. This decline in VES is linked to the increased rates of hydrolysis at high temperatures.

All the VES results are within acceptable limits of ± 20 % around the targets. A slight declining trend is visible over the 12 months at 5±3 °C. For the elevated condition of 30±2 °C, VES trend is more obvious than Formula B where the VES concentration is the highest. This is to be expected as degradation of VES follows first order kinetics as seen in previous studies. Overall, the limits of ± 20% is not expected to be exceeded over the proposed shelf-life of the MVB of 12 months at 5±3 °C.

Sterility

Sterility is performed at the beginning of the stability testing program for all Formulae. All batches pass sterility tests at the start and at the 12-month time point in the program. Sterility tests are not repeated for all batches at the end of 4 weeks at 30±2 °C. Sterility test results for all batches of MVBs are satisfactory at initial and final test points of 12 months. Conclusion

Stability Studies on Monovalent Bulks

Available 12 months real time (5±3 °C) and 4 weeks accelerated (30±2 °C) stability data showed that MVBs formulated without any stabiliser result in the formation of flocculated sediments, and unstable HA resulting in lower starting concentrations. There is some decrease in the VES Content in the MVBs with Formula B (FluLaval® formula) but all are within the ± 20% acceptable limits over 12 months. This study showed that stabilisers are needed in order to maintain a high concentration of HA in MVB and to prevent aggregations of HA antigens. Overall, batches containing a combination of stabilisers including VES at concentrations between 350 and 450 Mg/ml, Tween® 80 at a concentration of 140 Mg/ml and Triton X-100® at a concentration of 500 Mg/ml, provide stabilisation effects to the HA antigens in the MVBs over a storage of 12 months at 5±3 °C.

Those formulae with reduced stabiliser (Formulae C, D and E) achieve the same or better stabilisation effects of HA when compared with the FluLaval® MVB formula (Formula B) that has a very high concentration of VES at 600 g/ml and Tween® 80 at 1,130 Mg/ml. These results confirm that reduced stabiliser concentration of VES and Tween® 80 while adding Triton X-100® at a concentration of 500 Mg/ml in the MVBs results in no loss of stability for HA Content. Available data support the proposed shelf-life of 12 months for the in Formulae C, D and E.

Example 11: Manufacturing of QIV Final Bulks and Filled Syringes in Various Formulations

Stabiliser System used for Manufacturing of QIV Final Bulks

Pre-formulation development study described above established a series of formulae with reduced Tween® 80 content in the presence of Triton X-100® and VES that are identified as potential stabilizers. Positive (Formula B, FluLaval® Quadrivalent) and negative controls (Formula A, stabiliser-free) are included in this study. Final QIV bulks are manufactured based on each of the 7 formulae below (Table 14). A total of 9 lots of QIV filled syringes are manufactured with different formulae in lots of 1 L and filled aseptically into 0.5 ml ceramic-coated tip (CCT) prefilled syringes within a sterile filling production area. Samples are taken from each lot for testing against established release specifications (Refer to Table 15).

Table 14: Target Stabiliser Concentrations for Final Quadrivalent Bulks and Filled Syringes

Representative samples taken from each the bulk and each lot of filled syringes are obtained for release testing to the specifications contained in Table 15. Abnormal toxicity test, in accordance with Chinese Pharmacopoeia (Ch.P.), is a vaccine product safety test involving slow administration by IM injection of doses of 0.5 (25 ml/kg) and 5.0ml (16.7 ml/kg) of the test samples into the mice (body weight of approximately 20g each) and guinea pigs (body weight of approximately 300g each) respectively after equilibration to room temperature. The doses tested in mice and guinea pigs represent 2,167 to 3,250 times the intended human dose calculated from an average body weight of 65 kg in China respectively. The gains in the body weight of treated animals are determined for 1 week and compared with the weight gain for groups of untreated animals. Absence of abnormal toxicity is indicated by comparable weight gains in both the treated and untreated animal groups.

In order to avoid unnecessary use of lab animals to detect any adventitious toxic contaminants or any unexpected toxic components, testing for abnormal toxicity is kept to a minimum. Testing is only performed on two lots, where the stabiliser concentrations are the highest (Batches RD-Fill-2014-B-0101 and RD-HII-2014-E- 0105).

Table 15: Release Specifications for Quadrivalent Final Bulks and Filled Syringes

Buffer Preparation

All the buffer solutions are prepared using WFI (Water for injection) except for the VES buffer solution is used as the solvent vehicle. The buffer preparations are conducted under Grade C environment following the standard operating procedure for Buffer preparation. All buffer solutions are filtered into sterile bags, sterile Nalgene® bottles or sterile glass bottles under the Unidirectional Airflow (UDF) bench. The 2.6% VES/95% Ethanol solution uses 95% ethanol solution as solvent vehicle. For detailed instructions on buffer preparation and comprehensive records and, refer to Secondary Laboratory Book 03.

Formulation

Laboratory experiments are performed to determine the critical process parameters such as stabiliser concentrations, order of stabilizers addition, stirring speed and stirring durations.

The final bulk preparation is performed in batch sizes of 1 L each within 2 L sterile glass bottles instead of the normal 10 L sterile plastic bags. The use of bottles instead of plastic bags does not have any impact on the formulations. This formulation process is similar to the existing influenza vaccine formulation process (FSOP0600B84) with more the use of a sterile bag and a close system (pump, tubing, filter) for product and solution transfer. This process provides a better formulation accuracy and provides a better sterile assurance as it consisted of a sterile bags.

The manufacturing of final bulks is described below in the Figure 3. All the formulations are performed under laminar airflow (LAF) within a Grade B background to ensure formulation sterility (Refer to Laboratory Book 07). Required volumes of monovalent bulks and stabilizers are calculated according to FSOP0600A25V00.

The formulation of all the final bulks for the series A-E or B2, M and N campaigns are completed within the same day with all the samples tested on the following day. In this way, variation in test results for the initiation of stability testing is minimised.

The only test item performed at in-process control (IPC) for the final bulks is HA content. If HA content of any strain exceeds ±10 % of the target, adjustment of formulation is normally required. Due to the presence of cross-interference between the absorbance of Tween® 80 and Triton X-100® at the measured wavelength in the current Chinese Pharmacopoeia (Ch.P.) test method for Tween® 80 is unsuitable for use in the study. Therefore, the concentrations of these stabilisers in the final bulks and finished products are reported as "quantified by input".

Table 16: Manufacture of Final Bulks and Filled Syringes

Note: Due to low yield of HA in the A/California/7/2009NYMC X-179A monovalent bulk and its poor stability under accelerated condition, a repeat of H1N1-179A monovalent bulk is conducted in parallel with a second H1N1 strain (A/Christchurch/16/2010 NIB-74xp). Final QIV bulks are manufactured using H1N1- 74xp strain. Required volumes of MVBs needed for each QIV final bulk are determined from the HA contents shown in Table 32. Target volumes of stabilizer buffers needed, as required for each formula, are calculated according to manufacturing instructions: FSOP0600A25. Both target and actual volumes dispensed are recorded (Tables 17 and 18).

Table 17: Target and Actual Volumes of MVBs Added to the Final Bulks

Note: * The HA content of final bulk batch (RD-Form-2014-A-0109 is corrected as it exceeds IPC range of 10% target.

Table 18: Required and Actual Volume of Stabilizers Added in Final Bulk

Filling

The formulated final bulks are filled into ceramic coated tip (CCT) glass syringes using the existing Corimar filling machine under laminar airflow unit (LAF) in Grade A area. The filling process used for this study is similar to that used for the existing QIV (Master Manufacturing Instructions FSOP0600B69). Each final bulk is filled into filled syringes (PFS) in lots of approximately 1,000 each. The different batches of final bulks are filled on the same day according to the process flow in Figure 4.

Results and Discussion

QIV Final Bulks and Filled Syringes in Different Formulae

Final Bulk Formulation

Nine batches of QIV final bulk formulations are manufactured in accordance with Tables 17 and 18. One of these batches (A-0109) has HA contents adjusted since they are outside of the 10% IPC limits. Batch analysis results for all 9 batches are included in Table 35.

Table 35: Batch Analysis Data for Quadrivalent Final Bulks

* HA content of this batch is adjusted.

Due to presence of interference with the Chinese Pharmacopoeia (Ch.P.) analytical test methods for Tween® 80 in the presence of Triton X-100®, data for Tween® 80 and Triton X- 100® are provided as "quantified by inputs" or actual quantities added to the formulae. Since the limits for these stabilisers are 20% around the target stabiliser concentrations, these targets are expected to be achievable given the small-scale trial operations with more accurate controls. Degradation of two surface active agents Tween® 80 and Triton X-100® is not expected to be significant. All the VES Contents for those formulae containing this stabiliser are measured by HPLC against a specification limit of target ± 20%. All 9 batches of QIV final bulks meet specifications and are acceptable for filling into CCT pre-filled syringes. Filling of QIV Final Bulks into Syringes

A total of 9 batches of filled syringes are manufactured in PFS syringes in accordance with existing routine filling procedures (FSOP0600B69). All IPC results for final products are within control limits. All syringes are stored in nests within plastic tub, with protection from light. Batch analysis data for these batches are given in Table 36.

Table 36: Batch Analysis Data at Release for QIV Final Bulk

11.3 Batch Analysis Data for Filled Syringes

Each lot of naked filled syringes (non-blister packed nor filled into carton) is tested against proposed release specifications (Refer to Table 37). All formulations are able to be successfully formulated to within targets of 26 - 40 pg/ml and 27 - 41 pg/rnl for the two A strains (HlNl & H3N2) and the two B strains (B-YMG & B-VIC) respectively. These lots are filled into pre-filled syringes. Due to interference with the Chinese Pharmacopoeia (Ch.P.) test method for Tween® 80, the contents of Tween® 80 are given as "quantified by input". Release data for the finished products containing different stabilizer combinations show that they are all within acceptable ranges for Appearance, pH, Osmolality, VES Content, HA Content and Total Protein Content. Sterility tests conducted on all these batches show absence of microbial contaminations.

Table 37: Batch Analysis Data at Release for QIV Filled Syrin es

d h

640]

125

Not No

Com pi i Not [104 250

M - 0512 7.2 70* owt es tested * 34 35 33 32 teste gr d h

156]

126

Not No

Com pi i Not [104 190

N - 0513 7.2 50* t es tested * 34 34 33 32 teste grow d h

156]

196

Not No

D2 Com pi i Not [160 250

7.1 70*

0514 es tested * 32 35 32 32 teste growt d h

240]

* quantified by input

Conclusions

QIV Final Bulks and QIV Filled Syringes Manufactured

Nine batches of final bulks and filled syringes have been successfully manufactured following the set procedure without any deviation. All IPC results for final bulks and final products are within the control limits. All formulated QIV final bulks and filled syringes met all pre-established release limits. Data from release testing show that QIV Final Bulk formulations with reduced stabilisers can be manufactured using existing manufacturing process with minor adaptations. Batch analysis data for final bulks and filled syringes confirm no significant differences in HA Content between these batches. Stabilisers used do not impact product sterility. In addition, test results from two batches selected to represent worst cases in terms of content of stabilisers are tested for abnormal toxicity in mice and guinea pigs. Both batches do not show any sign of toxicity. Hence, samples from each batch of filled syringes are satisfactory for use in the real time and accelerated stability studies. Example 12: Stability Studies for QIV Filled Syringes

QIV Filled Syringes Batches

Opaque plastic tubs of naked filled syringes (non-blister packed and not packed into carton) are stored in tubs and placed under 30±2 °C for up to 4 weeks (accelerated conditions) and at 5±3 °C for up to 12 months (real time conditions) within constant temperature stability chambers or validated stability rooms. A total of 9 batches of filled syringes are included in the stability testing programs. During stability studies, samples are periodically withdrawn at the designated time points, in sufficient quantities, for the following tests: product appearance, pH, HA content, VES content and sterility in accordance with established stability testing protocols. Details of the stability studies are included in Tables 19. The content of Tween® 80 and Triton X-100® in the final bulks and filled syringes are reported as "quantified by input". Filled syringes stored under real time conditions (5±3 °C) are required to meet the End-of Life Specifications for filled syringes (Table 20).

Stability Testing Programs QIV Filled Syringes

Details of the 3 stability programs initiated to assess the stability of the 9 batches of QIV filled syringes in various formulae (A, B, C, D, E, B2, M, N and D2) are shown in Table 19. All stability testing programs are performed within temperature controlled enclosures that have been previously validated for the purpose. The samples at real time conditions of 5±3 °C are expected to meet the End-of-Life specification in Table 20.

Table 19: Stability Testing Programs for QIV Filled Syringes

Table 20: End-of-Life Specifications for Quadrivalent Filled Syringes

Results and Discussion

Stability Testing Program for QIV Filled Syringes in Various Formulae

Stability Data for Fill Syringes at 5±3 °C and 30±2 °C

Stability data for Appearance of finished products at 5±3 °C and 30±2 °C are included in Tables 65 and 66 respectively.

Data for pH values of finished products at 5±3 °C and 30±2 °C are included in Tables 67 and 68 respectively (Figures 31 to 34). Stability data for HA Content of finished products at 5±3 °C and 30±2 °C are included in Tables 69 to 77 (Figures 35 to 50). Stability data for VES Content of finished products at 5±3 °C and 30±2 °C are included in Tables 78 and 79 (Figure 51 to 54) respectively. Data for Sterility test on finished products at 5±3 °C are included in Tables 80. Sterility test is not performed at 30±2 °C.

5

Table 65 : Appearance of OIV Filled Syringes for 5±3 °C

Table 66 : Appearance for QIV Filled Syringes at 30±2 °C

Table 67 : DH Results for OIV Filled Syrinaes at 5±3 °C

B2

7.0 N/A 7.0 7.0 7.0

0511 7.0 7.0 7.0

M - 0512 7.2 7.1 7.1 7.2 N/A 7.1 7.2 7.2

N - 0513 7.2 7.1 7.1 7.2 N/A 7.1 7.1 7.2

7¾fr/e 6ff / £>H flesE/fe for O/V Filled Syrinaes at 30±2°C

Table 69 : HA Content for H1N1-179A for OIV Filled Syrinaes at 5±3 °C

Table 70 : HA Content (A/Christchurch H1N1-74XD) for OIV Filled Syrinaes at 5±3 °C

Table 71 : HA Content (A/California H IN 1-179A) for OIV Filled Syrinaes at 30±2 °C

Table 72 : HA Content (A/Texas H3N2) for OIV Filled Syrinaes (5±3 °C)

Table 73 : HA Content (A/Texas H3N2) for OIV Filled Syrinaes (30±2 °C)

Table 74 : HA Content (B/Massachusetts B-YMG) for OIV Filled Syrinaes at 5±3 °C

Table 75 : HA Content (B/Massachusetts B- YMG) for OIV Filled Syrinaes (30±2 °C)

Table 76 : HA Content (B/ Brisbane B-VIC) for ON Filled Syrinaes at 5±3 °C

Table 77 : HA Content (B/ Brisbane B- VIC) for OIV Filled Syrinaes at 30±2 °C

Table 78 : VES Content (ua/ml) for OIV Filled Syrinaes at 5±3 °C

Table 79 : VES Content iua/m I) for OIV Filled Syrinaes at 30±2 °C

Table 80: Sterility Data for Filled Syrinaes at 5 ±3 °C

Stability Data for Filled Syrinaes at 5±3 °C and 30±2 °C

Appearance

The appearance of batches RD-Fill-2014-B-OlOl, C-0102, D-0103, E-0105, B2-0511, M- 0512, N-0513 and D2-0514 remain unchanged for 12 months at 5±3 °C. Batches RD-Fill-2014- A-0106 that contains no VES/Tween® 80/Triton X-100®, develops small amount of flocculated sediments after 1 to 3 months. The appearance of batches RD-Fill-2014-B-0101, C-0102, D- 0103, E-0105, B2-0511, M-0512, N-0513 and D2-0514 remains unchanged for 4 weeks at 30±2 °C. At elevated temperature of 30 °C, flocculated sediments are seen after 2 weeks with batch RD-RII-2014-A-0106 (Formula A).

Small amount of flocculated sediments is seen at the bottom of all batches with Formula A after 1 month at 5±3 °C. At elevated temperature of 30±2 °C, flocculated sediments appear after just 1 week. These floccules appear like protein precipitates. These data show that stabilisers are needed for QIV formulations.

2d

pH values for all batches RD-RII-2014-A-0106, B-0101, C-0102, D-0103, E-0105, B2- 0511, M-0512, N-0513 and D2-0514 remain unchanged at between 7.0 and 7.2 for 12 months 5 ± 3°C The pH values for all batches RD-FNI-2014-A-0106, B-0101, C-0102, D-0103, E-0105, B2-0511, M-0512, N-0513 and D2-0514 remain unchanged at between 7.0 and 7.2 for 4 weeks 30±2 °C.

The pH of all the QIV formulations remain unchanged at between 7.0 and 7.2 throughout the 12 months at 5±3 °C and 4 weeks at 30±2 °C. Therefore, different amounts of stabilizer, storage temperature and duration of storage have no impact on the product pH.

HA Content fA/California H1N1-179A & A/Christchurch H1N1-74XP1

HA Contents for A/California H1N1-179A strain for all batches RD-RII-2014-A-0106, B- 0101, C-0102, D-0103, E-0105, B2-0511, M-0512, N-0513 and D2-0514 decline about 10 % during the first month at 5±3 °C. From the second month onwards the decrease is more gradual until 12 months. The HA Contents (for A/California H1N1-179A) for all batches of filled syringes RD-FNI-2014-A-0106, B-0101, C-0102, D-0103 and E-0105 decline markedly in the 4 weeks at 30±2 °C. HA Contents are below 24 pg/rnl for all batches after 2 weeks at 30±2 °C except for batch A-0106. These data demonstrate that A/California H1N1-179A is very temperature-sensitive. In contrast, the HA contents for A/Christchurch HlNl-74xp batches RD- RII-2014-B2-0511, M-0512 and N-0513 and D2-0514 are relatively stable (loss of 10 %) after 4 weeks. The greater thermal stability for the A/Christchurch H lNl-74xp strain is noted

At 30 °C the HA degrades at a fast rate such that after 2 weeks at 30°C±2 °C, 5 of the batches have HA Content below 24 pg/rnl. There is a large difference between the stability profiles for HA Content of the two H1N1 strains (A/California-179A and A/Christchurch-74xp). The HA from the A/Christchurch HlNl-74xp strains is much more stable at elevated temperature of 30 °C However, this difference is not as obvious at the 5±3 °C where the HA Content for A/California H1N1-179A is stable after initial loss following 1 month of storage.

No real difference in HA Content stability is seen between the different formulae. This result shows that the stability of HA Content is more dependent on the storage temperature than the amounts or types of stabilizer present.

HA Content (A/Texas H3N21

The HA Contents for A/Texas H3N2 strain for all batches RD-RII-2014-A-0106, B-0101, C-0102, D-0103 and E-0105 remain within the proposed specifications over 6 months. Due to test method variability HA content for batch B-0101 (Formula B) is above proposed specification at 1 month test point at 5 ± 3°C HA Contents for batches RD-FMI-2014-B2-0511, M-0512 and N-0513 remain unchanged after 3 months. Batch RD-FNI-2014-D2-0514 is not tested for HA Content for A/Texas H3N2. HA Contents for A/Texas H3N2 strain for batches RD-Fill-2014-B-0101, C-0102, D-0103, E-0105, B2-0511, M-0512 and N-0513 are relatively stable at 30 ± 2°C for 4 weeks. This is within the method variability (about ±10 %). Batch RD-FNI-2014-A-0106 is much less stable such that after 2 weeks their HA Contents are at or below 24pg/ml. The instability with these two batches is due to absence of stabiliser.

HA Contents for the various batches A/Texas H3N2 appear to be very stable at 5±3°C. All batches are within specifications of between 24 and 36 pg/ml. At the elevated temperature of 30 °C, one batch RD-FNI-2014-A-0106 is found to be less stable. The results show that A/Texas H3N2 requires inclusion of stabilisers for its stabilisation.

HA Content fB/Massachusetts -YMO

The HA Contents for B/ Massachusetts B-YMG for all batches RD-FNI-2014-A-0106, B-

0101, C-0102, D-0103 and E-0105 remain within the proposed specifications over 12 months at 5 ± 3°C Batch RD-Fill-2014-B-0101 seems to be declining faster than expected because it has a low starting point. HA Contents for batch RD-FNI-2014-B2-0511 decline by 16% after 2 months while batches RD-FNI-2014-M-0512 and N-0513 remain unchanged. Batch RD-Fill-2014- D2-0514 is not tested for HA Content for B/Massachusetts B-YMG.

HA Contents for B/ Massachusetts B-YMG for batches RD-RII-2014-A-0106, B-0101, C-

0102, D-0103, E-0105, B2-0511, M-0512 and N-0513 are declining relatively steadily at 30 ± 2°C with batches A-0106 and B2-0101 being the most rapid.

Batches RD-Fill-2014-B-0101 and B2-0101 seem to be declining faster than expected because they have lower starting points. The instability with A-0110 is due to absence of stabiliser.

The HA Content results for the B/Massachusetts - YMG are all within specifications. All other stabiliser combinations containing VES, Tween® 80 and Triton X-100® are useful to stabilise the HA. At 5±3°C, batch RD-FNI-2014-A-0106, which does not contain any stabiliser is relatively stable while for the FluLaval® Quadrivalent (RD-Fill-2014-B-0101 and B2-0511) with high contents of VES/Tween® 80 are unexpectedly showing least stability. Higher decrease in HA Contents are also seen at the elevated temperature for batches RD-FNI-2014-A-0106 and B2-0511. These anomalies can be due to analytical variability with the adapted SRID test method and low starting content for this batch. The results show that in QIV formulation HA content for B/Massachusetts-YMG require an addition of a combination of stabilisers in order to meet the proposed product shelf-life. HA Content fB/Brisbane - VIC)

The HA Contents for B/Brisbane - VIC for batches RD-FNI-2014-A-0106, B-0101, C-0102, D-0103 and E-0105 remain within the limits of 24 to 36 pg/rnl of targets for 12 months at 5 ± 3°C HA Contents for batches RD-RII-2014-B2-0511, M-0512, N-0513 and D2-0514 remain within acceptable limits after 2 months. Batch RD-Fill-2014-B-0101 appears to trend lower than the other because of low starting point.

The HA Contents for B/Brisbane - VIC for all batches RD-RII-2014-C-0102, D-0103, E- 0105, B2-0511, M-0512, N-0513 and D2-0514 decline at 30 ± 2°C rapidly after the first week then gradually over the 4 weeks. Batches RD-FNI-2014-A-0106 and B-0101 have the most obvious decline in HA Content. Batch RD-FNI-2014-B2-0511 has a low starting value.

All the HA Contents for B/Brisbane - VIC are within specifications. They all seem to show gentle declining trends following a faster decline over the first month of storage at 5±3°C At 30±2°C, the HA contents are declining fastest for batches RD-FNI-2014-A-0106 (where no stabiliser has been added to the formula), B-0101 (stabiliser system as per FluLaval® Quadrivalent) and RD-FNI-2014-B2-0511 (stabiliser as per FluLaval® QIV). As with the B/Massachusetts strain, the FluLaval® Quadrivalent formula (batches RD-Fill-2014-B-0101 and B2-0511) are unexpectedly showing the worst stability. This anomaly can be due to analytical variability with the adapted SRID test method and low starting content for this batch. The results show that in QIV formulation HA content for B-VIC requires an addition of a combination of stabilisers in order to meet the proposed product shelf-life.

VES Content

VES Contents for all batches RD-RII-2014-A-0106, B-0101, C-0102, D-0103 and E-0105 remain within the limits of ± 20% of targets for 12 months at 5±3°C VES Contents of batches RD-FMI-2014-B2-0511, M-0512, N-0513 and D2-0514 remain within the limits of ± 20% of targets for 12 months at 30±2°C. VES Contents for all batches (RD-FNI-2014-) A-0106, B-0101, C-0102, D-0103, E-0105, B2-0511, M-0512, N-0513 and D2-0514 decline about 10% after 4 weeks at 30±2°C This decline in VES is linked to the increased rates of hydrolysis at high temperatures.

All the VES results are within acceptable limits of ± 20% around the targets. No real declining trend is visible over the 12 months. For the elevated condition of 30±2°C, VES trend is more obvious for batch 0101-B where the VES concentration is the highest. This is to be expected as degradation of VES followed first order kinetics as seen in previous studies. Overall, the limits of ±20% will not be exceeded over the proposed shelf-life of the product of 12 months at 5±3°C Sterility

Sterility is performed at the beginning of the stability testing program for all batches RD- RII-2014-A-0106, B-0101, C-0102, D-0103, E-0105, B2-0511, M-0512, N-0513 and D2-0514. Sterility test is repeated for all batches at the end of the stability programs. All batches pass sterility tests at the start and end of the 12 months stability testing program.

Conclusions

Stability Studies on QIV Filled Syringes

Available 12 months real time (5 ± 3°C) and 4 weeks accelerated (30 ± 2°C) stability data show that QIV formulated in the PBS solution without any stabiliser result in formation of flocculated sediment and unstable HA. There is some decrease in the VES concentration in the formulations but all are within acceptable limits over 12 months.

This study shows that stabilisers are needed in order to prevent aggregations of some strains of HA proteins. There is evidence that some strains (particularly A/Texas H3N2) need stabiliser in order to maintain optimum HA content stability. Overall, batches containing a combination of stabilisers including VES, Tween® 80 and Triton X-100®, on the other hand, provide stabilisation effects to the HA antigens in the formulation.

Those formulae with reduced stabiliser concentrations (Formulae C, D and E) achieve the same or better stabilisation effects of HA when compared with the FluLaval® (Formula B) that has a very high concentration of VES and Tween® 80. These results confirm that reduced stabiliser concentration in the QIV final bulk formulation for splitting agents of Tween® 80 (NMT 80 pg/rril) and Triton X-100® (NMT 300 pg/ml) result in no loss of stability for HA content. Formulae with even lower concentrations in stabiliser (Formulae M and N) although appear promising to ensure product stability.

Available 12 month real time stability data support the proposed shelf-life of 12 months

(at 5 ± 3°C) for the QIV filled syringes in Formulae C, D and E. Based on the satisfactory abnormal toxicity and stability data and considering the needs to have minimum additives in a vaccine and a robust stabiliser system to cater for anticipated changes in viral strains from year to year, it is recommended that Formula C be chosen for further stability evaluations.

Claims

Claims

1. A vaccine composition comprising a split influenza virus preparation or subunit influenza virus preparation and a pharmaceutically acceptable excipient said excipient comprising polyoxyethylene sorbitan monooleate (Tween® 80 or Polysorbate 80), alpha-tocopherol and t-octylphenoxypolyethoxyethanol (Triton X-100®), wherein the amount of polyoxyethylene sorbitan monooleate per human dose is <200 pg and the amount of t- octylphenoxypolyethoxyethanol per human dose is >100 pg.

2. The vaccine composition of claim 1 wherein the derivative of alpha tocopherol is alpha tocopherol succinate.

3. The vaccine composition of claims 1 or 2 wherein the amount of polyoxyethylene sorbitan monooleate per human dose is from 20 to 200 pg and the amount of t- octylphenoxypolyethoxyethanol per human dose is from 100 to 600 pg.

4. The vaccine composition of any preceding claim wherein the amount of polyoxyethylene sorbitan monooleate per human dose is from 50 to 150 pg and the amount of t- octylphenoxypolyethoxyethanol per human dose is from 100 to 400 pg.

5. The vaccine composition of any preceding claim wherein the amount of polyoxyethylene sorbitan monooleate per human dose is (i) from 20 to 50 pg per 0.25 ml paediatric dose or (ii) from 60 to 90pg per 0.5 ml adult dose.

6. The vaccine composition of any preceding claim wherein the amount of Triton X-100™ per human dose is (i) from 120 to 170pg per 0.25 ml paediatric dose or (ii) from 250 to 350pg per 0.5ml adult dose.

7. A multi-dose or single dose vial comprising the vaccine composition of any one of claims l to 6.

8. A process for producing a vaccine composition comprising the steps of: (i) preparing a split influenza virus preparation or subunit influenza preparation, (ii) optionally inactivating the split influenza virus preparation before or after the splitting step, (iii) purifying and filtering said preparation and (iv) adding a pharmaceutically acceptable excipient to the preparation, said excipient comprising the following stabilisers: polyoxyethylene sorbitan monooleate, alpha-tocopherol and t- octylphenoxypolyethoxyethanol, wherein the concentration of polyoxyethylene sorbitan monooleate is <200 pg/ml and the concentration of t-octylphenoxypolyethoxyethanol is from 300 to 700 pg/ml.

9. The process of claim 8 wherein the wherein the concentration of polyoxyethylene sorbitan monooleate is from 50 to 200 pg/ml and the concentration of t- octylphenoxypolyethoxyethanol is from 400 to 600 pg/ml.

10. The process of claim 8 or 9 further comprising storage of the vaccine composition in bulk after excipient has been added.

11. The process of any one of claims 8-10 further comprising putting the vaccine composition containing excipient into individual or multi-dose vials ready for storage or administration.

12. A vaccine composition according to any one of claims 1 to 6 for use in medicine.

13. A vaccine composition according to any one of claims 1 to 6 for use in the treatment, prevention and/or vaccination against disease caused by influenza.

14. A method of inducing an immune response in a human subject, said method comprising administering to the subject a vaccine composition of any one of claims 1 to 6.

15. A method of treatment, prevention and/or vaccination against influenza disease, comprising the administration of a vaccine composition according to any one of claims 1 to 6 to a person in need thereof (such as an elderly person age 50 or over, particularly age 65 or over).