Removal of agglomerates

Abstract

The present disclosure provides a method of dispersing agglomerated material in a formulation comprising influenza proteins. The method comprises subjecting the formulation to ultrasonication.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to Australian application No. 2018902497 entitled "Removal of Aggregates", filed on July 10, 2018, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This disclosure relates to a method for dispersing an agglomerate in a formulation containing an influenza antigen, and more particularly to the use of this method in the production of influenza vaccines.

[0004] background

[0005] Influenza vaccines are considered the most effective way to prevent infection. The first influenza vaccine was a whole-virus preparation[1]. The current inactivated trivalent and quadrivalent influenza vaccines (TIV and QIV, respectively) are produced using processes based on the chemical destruction or “lysis” of the influenza virus, which began in the 1960s[2]. It was found that chemical destruction (by detergents or solvents) reduces the reactivity of the vaccine, but in many cases does not impair immunogenicity. Due to the high volatility of solvents, all commercially available influenza vaccines are destroyed or lysed by detergents. However, the concentration of detergents used to destroy whole virus particles exceeds acceptable limits in vaccines and must be removed to permissible levels.

[0006] The removal of detergents results in the formation of aggregates or clusters of the resulting lysed viral particles. The occurrence of these aggregates is related to the strain and the level / type of detergent used during the lysis process. In many cases, vaccines and other pharmaceuticals contain residual detergents or other detergent agents / chemicals to maintain appropriate quality properties. However, it has long been established that the presence of detergents, particularly in vaccines, makes antigens more soluble, thereby leading to decreased immunogenicity and thus reducing vaccine effectiveness.

[0007] Brief

[0008] This disclosure provides a method for dispersing aggregates in a formulation containing influenza proteins or viruses, the method comprising sonicating the formulation.

[0009] This disclosure also provides a method for producing an influenza vaccine, the method comprising producing a formulation containing inactivated or lysed influenza virus particles and subjecting the formulation to ultrasonic treatment. Attached Figure Description

[0010] Figure 1Reproducibility of optical density turbidity (ODT) assays of IVV vaccine matrix for influenza virus strains A / Victoria / 361 / 2011 (H3N2), A / California / 07 / 2009 (H1N1), and B / Hubei-Wujiagang / 158 / 158 / 2009 (B Yamagata).

[0011] Figure 2 ODT analysis of the effects of sonication on the dispersion of the H3N2 A / Victoria / 361 / 2011IVV drug matrix in terms of energy (joules / mL) and heat (37°C, 30 min).

[0012] Figure 3 ODT analysis of the effect of ultrasonic treatment on the dispersion of H3N2A / Victoria / 361 / 2011IVV drug matrix in terms of energy input intensity (amplitude).

[0013] Figure 4 Compared with untreated and PS80-treated samples, ODT results demonstrated that sonication effectively dispersed aggregates and maintained the dispersion of the H3N2 A / Victoria / 210 / 2009IVV drug matrix over time.

[0014] Figure 5 The average particle size distribution intensity (PSDs) (n=5) of the IVV drug matrix of H3N2A / Victoria / 210 / 2009 (untreated, PS80 treated, and sonicated) was analyzed by DLS over 24 weeks.

[0015] Figure 6 One-way radioimmunodiffusion (SRID) analysis was performed on untreated, detergent-treated (PS80) and sonicated IVV drug matrices (H3N2; A / Victoria / 210 / 2009) within 24 weeks.

[0016] Figure 7 : EM micrographs of IVV drug matrix (MPH), sonicated IVV drug matrix (sonicated MPH), and IVV drug matrix in the presence of polysorbate 80 (MPH+PS80). All samples were analyzed by EM at 0, 1, 2, and 6 months.

[0017] Figure 8ODT results of four seasonal influenza strains (A / Victoria / 361 / 2011(H3N2), A / California / 7 / 2009(H1N1), B / Hubei-Wujiagang / 158 / 2009(B Yamagata), and B / Brisbane / 60 / 2008(BVictoria)) before and after ultrasound (Son) treatment at 0, 3, and 6 months.

[0018] Figure 9 Mean particle size distribution (PSD) of untreated (left) and sonicated (right) IVV drug matrix materials analyzed by DLS over six months: A / Victoria / 361 / 2011 (A, B), A / California / 7 / 2009 (C, D), B / Hubei-Wujiagang / 158 / 2009 (E, F) and B / Brisbane / 60 / 2008 (G, H) (n=5).

[0019] Figure 10 For various IVV drug matrix batch volumes of 60, 500, and 1000 ml, the ultrasonic exposure time (minutes) corresponding to an ODT ≥ 80% for an IVV drug matrix batch volume (MPH volume mL).

[0020] Figure 11 For batches of various IVV drug matrices of 60, 500 and 1000 ml, the ultrasonic energy input corresponds to ODT ≥ 80% within the processing time.

[0021] Figure 12 The ultrasonic energy input required to achieve ODT ≥ 80% relative to the batch volume (60, 500, and 1000 ml) of the IVV drug matrix using a flow-through device.

[0022] Figure 13 Linear relationship between the amount of aggregation in the IVV drug matrix (expressed as %ODT) and the number of predicted glycosylation sites on the HA molecule. Detailed Implementation

[0023] This disclosure describes the efficient dispersion of aggregated material in influenza virus vaccine (IVV) pharmaceutical materials or IVV pharmaceutical products using ultrasound, and hereinafter referred to as IVV pharmaceutical matrix. The feasibility of ultrasound treatment was evaluated on the H3N2 strain of influenza virus, as this influenza A substrain exhibited the highest level of aggregation compared to non-H3N2 strains and influenza B virus.

[0024] Since inhibition of protein aggregation is typically achieved by adding compatible excipients to formulations, one example of this disclosure provides a previously unseen method. For example, excipients such as sugars, polyols, amino acids, salts, polymers, and surfactants have been found to stabilize aggregates through preferential interactions [(Arakawa et al. (1991); Timasheff (1998)], increase protein folding rates [Wang et al. (1995); Frye and Royer (1997)], reduce solvent accessibility and conformational migration [Kendrick et al. (1997)], and increase solvent viscosity [Jacob and Schmid (1999)].

[0025] This disclosure avoids the need for these additives, which could adversely affect the immunogenicity of the vaccine and cause undesirable side effects in individuals receiving the vaccine.

[0026] In one embodiment, this disclosure provides a method for dispersing aggregates in a formulation containing influenza proteins, the method comprising sonicating the formulation.

[0027] In another embodiment, this disclosure also provides a method for producing an influenza vaccine, the method comprising producing a formulation containing inactivated or lysed influenza virus particles and subjecting the formulation to ultrasonic treatment.

[0028] The formulation processed by sonication can be an intact virus particle, a lysed virus particle, a subunit vaccine, or a recombinant vaccine. To facilitate filtration for formulation sterilization, it is preferable that the percentage of agglomerates in the final formulation is less than 10%.

[0029] In one embodiment, the formulation comprises influenza hemagglutinin, for example, the formulation comprises lysed influenza virus particles. In one embodiment, the formulation is substantially free of detergents. As used herein, the phrase "substantially free of detergents" means having a level of less than 0.02%. For example, the detergent content is less than 200 ppm. In other examples, the detergent content is less than 50 ppm.

[0030] Typically, sonication is performed at a certain intensity for a period of time to disperse at least 50% of the aggregates present in the formulation. The ultrasonic energy generated to disrupt the aggregates can be transferred to the target IVV drug matrix in one of three ways: (1) directly through an ultrasonic generator probe suspended in the IVV drug matrix; (2) through the vibrating tip of the ultrasonic generator probe enclosed in a flow-through device; or (3) indirectly through a metal or glass tube through which the IVV drug matrix passes. All sonication methods require the transfer of at least 89 joules / mL of energy to disperse at least 50% of the aggregates in the IVV drug matrix.

[0031] The vaccines produced using the methods of this disclosure can be monovalent seasonal vaccines or monovalent pandemic vaccines. In another embodiment, the vaccines of this disclosure are multivalent vaccines, such as trivalent and quadrivalent vaccines.

[0032] Vaccines produced by the methods of this disclosure typically contain influenza A and influenza B antigens and are, for example, substantially free of agglomerative material. In one embodiment, the phrase "substantially free of agglomerative material" means that more than 50% (about 90 joules / mL sonication) of the material does not agglomerate. In another embodiment, at least 60% (about 134 joules / mL sonication), 70% (about 178 joules / mL sonication), or 80% (about 223 joules / mL sonication) of the material does not agglomerate. In yet another embodiment, at least 90% (about 267 joules / mL sonication) of the material does not agglomerate.

[0033] As described below, the method of this disclosure offers several unexpected advantages. First, it is highly effective in dispersing aggregates present in influenza antigen preparations. Surprisingly, these dispersed aggregates do not re-aggregate even after prolonged storage (4°C). Furthermore, it has been shown that vaccines produced using the method of this disclosure can elicit a stronger immune response in ferret models than vaccines containing the same antigen but untreated or treated with additives to disperse aggregates.

[0034] Throughout this specification, unless the context requires otherwise, the word “comprising” or its variations such as “including” or “containing” shall be understood to imply inclusion of the stated element or integer or group of elements or integers, but not to exclude any other element or integer or group of elements or integers.

[0035] References to any prior publications (or information derived therefrom) or any known things in this specification are not, and should not be construed as, an admission or endorsement, or in any way as suggesting that prior publications (or information derived therefrom) or any known things form part of the common general knowledge in the field to which this specification pertains.

[0036] All publications mentioned in this specification are incorporated herein by reference in their entirety.

[0037] It must be noted that, unless otherwise specified, the singular forms “an,” “a,” and “the” used in this specification include plural meanings. Therefore, for example, a reference to “an agent” includes a single agent as well as two or more agents; a reference to “an molecule” includes a single molecule as well as two or more molecules, etc.

[0038] Example

[0039] method

[0040] Ultrasonic treatment of IVV drug matrix

[0041] Direct probe ultrasound method

[0042] IVV drug matrix is ​​treated using a “direct” sonication method, in which the horn / probe of the sonicator is directly immersed in the sample in a beaker.

[0043] Ultrasonic processing is performed using a Branson 450 ultrasonic generator (Branson Ultrasonics), which consists of four components, including a power unit, a 102C transducer, and a 0.5-inch tapped horn.

[0044] To evaluate the efficacy of sonication, a Bandelin SONOPULS ultrasonic generator (Bandelin Electronic.KG) was used. The device included a GM3200 ultrasonic generator power unit, a UW3200 converter, an SH213G booster, and a TT13 13mm titanium tip. The IVV drug substance was sonicated in a beaker according to the method described for the Bandelin ultrasonic generator.

[0045] IVV drug matrix samples were prepared in 10 mL batches and placed in clean 30 mL beakers, which were fixed in a fixture. The ultrasonic generator probe was fully immersed in the sample solution (i.e., the distance between the ultrasonic generator tip and the bottom of the beaker was approximately 1 mm). Ultrasonic treatment was then performed to transfer a specific range of energy input while varying the rate (amplitude) of energy transfer.

[0046] The range of ultrasound energy delivered is measured in joules (energy) per milliliter of IVV drug matrix, including: 0 joules / mL, 83 joules / mL, 165 joules / mL, 259 joules / mL, and 345 joules / mL.

[0047] To avoid overheating the sample, it should be sonicated in at least two parts, with the beaker briefly cooled on ice between the two treatments. The final sonicated IVV drug sample should be transferred to a plastic tube and stored at 2–8°C for further analysis.

[0048] Flow-through ultrasonic treatment

[0049] To evaluate the scalability of sonication, a Bandelin SONOPULS ultrasonic generator (Bandelin Electronic, KG) equipped with a flow-through device was used. This device includes a GM3200 ultrasonic generator power unit, a UW3200 converter, an SH213G booster, a TT13 13mm titanium tip, and a DG 4G flow-through treatment vessel. IVV drug material was circulated and ultrasonically treated via the ultrasonic generator's flow-through device using a 520U peristaltic pump (Watson & Marlow, Australia).

[0050] Determination of aggregated substances

[0051] Optical density turbidity (ODT) measurement

[0052] To assess the extent of non-aggregated material in the IVV pharmaceutical intermediate, the recovered protein level was determined by optical density (OD) at 280 nm A in the supernatant after applying moderate centrifugation (Tay et al., Investigation into alternative testing methodologies for characterization of influenza vaccine, Human Vaccine Immunotherapy 2015 11(7)1673-84). Since the proportion of protein in the precipitate after centrifugation is directly related to the degree of aggregation in the sample, a higher recovery rate in the supernatant corresponds to a larger proportion of dispersed protein. This assay is called optical density turbidity (ODT). The protein recovery value ranges from 0 to 100% (expressed as %ODT), and this value increases with increasing protein dispersion in the sample.

[0053] Several properties of the assay were evaluated for three influenza strains: H1N1 (A / California / 07 / 2009), H3N2 (A / Victoria / 361 / 2011), and B (B / Hubei Wujiagang / 158 / 158 / 2009). Validation results showed repeatability (%CV) of 2.9, 3.1, and 3.1%, respectively; batch-to-batch precision variation (%CV) of 8.8, 6.5, and 3.4%, respectively; intermediate precision of 0.4%, 6.3%, and 0.2%, respectively; no statistically significant differences between operators (p-values ​​of 0.81, 0.13, and 0.78), and predictable linearity between expected and observed results (R = 0.9685). Figure 1 The document details typical %ODT curves for repeated batches of IVV drugs representing seasonal vaccine subtypes H1N1, H3N2, and influenza B.

[0054] Furthermore, this assay is relevant to alternative methods for characterizing aggregation, including dynamic light scattering (DLS) and asymmetric field flow hierarchical separation (A4F). The low variability in the ODT analysis suggests that this assay is well-suited for assessing aggregation in intermediate vaccine materials.

[0055] Dynamic light scattering (DLS)

[0056] Particle size analysis via DLS is used as a complementary method to ODT experiments to further understand the aggregation characteristics of samples. DLS is based on the measurement of Brownian motion of proteins in solution, which is the random motion caused by collisions with surrounding solvent molecules. Brownian motion causes time-dependent fluctuations in the intensity of scattered light, which can be measured by DLS to produce the particle size distribution (PSD) of the sample.

[0057] DLS measurements were performed using a Malvern Zetasizer Nano Series ZS (Malvern Instruments Ltd.). For each sample analyzed, several properties affecting the Brownian motion of particles in solution were pre-determined, including density (DA-100M densitometer, Mettler Toledo), viscosity (Lovis 2000M microvisometer, Anton Paar), and refractive index (30GS refractometer, Mettler Toledo). Sample pretreatment involved centrifugation at 8000 rpm for 1 minute to remove any precipitates or foreign matter containing unstable large particles that could interfere with subsequent analyses. The supernatant composition obtained from each sample was then evaluated by DLS; each sample measurement was based on five replicates (n=5) performed at a backscattering angle of 173° and equilibrated at 25°C for three minutes.

[0058] Influenza antigenicity assessment

[0059] One-way radiation-induced immunodiffusion (SRID)

[0060] SRID analysis was performed as previously described [Williams et al., 1980]. Briefly, the reference and test antigens were diluted in 1% zwitterionic solution containing PBS (Calbiochem, Darmstadt, Germany) at ratios of 1:1, 2:3, and 1:3, and added in duplicate to the wells of an agarose gel containing polyclonal antiserum. The gels were incubated in a humidified chamber for 72 hours, dried on glass plates, and stained with Coomassie Brilliant Blue R-250 (Sigma, California, USA). Circular regions of antigen-antibody precipitation were measured, and HA concentrations were calculated using a parallel-line bioassay compared to IZP standards (15), with the validity of the test confirmed by the "g" test (g ≤ 0.061) (16).

[0061] Electron microscopy (EM) imaging

[0062] Negative staining EM was performed using the agar diffusion filtration method employed by Hayat and Miller (1990). Three grids were prepared for each sample; the IVV drug was diluted 1:50 in phosphate-buffered saline (PBS; pH 7.2) to provide a discontinuous monolayer. A sample (1 μL) was applied to a copper electron microscope grid coated with polymethyl methacrylate (Formvar) and then inverted onto a 2% w / v agar plate. After the grid had settled onto the agar plate (i.e., the liquid was absorbed by the agar), the grid was floated on a drop of negative dye (2% w / v sodium tungstate phosphate, pH 7.0). After twenty seconds, the grid was lifted, and excess dye was removed by bringing the grid edge into contact with a small strip of torn Whatman No. 1 filter paper. Any remaining dye film was air-dried before examination under an electron microscope.

[0063] result

[0064] Destruction of aggregated IVV drug matrix by ultrasonic treatment

[0065] The most effective method for dispersing aggregates was determined using the H3N2 influenza substrain, as it tends to form the largest amount of aggregates after detergent degradation. Several methods were evaluated to determine whether aggregated material could be dispersed after detergent degradation.

[0066] The first method involves dispersing aggregates in an IVV drug material of subtype A / Victoria / 361 / 2011 H3N2 using high-frequency acoustic waves through ultrasonic treatment. Figure 2Direct sonication allows for the direct transfer of localized, high-intensity ultrasonic energy from a probe into a beaker containing the sample. Ultrasonic input energy ranging from 83 J / mL to 345 J / mL was applied to samples (prepared in 10 mL batches), and the degree of dispersion was assessed by ODT analysis. Results showed a linear relationship between energy input and the level of dissociated aggregates; the %ODT value after sonication was between 50% and 80%, while the untreated value was 40%. Figure 2 Interestingly, the dispersion level reached its maximum at 259 joules / mL, from which no further increase in dispersion was observed. In contrast, mild heating and stirring of the IVV drug (at 37°C and 600 rpm for 30 minutes) did not alter the level of dispersion.

[0067] The intensity of energy transfer to the IVV drug was investigated by adjusting the ultrasonic amplitude (0–100%), while the exposure time for all samples was kept constant at 0.70 s / ml. As previously observed in the energy transfer amount, the ODT results showed a predictable trend that the dispersion of the IVV drug increased with increasing ultrasonic amplitude when the exposure time was kept constant. Figure 3 The optimal rate of ultrasonic energy transfer was achieved at 80% amplitude; beyond this rate, the dissociation level of the aggregates no longer increased. These results indicate that 80% is the optimal amplitude (i.e., energy transfer rate) for dissociating aggregates.

[0068] Feasibility of Ultrasonic Dispersion of IVV Drug Matrix Materials

[0069] Achieving acceptable inactivated vaccines requires quality stability. That is, if a method has been shown to disperse aggregated substances, it is important that the substances retain this property.

[0070] A 24-week study evaluated the feasibility of ultrasound as a means of dispersing agglomerated material in an IVV drug matrix of H3N2 A / Victoria / 210 / 2009, with or without the detergent polysorbate 80 (PS80). Multiple assays were performed to monitor various characteristics of the samples, including ODT and DLS for agglomeration assessment, SRID for antigenicity, and electron microscopy for morphological imaging.

[0071] ODT and DLS reunification behavior

[0072] Aggregation characteristics of untreated, sonicated, and PS80-treated IVV drug matrices were assessed using ODT and DLS assays. ODT analysis showed that sonication (at least 200 joules / mL) resulted in 80% dispersion of the IVV drug matrix, compared to 40% for the untreated matrix (at time 0). Figure 4This level of dispersion persisted for 24 weeks at 4°C, indicating an irreversible dispersion state. Figure 4 Furthermore, during this period, there was no increase in dispersion or further aggregation of the control (untreated) material. The addition of detergent (0.1% PS80) had no significant effect on the aggregate level in this material compared to the initial level. Therefore, this indicates that the aggregation level of the IVV drug matrix material is determined after detergent breakdown.

[0073] Combined with ODT analysis, DLS analysis was used to further characterize the aggregation of samples. For each sample, DLS measurements (n=5) produced the intensity-size distribution (PSD), which shows the relative intensity of scattered light within different particle size groups. The DLS results showed a strong correlation with the ODT results for all three types of samples. Figure 5 For example, untreated and PS80-treated IVV drug matrix samples showed multi-peak PSD with peaks at 60, 400, and 7000 nm, indicating the presence of aggregates. Figure 5 A and B). However, ultrasonic treatment produced a single-peak distribution with a distinct peak at 300 nm, indicating that the sample was uniform and well-dispersed, without agglomerates. Figure 5 C). At each time point of analysis, all samples produced reproducible PSDs that remained unchanged over 24 weeks.

[0074] Antigenicity analysis using SRID and EIA

[0075] SRID was used to assess antigenic material levels to determine whether sonication or the addition of detergents affected influenza antigens. SRID analysis showed that regardless of whether the material was sonicated or treated in the presence of detergents, the levels of antigenic material remained at the same potency level as untreated samples and remained unchanged over time (Table 1). Figure 6 ).

[0076] Table 1: SRID results of untreated, detergent-treated (PS80) and sonicated IVV drug matrix (H3N2; A / Victoria / 210 / 2009) over time.

[0077]

[0078] Morphological imaging of EM

[0079] EM imaging was used to examine the morphological appearance of samples at 0, 1, 2, and 6 months. Figure 7Significant differences were found between sonicated and control IVV drug matrix samples. Control / untreated IVV drug matrices (with and without PS80) contained substantial aggregates throughout the 6-month period, as shown by the darker areas in the micrographs. In contrast, sonicated IVV drug matrices contained fewer aggregates, which were smaller, and the appearance of the material remained consistent throughout the examined time period. These observations reflect the results of ODT and DLS analyses, where the sonicated material was significantly more dispersed than untreated or detergent-containing IVV drug matrices (with PS80).

[0080] The applicability of ultrasound to all seasonal influenza virus strains

[0081] Compared to other seasonal strains, H3N2 exhibits the greatest degree of aggregation ( Figure 1 This method must demonstrate its ability to disrupt aggregates of all strains. The levels of dispersed aggregates in four seasonal strains before and after ultrasonic treatment (at least 200 joules / mL) over a six-month period must be measured. Figure 8 Of all the viral preparations examined, the application of ultrasound increased the level of dispersed matter, which remained dispersed for six months. The increase in aggregate dispersion of the two influenza A virus substrains, H3N2 and H1N1, after ultrasound treatment was more pronounced compared to the two influenza B viruses from the Yamagata and Victoria strains. For example, the level of dispersed matter increased by approximately 60-100% for the two A strains, while the level of dispersed matter increased by 3-10% for the B strain.

[0082] Data obtained from DLS analysis of all four influenza strains further confirmed the ODT results. Figure 8 The intensity PSD of each sample is shown from five repeated measurements at 0, 3, and 6 months. The untreated sample from A / Victoria / 361 / 2011 is shown in its multimodal profile. Figure 9 A) describes the presence of aggregates in groups of various sizes, while the PSDs of A / California / 7 / 2009, B / Hubei-Wujiagang / 158 / 2009, and B / Brisbane / 60 / 2008 exhibit more unimodal structures, thus indicating a more homogeneous particle population (respectively). Figure 9 C, E, and G). After ultrasonic treatment, all four strains exhibited the characteristics of a well-dispersed IVV drug matrix, free of any aggregates. Figure 9 B, D, F, H).

[0083] Stability and batch consistency of ultrasound-treated IVV drug matrix

[0084] IVV drug matrices representing four vaccine candidate types / subtypes were sonicated to determine the consistency and stability of applying controlled energy levels (joules / mL) to achieve target levels of aggregate dispersion (Table 2). Figure 8 Each representative strain of IVV drug was divided into six aliquots. Three of the six aliquots were individually exposed to sonication at a minimum of 200 joules / mL and stored at 2–8°C for up to six months along with their unsonicated control groups. Samples were taken from all groups at 0, 1, 3, and 6 months, and the levels of aggregates present were analyzed by ODT. Significant changes in the levels of dispersed aggregates were observed in all sub-batches of influenza A virus subtypes A / California / 07 / 2009 and A / Victoria / 361 / 2011 compared to the unsonicated control groups. For the two representative influenza A virus strains, there was high consistency between the independently sonicated sub-batches. At 0 months, all three sub-batches of the H1N1 and H3N2 strains reached the target ODT% (>80%) after sonication, with CV% of 2.1% and 1.1%, respectively, within the 10% limit. Due to the low aggregate content in the control group, the observed variation in aggregate dispersion levels in batches representing strain B was small. Following ultrasonic treatment, the sub-batches of strain B showed the same level of inter-batch consistency as those representing strain A, with CV% of 1.2% and 0.7% for B / Hubei Wujiagang / 158 / 2009 and B / Brisbane / 60 / 2008 at 0. Importantly, the sub-batches representing all four seasonal influenza virus strains maintained their high levels of %ODT and inter-batch consistency throughout the 6-month period, indicating that the aggregate disruption caused by ultrasonic treatment is permanent.

[0085] Table 2: IVV drug matrix batches (n=3) representing four seasonal influenza strains before and after ultrasound: A / California / 07 / 2009, A / Victoria / 361 / 2011, B / Hubei Wujiagang / 158 / 2009, and B / Brisbane / 60 / 2008, over a 6-month period.

[0086]

[0087] A linear relationship between the number of glycosylation sites predicted on HA and the required sonication amount.

[0088] Envelope glycoprotein; hemagglutinin (HA) is a sialic acid receptor-binding protein of influenza viruses, anchoring them to host cells and allowing them to evade digestion once phagocytosed into endosomes. The globular head region of the HA molecule contains N-linked glycosylation sites that overlap with antigenic sites and are thought to be involved in protecting these antigenic sites from binding to antibodies and major histocompatibility complexes (Skehel et al., 1984; Jackson et al., 1994). Furthermore, the structural complexity of the N-glycan is positively correlated with HA receptor binding specificity (Tsuchia et al., 2002). During the evolution of H1N1 and H3N2 human influenza A viruses, the number of N-linked glycosylation sites in the HA globular head region increased (Suzuki, 2011). We have determined the relationship between the predicted number of glycosylation sites on the HA molecule of influenza A and the level of aggregation present. Using algorithms available on the NetNGlyc 1.0 server, the number of glycosylation sites was predicted by calculating probability scores. http: / / www.cbs.dtu.dk / services / NetNGlyc / To generate a probability score, the HA protein sequence of the relevant strain is input into the algorithm's panel and submitted for analysis. The software generates a list of predicted glycosylation sites within the input sequence, scored using 1-3 plus signs (+) based on probability strength. The probability score for predicted HA glycosylation sites (pGly score) is defined as the sum of the plus signs for a given output sequence. We recommend that H3N2 strains with a pGly score ≥16 require sonication at ≥90 J / mL, and H1N1 strains with a pGly score ≥11 require sonication at ≥90 J / mL (where more than 50% of the material does not aggregate).

[0089] Table 3: Relationship between the number of predicted glycosylation sites on the HA1 molecule of influenza A and the level of aggregation present.

[0090]

[0091] Alternative physical destruction methods for dispersing IVV drugs

[0092] To evaluate the unique ability of ultrasonication to disperse IVV drug matrices, other physical disruption methods were also investigated. These included local heating (microwave for 1 and 10 seconds) and shear force (25 and 100 Dounce homogenizations). In both cases, there was no significant difference in dispersibility compared to untreated material (Table 4).

[0093] Table 4: ODT results of IVV drug substances after microwave heating and shear force Dounce homogenization (A / Victoria / 361 / 2011).

[0094]

[0095] in conclusion

[0096] Direct sonication has been found to be an effective method for dispersing aggregates within IVV drug matrices. The H3N2 subtype of influenza virus strains exhibited the highest levels of aggregation. Optimization of the process revealed that the energy and transfer rate to the IVV drug matrix are key to controlling aggregate dispersion levels. Increasing the sonication rate (amplitude) and / or exposure time (seconds) resulted in a linear increase in aggregate dissociation levels. A plateau in dispersion levels was observed after 97% ODT, after which little or no further aggregate dispersion (measured by ODT) was observed. The method of using sonication to disperse aggregates in IVV drug substances was evaluated over a 24-week (6-month) timeframe with and without detergent (PS80). Several characterization analyses, including ODT, DLS, and EM, showed a significant increase in the amount of dispersed material in the sonicated IVV drug matrix compared to untreated and detergent-treated samples. The amount of sonication required to achieve the target aggregate dispersion level can be predicted and is consistent across batches. Furthermore, the dispersion level remained consistent throughout the feasibility study, indicating a stable and permanent aggregate dispersion. Immunological evaluation using SRID confirmed that neither sonication nor detergent treatment impaired the antigenicity of the IVV drug matrix.

[0097] This work strongly demonstrates the value of sonication as a simple, practical, and effective method to improve the quality properties of influenza vaccines containing highly aggregated IVV drug matrices. Furthermore, this method has been shown to be applicable to all seasonal influenza virus strains. Following sonication, increased levels of dispersed material in the IVV drug matrices of H3N2, H1N1, and two influenza B virus strains (Yamagata and Victoria lines) were observed and maintained well over 6 months.

[0098] As an alternative to direct ultrasound methods designed for laboratory-scale studies, a continuous flow ultrasound configuration was investigated for processing commercial-scale IVV drug matrices. In short, the ultrasound apparatus was powered by a high-frequency generator and a 20 kHz transducer; a connected booster horn was housed within a flow-through processing vessel containing a sample that was continuously recirculated at a specified flow rate.

[0099] To evaluate the system's scalability, the effects of various process parameters, including ultrasonic intensity (amplitude) and product recirculation flow rate, were investigated to determine their impact on the dispersion efficiency of aggregates within the IVV drug substance. Using a constant recirculation flow rate of 120 ml / min and a fixed amplitude of 80%, a strong linear correlation (R0.05) was demonstrated between batch volume (60, 500, and 1000 ml) and the ultrasonic time required to reach 80% of the ODT threshold (over 60 minutes). 2=0.989). In ultrasonic energy (joules) and time (R... 2 This trend was also observed between ≥0.998 and, consequently, between ultrasonic energy and IVV drug matrix batch volume (R). 2 =0.981, Table 5 and Figure 10-12 These data demonstrate that the sonication process is scalable in terms of IVV drug batch size. Furthermore, the data show that a fixed energy input of at least 300 joules / mL is sufficient to dissociate aggregates to ≥80% of the ODT level, regardless of the volume within the system.

[0100] Table 5: Ultrasonic time and energy input required for IVV drug matrix to achieve ODT≥80% during recirculation through the ultrasound machine at different batch volumes (80% amplitude; IVV drug matrix flow rate of 120 ml / min).

[0101] IVV drug matrix Ultrasound time (seconds) Energy input Energy / ml ODT (ml) (Second) (joule) (joules / mL) (％) 60 150 18272 305 83 500 660 89354 179 82 1000 1500 222953 223 83

[0102] We have determined a linear relationship between the number of predicted glycosylation sites on the influenza A HA molecule and the degree of aggregation found in the IVV drug matrix of this strain. Figure 13 The correlation coefficient (r-value) between the predicted glycosylation sites and IVV drug matrix aggregation degree of 12 H3N2 strains produced between 2005 and 2017 was calculated. The r-value was 0.74, indicating a moderate correlation between the two properties. We recommend that H3N2 strains with a pGly score ≥16 require sonication at a rate of ≥90 joules / mL, and H1N1 strains with a pGly score ≥11 require sonication at a rate of ≥90 joules / mL.

[0103] References

[0104] [1] CDC. Seasonal Influenza Vaccine Safety: A Summary for Clinicians. 2011 [Cited; available at: http: / / www.cdc.gov / flu / professionals / vaccination / vaccine_safety.htm]

[0105] [2] Fuminger IGS. Vaccine production. In the following books: Nicholson KG, Webster RG, Hay AJ eds. Textbook of influenza Oxford: Blackwell Science, 1998: 324-32.

[0106] Smith TL, Jennings R., Specificity and in vitro transfer of the immunosuppressive effect of detergent-disrupted influenza virus vaccine. Clin Exp Immunol. 1990 Jan; 79(1):87-94.

[0107] Williams MS, Mayner RE, Daniel NJ, Phelan MA, Rastogi SC, Bozeman FM, et al. New developments in the measurement of the hemagglutinin content of influenza virus vaccines by single-radial-immunodiffusion. J Biol Stand. 1980; 8(4):289-96

[0108] van de Donk HJ, de Jong JC, van Olderen MF, Osterhaus AD. Monoclonal antibodies for the control of influenza virus vaccines. Developments in biological standardization. 1984; 57:251-5

[0109] Bhatti A, Siddiqui, YM., Micusan, VV. Highly sensitive fluorogenic enzyme-linked immunosorbent assay: detection of staphylococcal enterotoxin B1. Journal of Microbiological Methods. 1994; 19:179-87

[0110] Skehel, JJ, Stevens, DJ, Daniels, RS, Douglas, AR, Knossow, M., Wilson, IA, and Wiley, DC (1984). A carbohydrate sidechain on hemagglutinins of Hong Kong influenza viruses inhibits recognition by a monoclonal antibody. Annu. Rev. Biochem. 69, 531-569.

[0111] Jackson, DC, Drummer, HE, Urge, L., Otvos, L. Jr., and Brown, LE (1994) Glycosylation of a synthetic peptide representing a T cell determinant of influenza virus hemagglutinin results in loss of recognition by CD4+ T-cell clones. Virology 199, 422-430.

[0112] Tsuchiya, E., Sugawara, K., Hongo, S., Matsuzaki, Y., Muraki, Y., Li, Z.-N., and Nakamura, K. (2002) Effect of addition of new oligosaccharide chains to the globular head of influenza A / H2N2 virus hemagglutinin on the intracellular transport and biological activities of the molecule. J. Gen. Virol. 83, 1137-1146.

[0113] Suzuki, Y. (2011) Positive selection for gains of N-linked glycosylationsites in hemagglutinin during evolution of H3N2 human influenza A virus. Genes Genet. Syst. 86, 287-294.

Claims

1. A method for dispersing aggregates in a formulation containing influenza virus proteins, the method comprising sonicating the formulation, wherein: a) The ultrasonic treatment transfers at least 90 joules / ml of energy. b) Ultrasound is performed at a rate of at least 40% amplitude; and c) the energy transfer rate is 0.70 s / ml, wherein the formulation comprises a protein of influenza virus H3N2 strain having a predicted HA glycosylation site probability score (pGly score) ≥16, or comprises a protein of influenza virus H1N1 strain having a pGly score ≥11.

2. The method according to claim 1, wherein, The formulation contains influenza hemagglutinin.

3. The method of claim 1 or 2, wherein the formulation comprises lysed viral particles.

4. The method of claim 3, wherein the content of the detergent is less than 200 ppm.

5. The method of claim 1 or 2, wherein the ultrasonic treatment is performed at a certain intensity for a period of time to disperse at least 50% of the agglomerates present in the formulation.

6. The method of claim 1 or 2, wherein the ultrasound is performed at a rate of 80% amplitude.

7. A method for producing an influenza vaccine, the method comprising producing a formulation containing inactivated or lysed influenza virus particles and sonicating the formulation, wherein: a) The ultrasonic treatment transfers at least 90 joules / ml of energy; b) Ultrasound is performed at a rate of at least 40% of amplitude; and, c) The energy transfer rate is 0.70 s / ml, wherein the formulation contains a protein of influenza virus H3N2 strain with a predicted HA glycosylation site probability score (pGly score) ≥16, or contains a protein of influenza virus H1N1 strain with a pGly score ≥11.

8. The method of claim 7, wherein the vaccine comprises at least three different influenza virus strains.

9. The method of claim 7 or 8, wherein the vaccine is a monovalent vaccine.

10. The method of claim 7 or 8, wherein the vaccine is a quadrivalent vaccine.

11. The method of any one of claims 7 or 8, wherein the vaccine further comprises an influenza B vaccine.

12. The method of claim 7 or 8, wherein more than 50% of the substance does not agglomerate.

13. The method of claim 7 or 8, wherein the ultrasound is performed at a rate of 80% amplitude.

Images