Controlled ice nucleation lyophilization method for bispecific molecules

Abstract

The present disclosure provides a method of preparing a lyophilized bispecific molecule composition comprising (a) inducing ice nucleation in a liquid protein composition having a bispecific molecule concentration of at least about 10 mg / ml in a vial exposed to a first temperature of about-18 DEG C to about-10 DEG C for about 60 minutes to about 270 minutes; (b) exposing the vial of (a) to a second temperature of about-25 DEG C to-50 DEG C for about 1 hour to about 5 hours; and (c) drying the composition of (b) at a third temperature of about-5 DEG C to about-25 DEG C for about 25 hours to about 70 hours, and (d) further drying the composition of (c) at a fourth temperature of about 25 DEG C to about 50 DEG C for about 4 hours to about 12 hours to obtain a vial comprising a lyophilized bispecific molecular composition, the composition has a percentage content of high molecular weight material (HMWS) of less than or equal to about 1.5% (mA / ). The disclosure further provides lyophilized bispecific molecule compositions prepared via the methods described herein.

Description

Technical Field

[0001] The present disclosure relates to a method of lyophilizing a composition comprising a bispecific molecule and the resulting lyophilized composition. Background Art

[0002] Bispecific T cell engager molecules (e.g. Molecules) have demonstrated clinical benefit in immuno-oncology. Typically, one binding domain of these molecules is specific for a selected tumor-associated surface antigen on the target cell, and the second binding domain is specific for CD3, a subunit of the T cell receptor complex on T cells. Through their specific design, The molecules are uniquely suited to transiently connect T cells to target cells and, at the same time, strongly activate the inherent cytolytic potential of T cells against target cells. Prior to storage, bispecific T cell engager molecules are typically lyophilized. Lyophilization is a widely used manufacturing method for improving the stability of pharmaceutical products by removing the water content of liquid products through freeze drying. Lyophilization is an intermittent process that is typically long and accounts for more than 50% of the pharmaceutical product processing time. In addition, manufacturing-scale freeze dryers are also limited in capacity; only a certain number of vials can be loaded in a batch, and the amount of liquid composition that can be dispensed into the vials is limited. Due to these limitations, lyophilized products are typically expensive. See, for example, Awotwe-Otoo et al., International Journal of Pharmaceutics, 450 (2013), 70–78; Esfandiary et al., J. Pharm. Sci., 105 (2016), 1427-1433. Efforts to improve the efficiency of lyophilization processes for biopharmaceutical products, such as bispecific T-cell engager molecules, can be hampered by inconsistent product attributes, such as excessive formation of high-molecular-weight species (HMWS), especially at higher product concentrations, which can lead to costly product losses.

[0003] The freezing step in conventional lyophilization methods involves uncontrolled or random ice nucleation that occurs at temperatures significantly below the normal freezing point. Therefore, such "supercooling" can produce small ice crystals and require very long drying times. One way to reduce the total lyophilization process time is to add an annealing step to the lyophilization process, which increases the average size of the ice crystals through a process called Ostwald ripening. However, as the present inventors have discovered, such an annealing step can lead to aggregation, i.e., the formation of high molecular weight species (HMWS), typically in high concentration bispecific molecule formulations (e.g., at a concentration of at least 5 mg / ml or even at least 10 mg / ml), which may be due to low temperature degradation mechanisms. Therefore, there is a need for an improved lyophilization method that can provide a bispecific T cell engaging molecule product with a low aggregation product rate even when starting from a higher concentration formulation. Summary of the Invention

[0004] In view of the above unmet needs, it is an object of the present invention to provide a resource-saving method for providing bispecific molecules, preferably bispecific (T cell engaging) molecules comprising an Fc domain with an extended half-life, at a concentration above 10 mg / ml, characterized by having only low levels (e.g., less than or equal to 1.5% of the total molecules) of undesirable aggregates (e.g., HMWS). This problem is solved by introducing a controlled ice nucleation (CIN) step in the lyophilization process to lyophilize high concentrations of bispecific molecules with an Fc domain, thereby reducing the overall process time compared to standard lyophilization methods. The reduction in process time reduces energy consumption and frees up the equipment for the next run more quickly. In a first aspect, it is envisaged in the context of the present invention to provide a method for preparing a lyophilized bispecific molecule composition, the method comprising

[0005] (a) (CIN) step: inducing ice nucleation in a liquid bispecific molecule composition having a bispecific molecule concentration of at least about 10 mg / ml in a vial exposed to a first temperature of about -10°C to about -18°C for about 60 minutes to about 270 minutes, wherein this step preferably includes a post-nucleation hold for up to 90 minutes, preferably about 30 minutes,

[0006] (b) a freezing step of exposing the vial to a second temperature of about -25°C to -50°C for about 1 to 5 hours;

[0007] (c) a drying step of drying the composition of (b) at a third temperature of about -5°C to about -25°C for about 25 to 70 hours; and

[0008] (d) further drying the composition of (c) at a fourth temperature of about 25° C. to 50° C. for about 4 to 12 hours to provide a vial comprising a lyophilized bispecific molecule composition having a high molecular weight species (HMWS) percentage content of less than or equal to 1.5% (m / V), wherein the bispecific molecule comprises at least three domains, wherein:

[0009] The first domain binds to tumor antigens on target cells;

[0010] The second domain binds to an extracellular epitope of the human and / or macaque CD3ε chain; and

[0011] Preferably, the third domain that provides the extended half-life of the bispecific molecule is fused to the second domain via a peptide linker, the third domain comprising two polypeptide monomers, each polypeptide monomer comprising a hinge, a CH2 domain and a CH3 domain, wherein the two polypeptide monomers are fused to each other via a peptide linker.

[0012] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein step (a) comprises exposing the vial to a first temperature for about 90 minutes to about three hours (CIN step).

[0013] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein step (a) comprises exposing the vial to the first temperature for about 90 minutes to about two hours, preferably 110 minutes.

[0014] Within said aspects, it is also envisaged in the context of the present invention to provide a process wherein the first temperature in step (a) is from about -12°C to about -17°C.

[0015] Within said aspects, it is also envisaged in the context of the present invention to provide a process wherein the first temperature in step (a) is about -15°C.

[0016] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein step (a) further comprises maintaining the ice-nucleated bispecific molecule composition at that temperature for a post-nucleation period of up to 90 minutes.

[0017] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein the post-nucleation time period is from about 20 minutes to about 90 minutes.

[0018] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein the post-nucleation time period is about 30 minutes.

[0019] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein step (b) comprises exposing the vials to a second temperature of about -45°C (freezing step).

[0020] Within said aspects, it is also contemplated in the context of the present invention to provide a method wherein the second period of time is from about two hours to about four hours.

[0021] Within said aspect, it is also envisaged in the context of the present invention to provide a method as claimed in claim 10, wherein the second period of time is about three hours.

[0022] Within said aspects, it is also contemplated in the context of the present invention to provide a method wherein the first temperature is transitioned to the second temperature at a rate of about 0.01°C to about 0.5°C per minute.

[0023] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein the first temperature is transitioned to the second temperature at a rate of about 0.2°C per minute.

[0024] Within said aspects, it is also envisaged in the context of the present invention to provide a process wherein the third temperature of step (c) is from about 0°C to about -20°C (drying step), preferably at about 70 to 120 mTorr, preferably 100 mTorr.

[0025] Within said aspects, it is also envisaged in the context of the present invention to provide a process wherein the third temperature of step (c) is from about -5°C to about -10°C.

[0026] Within said aspects, it is also envisaged in the context of the present invention to provide a process wherein the third temperature of step (c) is about -8°C.

[0027] Within said aspects, it is also contemplated in the context of the present invention to provide a method wherein step (c) comprises exposing the vial containing the bispecific molecule composition to a temperature increase at a rate of about 0.01°C to about 0.5°C per minute.

[0028] Within said aspects, it is also contemplated in the context of the present invention to provide a method wherein the transition from step (b) to step (c) comprises increasing the temperature at a rate of about 0.2°C to about 0.7°C per minute and maintaining the vial at a temperature of about -40°C to about -30°C for about 15 minutes to about one hour.

[0029] Within said aspects, also contemplated in the context of the present invention is a method wherein the liquid bispecific molecule composition is exposed in step (a) to a first temperature of about -15°C for about 110 minutes plus about 30 minutes after nucleation, and in step (b) to a second temperature of about -45°C for about 3 hours; in step (c) the composition of (b) is dried at a third temperature of about -8°C for about 50 hours; and the composition of (c) is further dried at a fourth temperature of about 40°C for about 8 hours, wherein step (d) is preferably carried out at about 70 to 120 mTorr, preferably 100 mTorr.

[0030] Within said aspects, it is also envisaged in the context of the present invention to provide a method, wherein the bispecific molecule is a single chain molecule.

[0031] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein the bispecific molecule is present in the composition at a concentration of about 10 mg / ml - about 30 mg / ml, or about 20 mg / ml to about 30 mg / ml or, preferably about 15 mg / ml to about 25 mg / ml.

[0032] Within said aspects, it is also contemplated in the context of the present invention to provide a method, wherein the method results in a vial containing a lyophilized bispecific molecule composition that exhibits aggregation in terms of high molecular weight species (HMSW) formation of less than or equal to about 1.5% (m / w), or preferably less than or equal to about 1.2, 1, 0.75 or even 0.5% (relative to the total bispecific molecular weight).

[0033] Within said aspects, it is also envisaged in the context of the present invention to provide a method, wherein the third domain comprises, in amino to carboxyl order: hinge-CH2-CH3-linker-hinge-CH2-CH3.

[0034] Within said aspects, it is also envisaged in the context of the present invention to provide a method, wherein each of said polypeptide monomers of the third domain has an amino acid sequence that is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 249-256, or an amino acid sequence that is identical to a sequence selected from the group consisting of SEQ ID NOs: 249-256.

[0035] Within said aspects, it is also envisaged in the context of the present invention to provide a method, wherein the first domain binds to CD33, CDH19, MSLN, FLT3, BCMA, CD19, MUC17, CDH3, CLDN18.2, CD70, EGFRviii, EpCAM, DLL3 and / or PSMA, preferably DLL3.

[0036] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein the first binding domain of the construct comprises a VH region comprising CDR-H1, CDR-H2 and CDR-H3 selected from the group consisting of: and a VL region comprising CDR-L1, CDR-L2 and CDR-L3 selected from the group consisting of:

[0037] (a) CDR-H1 as depicted in SEQ ID NO: 4, CDR-H2 as depicted in SEQ ID NO: 5, CDR-H3 as depicted in SEQ ID NO: 6, CDR-L1 as depicted in SEQ ID NO: 1, CDR-L2 as depicted in SEQ ID NO: 2, and CDR-L3 as depicted in SEQ ID NO: 3,

[0038] (b) CDR-H1 as depicted in SEQ ID NO: 29, CDR-H2 as depicted in SEQ ID NO: 30, CDR-H3 as depicted in SEQ ID NO: 31, CDR-L1 as depicted in SEQ ID NO: 34, CDR-L2 as depicted in SEQ ID NO: 35, and CDR-L3 as depicted in SEQ ID NO: 36,

[0039] (c) CDR-H1 as depicted in SEQ ID NO:42, CDR-H2 as depicted in SEQ ID NO:43, CDR-H3 as depicted in SEQ ID NO:44, CDR-L1 as depicted in SEQ ID NO:45, CDR-L2 as depicted in SEQ ID NO:46, and CDR-L3 as depicted in SEQ ID NO:47,

[0040] (d) CDR-H1 as depicted in SEQ ID NO: 53, CDR-H2 as depicted in SEQ ID NO: 54, CDR-H3 as depicted in SEQ ID NO: 55, CDR-L1 as depicted in SEQ ID NO: 56, CDR-L2 as depicted in SEQ ID NO: 57, and CDR-L3 as depicted in SEQ ID NO: 58,

[0041] (e) CDR-H1 as depicted in SEQ ID NO: 65, CDR-H2 as depicted in SEQ ID NO: 66, CDR-H3 as depicted in SEQ ID NO: 67, CDR-L1 as depicted in SEQ ID NO: 68, CDR-L2 as depicted in SEQ ID NO: 69, and CDR-L3 as depicted in SEQ ID NO: 70,

[0042] (f) CDR-H1 as depicted in SEQ ID NO: 83, CDR-H2 as depicted in SEQ ID NO: 84, CDR-H3 as depicted in SEQ ID NO: 85, CDR-L1 as depicted in SEQ ID NO: 86, CDR-L2 as depicted in SEQ ID NO: 87, and CDR-L3 as depicted in SEQ ID NO: 88,

[0043] (g) CDR-H1 as depicted in SEQ ID NO: 94, CDR-H2 as depicted in SEQ ID NO: 95, CDR-H3 as depicted in SEQ ID NO: 96, CDR-L1 as depicted in SEQ ID NO: 97, CDR-L2 as depicted in SEQ ID NO: 98, and CDR-L3 as depicted in SEQ ID NO: 99,

[0044] (h) CDR-H1 as depicted in SEQ ID NO: 105, CDR-H2 as depicted in SEQ ID NO: 106, CDR-H3 as depicted in SEQ ID NO: 107, CDR-L1 as depicted in SEQ ID NO: 109, CDR-L2 as depicted in SEQ ID NO: 110, and CDR-L3 as depicted in SEQ ID NO: 111,

[0045] (i) CDR-H1 as depicted in SEQ ID NO: 115, CDR-H2 as depicted in SEQ ID NO: 116, CDR-H3 as depicted in SEQ ID NO: 117, CDR-L1 as depicted in SEQ ID NO: 118, CDR-L2 as depicted in SEQ ID NO: 119, and CDR-L3 as depicted in SEQ ID NO: 120,

[0046] (j) CDR-H1 as depicted in SEQ ID NO: 126, CDR-H2 as depicted in SEQ ID NO: 127, CDR-H3 as depicted in SEQ ID NO: 128, CDR-L1 as depicted in SEQ ID NO: 129, CDR-L2 as depicted in SEQ ID NO: 130, and CDR-L3 as depicted in SEQ ID NO: 131,

[0047] (k) CDR-H1 as depicted in SEQ ID NO: 137, CDR-H2 as depicted in SEQ ID NO: 138, CDR-H3 as depicted in SEQ ID NO: 139, CDR-L1 as depicted in SEQ ID NO: 140, CDR-L2 as depicted in SEQ ID NO: 141, and CDR-L3 as depicted in SEQ ID NO: 142,

[0048] (1) CDR-H1 as depicted in SEQ ID NO: 152, CDR-H2 as depicted in SEQ ID NO: 153, CDR-H3 as depicted in SEQ ID NO: 154, CDR-L1 as depicted in SEQ ID NO: 155, CDR-L2 as depicted in SEQ ID NO: 156, and CDR-L3 as depicted in SEQ ID NO: 157,

[0049] (m) CDR-H1 as depicted in SEQ ID NO: 167, CDR-H2 as depicted in SEQ ID NO: 168, CDR-H3 as depicted in SEQ ID NO: 169, CDR-L1 as depicted in SEQ ID NO: 170, CDR-L2 as depicted in SEQ ID NO: 171, and CDR-L3 as depicted in SEQ ID NO: 172,

[0050] (n) CDR-H1 as depicted in SEQ ID NO: 203, CDR-H2 as depicted in SEQ ID NO: 204, CDR-H3 as depicted in SEQ ID NO: 205, CDR-L1 as depicted in SEQ ID NO: 206, CDR-L2 as depicted in SEQ ID NO: 207, and CDR-L3 as depicted in SEQ ID NO: 208;

[0051] (o) CDR-H1 as depicted in SEQ ID NO: 214, CDR-H2 as depicted in SEQ ID NO: 215, CDR-H3 as depicted in SEQ ID NO: 216, CDR-L1 as depicted in SEQ ID NO: 217, CDR-L2 as depicted in SEQ ID NO: 218, and CDR-L3 as depicted in SEQ ID NO: 219;

[0052] (p) CDR-H1 as depicted in SEQ ID NO: 226, CDR-H2 as depicted in SEQ ID NO: 227, CDR-H3 as depicted in SEQ ID NO: 228, CDR-L1 as depicted in SEQ ID NO: 229, CDR-L2 as depicted in SEQ ID NO: 230, and CDR-L3 as depicted in SEQ ID NO: 231; and

[0053] (q) CDR-H1 as depicted in SEQ ID NO: 238, CDR-H2 as depicted in SEQ ID NO: 239, CDR-H3 as depicted in SEQ ID NO: 240, CDR-L1 as depicted in SEQ ID NO: 241, CDR-L2 as depicted in SEQ ID NO: 242, and CDR-L3 as depicted in SEQ ID NO: 243.

[0054] Within said aspects, it is also envisaged in the context of the present invention to provide a method, wherein the first domain has an amino acid sequence selected from the group consisting of: SEQ ID No: 7, 8, 17, 27, 28, 37, 38, 39, 40, 41, 48, 49, 50, 51, 52, 59, 60, 61, 62, 63, 64, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 89, 90, 91, 92, 93, 100, 101, 102, 103, 104, 113, 114, 121, 122, 123, 124, 125 174, 175, 176, 177, 178, 179, 180, 181, 223, 235 and 246, preferably 100 to 104.

[0055] Within said aspects, it is also envisaged in the context of the present invention to provide a method wherein ice nucleation is induced via an ice mist method or a reduced pressure method.

[0056] In additional aspects of the invention, lyophilized bispecific molecule compositions are prepared via the methods described herein.

[0057] It should be understood that although multiple embodiments in the specification are presented using "comprising" language, in many cases, the relevant embodiments can also be described using "consisting of..." or "consisting essentially of..." language. The present disclosure contemplates embodiments described as "comprising" a feature to include embodiments that "consist of" or "consisting essentially of" the feature. The term "a or an" refers to one or more. Likewise, the terms "a / an," "one or more," and "at least one" are used interchangeably herein. Unless the context clearly requires otherwise, the term "or" should be understood to cover each item in an alternative or combined manner.

[0058] It should also be understood that when describing a range of values, this disclosure contemplates a single value found within the range. For example, "a pH from about pH 4 to about pH 6" can be, but is not limited to, pH 4.2, 4.6, 5.2, 5.5, etc., and any value between such values. Within any range described herein, the endpoints of the range are included within the range. However, the description also contemplates the same range excluding lower and / or higher endpoints. When the term "about" is used, it means the enumerated number plus or minus 5%, 10% or more of the enumerated number. The actual variation expected can be determined from the context.

[0059] From the entirety of this application (including the drawings and specific embodiments), additional features and variations of the present invention will be apparent to those skilled in the art, and all such features are intended to be aspects of the present invention. Similarly, the features of the present invention described herein can be reorganized into additional embodiments, which are also intended to be aspects of the present invention, regardless of whether the combination of these features is specified as an aspect or embodiment of the present invention. The entire document is intended to be narrated as a unified disclosure, and it should be understood that all combinations of the features described herein (even if described in separate sections) are contemplated, even if these combinations of features are not found in the same sentence, paragraph, or section of this document. Moreover, only the limitations described herein that are critical to the present invention should be considered as such; variations of the present invention that lack limitations (those not described herein as critical) are intended to be aspects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1A-1B: (A) A standard lyophilization cycle for bispecific molecules without a CIN step includes annealing during the freezing phase and a longer primary drying time. (B) Without an annealing step, a CIN lyophilization cycle relies on triggering ice nucleation at higher temperatures to form larger ice crystals, resulting in shorter primary drying and total cycle times.

[0061] Figures 2A-2C: Lyophilization curves for lyophilization cycles of samples of BCMAxCD3 and DLL3xCD3 bispecific molecules. Each lyophilization cycle used the same lyophilization tray arrangement. (A) Standard non-CIN cycle with annealing, where the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecule was -33.3°C, and the maximum Tp (Tp max) for the DLL3xCD3 bispecific molecule was -32.8°C. (B) Standard non-CIN cycle without annealing, where the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecule was -31.3°C, and the maximum Tp for the DLL3xCD3 bispecific molecule was -31.5°C. (C) CIN cycle, where the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecule was -27.8°C, and the maximum Tp for the DLL3xCD3 bispecific molecule was -28.7°C.

[0062] Figures 3A-3B: Changes in % HMWS after lyophilization for a standard non-CIN cycle, a standard non-CIN cycle without annealing, and two CIN cycles 1 and 2 with different nucleation conditions and post-nucleation hold times. Therefore, CIN cycle 2 (nucleation parameters of -15°C and 0.5 h post-nucleation) was selected for this experiment. (A) BCMAxCD3 bispecific molecule (B) DLL3xCD3 bispecific molecule.

[0063] Figures 4A-4B: % HMWS stability results over time for standard non-CIN cycles, standard non-CIN cycles without annealing, and CIN cycles for the BCMAxCD3 and DLL3xCD3 bispecific molecules. Values ​​are % HMWS recorded immediately after reconstitution (post-reconstitution) at each time point. The lines in the graphs are overlaid. (A) BCMAxCD3 bispecific molecule (B) DLL3xCD3 bispecific molecule.

[0064] Figure 5 Time zero moisture content of lyophilized BCMAxCD3 and DLL3xCD3 bispecific molecule samples. No significant differences in moisture content were found among the groups (p>0.05).

[0065] Figures 6A-6F: CEX-HPLC results of BCMAxCD3 and DLL3xCD3 bispecific molecules. (A) Main peak of BCMAxCD3 bispecific molecule (B) Acidic peak of BCMAxCD3 bispecific molecule (C) Basic peak of BCMAxCD3 bispecific molecule (D) Main peak of DLL3xCD3 bispecific molecule (E) Acidic peak of DLL3xCD3 bispecific molecule (F) Basic peak of DLL3xCD3 bispecific molecule

[0066] Figure 7 : Average cake resistance at a depth of 1 mm in a 6R vial with a fill volume of 1.3 ml. Compared to the standard freeze-drying method, the cake resistance is reduced by >50% when using CIN and is even significantly reduced compared to the standard method with additional annealing. The "CIN 1" process with CIN step parameters not according to the invention has a nucleation temperature of -7°C, whereas the "CIN 2" process according to the invention has a nucleation temperature of -15°C. DETAILED DESCRIPTION

[0067] The present disclosure provides a method for preparing a freeze-dried bispecific molecule composition, particularly a composition comprising a bispecific T cell engaging molecule containing an Fc domain with an extended half-life, the method comprising a controlled ice nucleation (CIN) step. Surprisingly, CIN is a better alternative to introducing an annealing step into the freeze-drying of the bispecific molecule according to the present invention, to simultaneously achieve desired product quality (i.e., HMWS significantly reduced compared to conventional freeze-drying with an annealing step) and a reduced process time compared to conventional methods without an annealing step or a CIN step. Compared to conventional freeze-drying methods without annealing, the savings in process time come from shorter steps (c) and (d) as described herein. For example, for a primary drying step (c), a conventional freeze-drying method may require up to 100 hours, while the method according to the present invention has a shorter drying time, preferably only about 50 hours for step (c). Taking into account the secondary drying step (d), the total time savings of the method relative to conventional freeze-drying methods is at least 30% or even more, such as 40%. Therefore, the consumption of resources such as energy is also reduced. An exemplary conventional or - also referred to herein as - "standard" lyophilization method for bispecific molecules (i.e., typically without a CIN step) is understood herein to have the following steps: a freezing step (b) without a CIN or annealing step, e.g., at about -45°C for about 2 hours, a primary drying step (c), e.g., at about -25°C and about 70 mTorr for about 100 hours, and a secondary drying step (d), e.g., at about 40°C and about 70 mTorr for about 8 hours.

[0068] Without wishing to be bound by theory, by incorporating a CIN step with the process parameters disclosed herein, larger ice crystals and, therefore, shorter cycle times are achieved. CIN introduces externally formed ice crystals into the drug product vial, resulting in larger ice crystals after freezing. In the present invention, it was shown that the application of a CIN step effectively mitigates aggregation in terms of HMWS, compared to, for example, an annealing step performed under selected conditions suitable for bispecific molecules, without affecting other product quality attributes, such as lyophilized cake appearance and moisture content. In fact, the inventive method including a CIN step advantageously reduces cake drag, for example, by at least 30%, typically by about 50%, compared to standard non-CIN lyophilization with or without an annealing step. As demonstrated in the examples, lyophilization cycles with annealing typically result in undesirable aggregation, typically at least 3% HMWS, whereas lyophilization methods including a CIN step as described herein maintain aggregation (i.e., HMWS) at or below 1.5%, which is close to the pre-lyophilization value (1%), or even lower.

[0069] The CIN step according to the present invention comprises an incubation period preferably in the range of -18°C to -10°C for about 1 to 5 hours, typically 90 to 120 minutes, for example 90, 110, or 120 minutes, to cool the composition (pre-nucleation) so that the ice crystals do not melt when introduced into the next step. Ice nucleation is induced by introducing ice crystals by incubation preferably in the range of -18°C to -10°C, with an additional time to allow ice crystal growth (post-nucleation). An advantage of the present invention is that a method comprising a CIN step based on "sweet spot" conditions is provided, as explained and demonstrated herein, which contribute to favorable bispecific molecule product quality (in terms of product homogeneity, preferably without HMWS at higher product concentrations) and make the method more efficient with shorter drying process lengths. Higher ice nucleation temperatures above -10°C (e.g., -7°C)—although having the advantage of faster drying times—typically result in bispecific molecule products with inferior product quality in terms of aggregation. In contrast, lower ice nucleation temperatures, below -18°C, typically result in longer drying times and may negate any benefits in terms of process length, resource, and energy savings. With respect to drying time, some process parameters may have a smaller impact on product quality and process economics than the CIN step temperature and duration. For example, the secondary drying pressure may not be set within a wider range, as it is known in the art to have a minor impact on the drying process and, therefore, may not be used to achieve significant energy savings.

[0070] The method preferably comprises inducing ice nucleation in a liquid bispecific molecule composition in a vial exposed to a first temperature in the range of about -18° C. to about -10° C. (e.g., about -18° C., -17° C., -16° C., -15° C., -14° C., -13° C., -12° C., -11° C., or -10° C.) for about 1 to about five hours, preferably 90 to 120 minutes; exposing the vial to a second temperature in the range of about -25° C. to -50° C. for a second time period; and drying the composition at a third temperature in the range of about -5° C. to about -25° C. The method results in a vial containing a lyophilized bispecific molecule composition having favorable product parameters (low percentage HMWS despite having a high and commercially beneficial product concentration) and having an aspect ratio greater than or equal to about 0.75 (e.g., greater than or equal to 0.8, 0.85, 0.9, or 0.95). In various aspects, the method obtains a vial containing a lyophilized bispecific molecule composition having an aspect ratio greater than or equal to about 1 (e.g., greater than or equal to 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9 or 2). It is noteworthy that the methods disclosed herein allow for a surprisingly high fill volume (i.e., aspect ratio) for a given vial while minimizing the risk of vial rupture and maintaining a reasonable drying time. "Aspect ratio" is the fill height of the vial (the height of the bispecific molecule composition in the vial) divided by the vial inner diameter (aspect ratio = fill height / vial inner diameter). Previous methods required a reduced volume (lower aspect ratio) introduced into the vial to prevent vial rupture and product cake collapse during the lyophilization process. The methods described herein allow for a higher aspect ratio during the lyophilization process, which provides many potential advantages. For example, for bispecific molecule therapeutics administered in large doses, fewer vials are required for administration. Alternatively (or additionally), the method allows the use of smaller vials compared to other freeze-drying methods while maintaining the same fill volume, resulting in more units per batch and therefore increased production. The use of smaller vials also provides the benefit of reduced storage space at the manufacturing site and clinic. These are just examples of the advantages of the method of the present invention, which are achieved when producing a product filter cake suitable for biopharmaceutical use.

[0071] The present disclosure describes various conditions for use in the freeze-drying method of producing freeze-dried bispecific molecule compositions. Generally speaking, the vials (e.g., glass vials) suitable for pharmaceutical compositions are filled with liquid bispecific molecule compositions and exposed to different temperatures and pressures to obtain freeze-dried products. The vials can be any size or shape suitable for freeze-drying methods, and can be formed by a variety of materials (e.g., glass, metal, or plastic (e.g., polycarbonate, polystyrene, polypropylene, or polyolefin). For example, the vials can be glass or glass-like and are tubular in shape. Molded glass vials are commercially available, with a variety of different sizes and sizes. In fact, vials of various sizes are commercially available (e.g., sizes 2R, 4R, 6R, 8R, 10R, 15R, 20R, 25R, 30R, 50R, or 100R). The vials can be constructed to include suitable stoppers, such as can be obtained from, for example, Daikyo Seiko, Ltd. or West Pharmaceutical Services. In some embodiments, the freeze-drying process comprises a freeze-drying chamber or an ice nucleation system. The freeze-drying chamber can be operated according to the manufacturer's instructions for use with pharmaceutical compositions. In various aspects of the present disclosure, the freeze-drying process does not include an annealing step. "Annealing" refers to a process in which the temperature of the preparation (e.g., from a low temperature to a higher temperature, and then back to a low temperature) is cycled. Various aspects of the present disclosure method allow the production of a freeze-dried product without such an annealing step.

[0072] The methods of the present disclosure include inducing ice nucleation in a liquid bispecific molecule composition in a vial (referred to herein as the "ice nucleation step"). Ice nucleation can be initiated using any of a variety of methods, including but not limited to ice mist, sudden / rapid decompression, and vacuum-induced evaporative cooling. Other methods for controlling ice nucleation include, for example, ultrasound, gap freezing, electrical freezing, temperature quench freezing, use of pre-cooled shelves, and mechanical stirring.

[0073] In various aspects of the present disclosure, ice nucleation is induced in a liquid bispecific molecule composition via an ice fog method. The ice fog method involves "seeding" externally generated ice crystals into a supercooled solution in a vial. An ice fog generator is used to produce a fine ice crystal suspension, which is injected into a freeze dryer chamber. The crystals from the ice fog act as ice seeds for the supercooled liquid product in the vial. Once the ice crystals from the fog enter the partially plugged vial and contact the surface of the supercooled liquid, ice nucleation occurs immediately within the vial at a specified shelf temperature. This occurs simultaneously in all vials, thereby improving uniformity within the batch. Ice fog systems are available from IMA Life (Tonawanda, New York) and Millrock Technologies (Kingston, New York). Ice fog technology is further described, for example, in Azzarella et al., BioPharm. Int. [International Biopharmaceuticals], 29(12) (2017), 36-41.

[0074] In various aspects of the present disclosure, ice nucleation is induced in the liquid bispecific molecule composition via a decompression method. Rapid decompression methods generally involve first pressurizing the lyophilizer chamber to 1.5 to 2 atmospheres (about 20-30 psig) using an inert gas (e.g., nitrogen), and then rapidly releasing the pressure to slightly above ambient pressure (e.g., within 3 seconds or less). The rapid change in pressure induces nucleation in the vial. Rapid decompression systems are available from SP Scientific (Gardner, New York) and are further described in, for example, Luoma et al., "Controlled Ice Nucleation Using Pressurization-Depressurization Method[use [Controlled Ice Nucleation by the Pressurization-Decompression Method]”, in: Ward K., Matejtschuk P. (eds.) Lyophilization of Pharmaceuticals and Biologicals. Methods in Pharmacology and Toxicology. Humana Press, New York, NY, 2019, pp. 57–77.

[0075] Vacuum-induced evaporative cooling typically involves reducing the freeze-drying chamber pressure to just above the boiling point of the solution and allowing the enhanced evaporative cooling effect of the liquid surface to induce nucleation. A vacuum-induced evaporative cooling system is available from HOF Sonderanlagenbau GmbH (Lora, Germany).

[0076] Induce ice nucleation (e.g., via an ice fog method) in the liquid bispecific molecule composition in the context of the present invention in a vial exposed to a first temperature of about -18°C to about -10°C (e.g., about -18°C, about -17°C, about -16°C, about -15°C, about -14°C, about -13°C, about -12°C, about -11°C, or about -10°C) for a period of about 60 minutes to about five hours. Alternatively, the ice nucleation step comprises exposing the vial to the first temperature for about 30 minutes to about two hours (e.g., about 90 minutes to about two hours). In various aspects, the vial is exposed to a first temperature of about -15°C. Higher temperatures (e.g., -7°C) result in significantly higher HMWS percentage levels, which are not desirable in the context of the present invention or in the production of therapeutic bispecific molecules in general. Therefore, the careful selection of a specific temperature range for the CIN step is not arbitrary. Optionally, this step of the method further comprises maintaining the ice-nucleated bispecific molecule composition at this temperature for a post-nucleation period of up to two hours (optionally at the same temperature). For example, the post-nucleation hold time can be from about 30 minutes to about 90 minutes (e.g., from about 45 minutes to about 75 minutes, such as 50 minutes, 55 minutes, 60 minutes, 65 minutes, 70 minutes or 75 minutes). In various aspects, the post-nucleation hold time is about 60 minutes.

[0077] The method further includes exposing the vial to a second temperature of about -25°C to about -50°C for a second time period (referred to herein as the "freezing step"). The second temperature can be, for example, about -25°C to about -45°C, about -25°C to about -40°C, about -25°C to about -35°C, about -35°C to about -50°C, about -40°C to about -50°C, or about -45°C to about -50°C (e.g., about -40°C, about -41°C, about -42°C, about -43°C, about -44°C, about -45°C, about -46°C, about -47°C, about -48°C, about -49°C, or about -50°C). In various aspects, the freezing step includes exposing the vial to a second temperature of about -45°C. The second time period is optionally about one hour to about five hours, for example, about two hours to about four hours. The second time period can be about 60 minutes, about 90 minutes, about 120 minutes, about 150 minutes, about 180 minutes, about 210 minutes or about 240 minutes (or any range including these endpoints). In various aspects, the second time period is about three hours.

[0078] In various aspects of the present disclosure, the cooling rate of the vial can be controlled. For example, the first temperature can be optionally transitioned to the second temperature at a rate of about 0.01°C to about 0.5°C per minute (e.g., about 0.05°C to about 0.45°C per minute, about 0.1°C to about 0.3°C per minute, or about 0.15°C to about 0.25°C per minute). In various aspects, the first temperature is transitioned to the second temperature at a rate of about 0.2°C per minute.

[0079] The method further comprises drying the composition resulting from the freezing step at a third temperature of about 0° C. to about 40° C. (referred to herein as the "drying step") to provide a vial containing the lyophilized bispecific molecule composition having an aspect ratio greater than or equal to about 0.75. In various aspects, the third temperature used in the drying step is about 0° C. to about 35° C., about 0° C. to about 30° C., about 5° C. to about 40° C., about 10° C. to about 40° C., about 15° C. to about 40° C., about 20° C. to about 40° C., about 25° C. to about 40° C., about 30° C. to about 40° C., or about 35° C. to about 40° C. Alternatively, the third temperature is about 0° C. to about 25° C., e.g., about 10° C. to about 25° C. (e.g., about 25° C.).

[0080] Alternatively, the transition from the freezing step to the drying step comprises increasing the temperature at a rate of about 0.2° C. to about 0.7° C. per minute, and maintaining the vial at a temperature of about −40° C. to about −30° C. for about 15 minutes to about one hour. For example, the transition can comprise increasing the temperature at a rate of about 0.2° C., 0.3° C., 0.4° C., 0.5° C., 0.6° C., or 0.7° C. The transition can occur over a period of, for example, about 15 minutes, 30 minutes, 45 minutes, or 60 minutes.

[0081] In various aspects of the disclosure, the drying step can comprise exposing the vial containing the bispecific molecule composition to a temperature that is elevated at a rate of about 0.01°C to about 0.5°C per minute (e.g., about 0.05°C to about 0.45°C per minute, about 0.1°C to about 0.3°C per minute, or about 0.15°C to about 0.25°C per minute). In various aspects of the disclosure, the drying step comprises (1) maintaining the vial at a temperature of about -5°C to about 5°C for about 8 hours to about 12 hours, and (2) maintaining the vial at a temperature of about 20°C to about 30°C for about 20 hours to about 50 hours. In exemplary aspects of the disclosure, drying step 1 comprises maintaining the vial at a temperature of about -5°C to about 0°C, about 0°C to about 5°C, or about -2°C to about 2°C; for example, about -5°C, about -4°C, about -3°C, about -2°C, about -1°C, about 0°C, about 1°C, about 2°C, about 3°C, about 4°C, about 5°C, or any range having these endpoints. In various aspects, the time period of drying step 1 is from about 8 hours to about 10 hours, from about 9 hours to about 11 hours, or from about 10 hours to about 12 hours; for example, about 8 hours, about 9 hours, about 10 hours, about 11 hours, or about 12 hours, or any range having these endpoints. In exemplary aspects of the present disclosure, drying step 2 comprises maintaining the vial at a temperature of from about 25°C to about 30°C, from about 20°C to about 25°C, or from about 23°C to about 27°C; for example, about 20°C, about 21°C, about 22°C, about 23°C, about 24°C, about 25°C, about 26°C, about 27°C, about 28°C, about 29°C, or about 30°C, or any range having these endpoints. The time period for drying step 2 is optionally from about 20 hours to about 45 hours, about 20 hours to about 40 hours, about 20 hours to about 35 hours, about 20 hours to about 30 hours, about 25 hours to about 50 hours, about 30 hours to about 50 hours, about 35 hours to about 50 hours, about 40 hours to about 50 hours, or about 45 hours to about 50 hours (e.g., about 20 hours, about 25 hours, about 30 hours, about 35 hours, about 40 hours, about 45 hours, or about 50 hours). Drying step (1) optionally includes maintaining the vial at a temperature of about 0° C. for about 10 hours. Drying step (2) optionally includes maintaining the vial at a temperature of about 25° C. for about 40 hours.

[0082] The methods of the present disclosure allow for the use of a larger volume of liquid composition in a vial prior to lyophilization. In this regard, the liquid bispecific molecule composition can fill at least 50% of the volume of the vial prior to the ice nucleation step. The method can include filling at least 50% of the volume of the vial with the liquid bispecific molecule composition prior to the ice nucleation step. The liquid bispecific molecule composition can fill at least 55%, 60%, 65%, or 75% of the volume of the vial.

[0083] The present disclosure further provides a freeze-dried bispecific molecule composition prepared by the methods described herein. The freeze-dried bispecific molecule formulation optionally further comprises a sugar, a surfactant and / or a buffer. The formulation also optionally has a pH of about 3 to about 7 (or about 3.5, 4, 4.5, 5, 5.5, 6, 6.5 or 7). In some cases, the pH is about 4 to about 6. In some preferred cases, the pH of the formulation is about 4 or about 4.2. In various cases, the pH of the formulation is about 5. In some embodiments, the pH of the formulation is about 6.

[0084] In some embodiments, the bispecific molecule of the lyophilized formulation is an antigen binding protein. An "antigen binding protein" is a protein comprising a domain that binds to a designated target antigen (e.g., HER2 or CD3 and / or DLL3, BCMA, or CD33). The antigen binding protein comprises a scaffold or framework portion that allows the antigen binding domain to adopt a conformation that facilitates binding of the antigen binding protein to the antigen.

[0085] The term "half-life" is understood herein to mean the time it takes for the concentration of a biological substance (e.g., a bispecific molecule of the invention) in plasma to decrease from its maximum concentration (C max ) is reduced to C max Compared to non-half-life extended bispecific molecules that typically exhibit a half-life of 24 hours or less, the extended half-life is preferably at least 40 hours, more preferably 50, 60, 70, 80, 90 or 100 hours.

[0086] The term "polypeptide" is understood herein to mean an organic polymer comprising at least one continuous, unbranched amino acid chain. In the context of the present invention, polypeptides comprising more than one amino acid chain are also contemplated. A polypeptide amino acid chain typically comprises at least 50 amino acids, preferably at least 100, 200, 300, 400, or 500 amino acids. In the context of the present invention, it is also contemplated that the amino acid chain of a polymer is linked to an entity that does not consist of amino acids.

[0087] The term "antigen-binding polypeptides" according to the present invention is preferably a polypeptide that is immunospecifically bound to its target or antigen. It typically comprises the heavy chain variable region (VH) and / or light chain variable region (VL) of an antibody, or comprises a domain derived therefrom. The polypeptide according to the present invention comprises the minimum structural requirements of the antibody that allows immunospecific target binding. This minimum requirement can be, for example, defined by the presence of at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region), preferably all six CDRs. The antigen binding molecules of the present invention are preferably T cell engaging polypeptides, and therefore they may be characterized in that there are three or six CDRs in one or two binding domains, and the technician knows where those CDRs are located (in what order) within the binding domain. Preferably, "antigen binding molecules" are understood as "antigen-binding polypeptides" in the context of the present invention. In an alternative embodiment, the antigen-binding polypeptides of the present invention can be an aptamer.

[0088] Alternatively, the molecule in the context of the present invention is an antigen-binding polypeptide, which corresponds to an "antibody construct", which typically refers to a molecule wherein structure and / or function is based on the structure and / or function of an antibody (such as a full-length or complete immunoglobulin molecule). Therefore, antigen-binding molecules can be extracted in conjunction with their specific target or antigen and / or from the variable heavy chain (VH) and / or variable light chain (VL) domains of an antibody or its fragment. In addition, the domain combined with the binding partner according to the present invention is understood herein as the binding domain of the antigen binding molecules according to the present invention. Typically, the binding domain according to the present invention comprises the minimum structural requirements for the antibody allowing target binding. This minimum requirement can be, for example, defined by the presence of at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in VL districts) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in VH districts), preferably all six CDRs. An alternative approach to defining the minimum structural requirements for an antibody is to define the antibody epitope within a specific target structure, the protein domain of the target protein constituting the epitope region (epitope cluster), or by reference to a specific antibody that competes with the epitope of the defined antibody. The antibodies on which the constructs according to the present invention are based include, for example, monoclonal antibodies, recombinant antibodies, chimeric antibodies, deimmunized antibodies, humanized antibodies, and human antibodies.

[0089] In the context of the present invention, the polypeptides of the present invention bind to their corresponding target structures in a specific manner. Preferably, each binding domain of the polypeptides according to the present invention comprises a paratope that "specifically or immunospecifically" binds to its corresponding target structure, "(specifically or immunospecifically) recognizes" its corresponding target structure, or "(specifically or immunospecifically) reacts" with its corresponding target structure. According to the present invention, this means that the polypeptide or its binding domain interacts or (immuno)specifically interacts with a given epitope on the target molecule (antigen) and CD3, respectively. This interaction or binding occurs more frequently, more rapidly, has a longer duration, has a greater affinity, or has some combination of these parameters in an epitope on a specific target than in an alternative substance (non-target molecule). However, due to sequence similarity between homologous proteins in different species, a binding domain that (immuno)specifically binds to its target (e.g., a human target) may cross-react with homologous target molecules from different species (e.g., from non-human primates). Thus, the term "specifically / immunospecifically binds" may include binding of a binding domain to epitopes in more than one species and / or structurally related epitopes. The term "(immuno)selectively binds" does not include binding to structurally related epitopes.

[0090] The binding domain of the antigen binding molecules according to the present invention can, for example, include the CDR groups mentioned above. Preferably, those CDRs are included in the framework of the antibody light chain variable region (VL) and the antibody heavy chain variable region (VH); However, it does not necessarily include both. For example, the Fd fragment has two VH regions and generally retains some antigen-binding functions of the complete antigen-binding domain. Additional examples of antibody fragments, antibody variants, or binding domain formats include (1) a Fab fragment, a monovalent fragment having VL, VH, CL, and CH1 domains; (2) a F(ab')2 fragment, a bivalent fragment having two Fab fragments connected by a disulfide bridge at the hinge region; (3) an Fd fragment having two VH and CH1 domains; (4) an Fv fragment having the VL and VH domains of a single arm of an antibody; (5) a dAb fragment having a VH domain (Ward et al., (1989) Nature 341:544-546); (6) isolated complementarity determining regions (CDRs), and (7) single-chain Fv (scFv), the latter being preferred (e.g., derived from an scFV library). Examples of embodiments of antigen binding molecules according to the present invention are described, for example, in WO 00 / 006605, WO 2005 / 040220, WO 2008 / 119567, WO 2010 / 037838, WO 2013 / 026837, WO 2013 / 026833, US 2014 / 0308285, US 2014 / 0302037, WO 2014 / 144722, WO 2014 / 151910 and WO 2015 / 048272.

[0091] In addition, within the definition of "binding domain" or "domain that binds to ..." are fragments of full-length antibodies, such as VH, VHH, VL, (s) dAb, Fv, Fd, Fab, Fab', F(ab')2 or "rIgG" ("half antibodies"). The antigen-binding molecules according to the present invention can also include modified antibody fragments, also referred to as antibody variants, such as scFv, two-scFv or two (two)-scFv, scFv-Fc, scFv-zipper, scFab, Fab2, Fab3, diabodies, single-chain diabodies, tandem diabodies (Tandab), tandem two-scFv, tandem three-scFv, "multi-antibodies" (such as tribodies or tetrabodies) and single domain antibodies comprising only one variable domain (which can be VHH, VH or VL, specifically binding to an antigen or epitope independently of other V regions or domains), such as nanobodies or single variable domain antibodies. Typically, the binding domain of the present invention comprises a paratope that promotes binding to its binding partner.

[0092] As used herein, the term "single-chain Fv", "single-chain antibody" or "scFv" refers to a single polypeptide chain antibody fragment comprising variable regions from heavy and light chains, but lacking a constant region. Generally speaking, a single-chain antibody further comprises a polypeptide linker between the VH and VL domains that allows it to form a desired structure that will allow antigen binding. Single-chain antibodies are discussed in detail in Pluckthun, The Pharmacology of Monoclonal Antibodies, Vol. 113, Rosenburg and Moore, eds. Springer-Verlag, New York, pp. 269-315 (1994). Various methods for producing single-chain antibodies are known, including those described in U.S. Patent Nos. 4,694,778 and 5,260,203; International Patent Application Publication No. WO 88 / 01649; Bird (1988) Science 242:423-442; Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883; Ward et al. (1989) Nature 334:54454; Skerra et al. (1988) Science 242:1038-1041. In certain embodiments, single-chain antibodies can also be bispecific, multispecific, human and / or humanized, and / or synthetic.

[0093] In the context of the present invention, a paratope is understood to be an antigen binding site that is part of a polypeptide as described herein and that recognizes an antigen and binds thereto. A paratope is typically a small region of about at least 5 amino acids. A paratope as understood herein typically comprises portions of antibody-derived heavy chain (VH) and light chain (VL) sequences. Each binding domain of a molecule according to the present invention provides a paratope comprising a set of 6 complementary determining regions (CDR loops) (wherein each three are contained within antibody-derived VH and VL sequences, respectively).

[0094] In addition, the definition of term " antigen binding molecules " includes preferably multivalent (polyvalent / multivalent) construct, and therefore includes bispecific molecules, wherein bispecific means that it is specifically combined with two cell types (i.e. target cell and effector cell) comprising different antigen structures. Due to the antigen binding molecules of the present invention are preferably multitargeted, these antigen binding molecules are typically also multivalent (polyvalent / multivalent) molecules, that is, they specifically bind to more than two antigen structures, preferably four different binding domains in the context of the present invention, which are two target binding domains and two CD3 binding domains.Term " multitargeting bispecific antigen binding molecules " includes term " multitargeting bispecific T cell engager molecule " and " multitargeting bispecific T cell engager polypeptides (MBiTEP) ".Preferred " multitargeting bispecific antigen binding molecules " are " multitargeting bispecific T cell engager molecule " or " multitargeting bispecific T cell engager polypeptides (MBiTEP) ".Term " multitargeting bispecific T cell engager molecule " is understood to include term " multitargeting bispecific T cell engager polypeptides ". Furthermore, the definition of the term "antigen binding molecule" includes molecules comprising only one polypeptide chain as well as molecules consisting of more than one polypeptide chain, which chains may be identical (homodimers, homotrimers or homooligomers) or different (heterodimers, heterotrimers or heterooligomers). In such molecules comprising more than one polypeptide chain (i.e., typically two chains), the chains are typically attached to each other as heterodimers via charged pair binding, for example, within an hetero-Fc entity (which serves as a spacer and half-life extending moiety between two bispecific entities as described herein). Examples of the above-identified antigen-binding molecules, e.g., antibody-based molecules, and variants or derivatives thereof, are described, inter alia, in Harlow and Lane, Antibodies a laboratory manual, CSHL Press (1988), and Using Antibodies: a laboratory manual, CSHL Press (1999), Kontermann and Dübel, Antibody Engineering, Springer, 2nd edition 2010, and Little, Recombinant Antibodies for Immunotherapy, Cambridge University Press 2009.

[0095] As used herein, the term "bispecific" means that the antigen binding molecule is "at least bispecific", i.e., it addresses two different cell types (i.e., target cells and effector cells) and comprises at least a first binding domain and a third binding domain and a second binding domain and a fourth binding domain, wherein at least two of the binding domains bind to two antigens or targets (preferably selected from CD20, CD22, FLT3, MSLN, CDH3, CLL1 and EpCAM), and the other two binding domains of the same molecule bind to another antigen (here: CD3) on effector cells (typically T cells). Thus, the antigen binding molecule according to the present invention has specificity for at least two different antigens or targets. For example, the two domains preferably do not bind to the extracellular epitopes of CD3 epsilon of one or more species as described herein.

[0096] The term "target cell surface antigen" refers to an antigen structure expressed by a cell, which is present on the cell surface so that it can be approached by antigen binding molecules as described herein. In the context of the present invention, preferred target cell surface antigens are tumor-associated antigens (TAA). It can be a protein, preferably the extracellular portion of a protein, or a carbohydrate structure, preferably a carbohydrate structure of a protein, such as a glycoprotein. It is preferably a tumor antigen. The term "bispecific antigen binding molecule" of the present invention also encompasses bispecific multi-targeted antigen binding molecules, such as three-targeted antigen binding molecules, which include three binding domains, or have more than three (e.g., four, five ...) specific constructs.

[0097] Preferred in the context of the present invention are "multi-targeting" molecules, which are understood herein as "typically targeting at least two targets (e.g., TAAs) / molecules of the present invention / target cells". In this regard, multi-targeting molecules, such as antigen binding molecules, have specificity for two typically identical effector structures on effector cells (e.g., CD3, more preferably CD3ε (CD3e, included herein whenever "CD3" is mentioned)) and at least two target cell surface antigens. Said specificity is conferred by corresponding binding domains as defined herein. Typically, "multi-targeting" refers to molecules that are specific for at least two (preferably different) target cell surface antigens (e.g., TAAs), which confer preferred properties of the multi-targeting antigen binding molecules according to the present invention, i.e., mitigating antigen loss and increasing selectivity, i.e., selectivity for killing target cells that co-express the target for which the molecule of the present invention has a binding domain directed, as well as target cells associated with the disease. Thus, the therapeutic window of the molecules of the present invention is increased relative to single-targeting bispecific molecules, which typically results in higher drug tolerance, as demonstrated herein.

[0098] T cell engaging antigen binding molecules, such as single-chain polypeptides according to the present invention, are preferably bispecific, which is understood herein to typically comprise a domain that binds to at least one target antigen and another domain that binds to CD3. Therefore, it is not naturally occurring, and its function is significantly different from that of naturally occurring products. Therefore, the polypeptide according to the present invention is an artificial "hybrid" polypeptide comprising at least two different binding domains with different specificities, and is therefore bispecific. Bispecific antigen binding molecules can be produced by a variety of methods (including fusion of hybridomas or connection of Fab' fragments). See, for example, Songsivilai and Lachmann, Clin. Exp. Immunol. [Clinical and Experimental Immunology] 79: 315-321 (1990).

[0099] At least four binding domains and variable domains (VH / VL) of the antigen binding molecules of the present invention typically comprise peptide linkers (spacer peptides).According to the present invention, term " peptide linker " comprises aminoacid sequence, and the aminoacid sequence of one (variable and / or in conjunction with) domain and another (variable and / or in conjunction with) domain of antigen binding molecules of the present invention is connected to each other by this aminoacid sequence.(wherein the peptide linker between the first and second binding domains and the third and fourth domains is preferably capable of combining two targets simultaneously, and these targets are preferably different targets (such as TAA1 and TAA2), preferably on identical cell) preferably flexible and length limited, such as 5,6,7,8,9,10,11,12,13,14,15,16,17 or 18 amino acid.Peptide linkers can also be used for merging spacers with other domains of antigen binding molecules of the present invention.The basic technical feature of such peptide linkers is that it does not comprise any polymerization activity. Suitable peptide linkers are those described in U.S. Patent Nos. 4,751,180 and 4,935,233 or WO 88 / 09344. Peptide linkers can also be used for attaching other domains or modules or districts (such as domains of half-life extension) to the antigen binding molecules of the present invention. However, typically the joint between the first target binding domain and the second target binding domain is different from the intra-binder linker connecting VH and VL in the target binding domain. The difference is that the joint between the first binding domain and the second binding domain is one more amino acid than the intra-binder linker in the binding domain, for example, six and five amino acids, such as SGGGGS vs. GGGGS. This unexpectedly makes the specific antigen binding molecule form as described herein have flexibility and stability simultaneously. The spacer (or synonymous spacer entity) between two bispecific entities as described herein is the specific embodiment of a joint, because spacer also plays the role of a joint, because it helps to connect two bispecific entities to preferably construct at least one continuous polypeptide chain including four binding domains or parts thereof. However, in addition, the spacer acts as an entity that spatially separates the two bispecific entities. Thus, a spacer in the context of the present invention is a specific embodiment of a linker, which together with two further shorter and flexible linkers at each end helps to connect the two binding domains (of two different bispecific entities), but first and foremost to space them apart so that the two bispecific entities can advantageously function as described herein, for example to show an unexpectedly high selectivity gap.

[0100] The antigen binding molecules of the present invention are preferably "antigen binding molecules generated in vitro". This term refers to an antigen binding molecule according to the above definition, wherein all or part of the variable region (e.g., at least one CDR) is generated in non-immune cell selection, such as in vitro phage display, protein chip or any other method that can test the ability of candidate sequences to bind to antigens. Therefore, this term preferably excludes sequences produced only by genomic rearrangement in animal immune cells. "Recombinant antibodies" are antibodies produced by using recombinant DNA technology or genetic engineering.

[0101] As used herein, the monoclonal antibody from which the term "monoclonal antibody" (mAb) or antigen-binding molecules is derived refers to an antibody obtained from a substantially uniform antibody population, that is, except for possible naturally occurring mutations and / or post-translational modifications (e.g., isomerization, amidation) that may exist in small quantities, the individual antibodies constituting the population are identical. Compared with conventional (polyclonal) antibody preparations typically comprising different antibodies for different determinants (or epitopes), monoclonal antibodies have a high degree of specificity for a single antigen side or determinant on an antigen. In addition to their specificity, monoclonal antibodies are also advantageous in that they are synthesized by hybridoma culture, and are therefore not contaminated by other immunoglobulins. The modifier "monoclonal" indicates the characteristic of an antibody obtained from a substantially uniform antibody population, and should not be construed as requiring antibody production by any ad hoc method.

[0102] For the preparation of monoclonal antibodies, any technology providing the antibody produced by continuous cell line culture can be used.For example, the monoclonal antibody to be used can be by the hybridoma method described first by Koehler et al., Nature [nature], 256:495 (1975), or can be prepared by recombinant DNA method (see, for example, U.S. Patent number 4,816,567). The example of other technology for producing human monoclonal antibodies includes tri-source hybridoma technology, human B cell hybridoma technology (Kozbor, Immunology Today [today's immunology] 4 (1983), 72) and EBV-hybridoma technology (Cole et al., Monoclonal Antibodies and Cancer Therapy [monoclonal antibodies and cancer therapy], Alan R.Liss, Inc. [Allen Liss Co., Ltd.] (1985), 77-96).

[0103] Standard methods such as enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance analysis such as Biacore can then be used to analyze the TM) Hybridomas are screened to identify one or more hybridomas that produce antibodies that specifically bind to a given antigen. Any form of the relevant antigen can be used as an immunogen, such as a recombinant antigen, a naturally occurring form, any variant or fragment thereof, and antigenic peptides thereof. Surface plasmon resonance, as employed in the Biacore system, can be used to increase the efficiency of phage antibodies that bind to epitopes of target cell surface antigens (Schier, Human Antibodies Hybridomas 7 (1996), 97-105; Malmborg, J. Immunol. Methods 183 (1995), 7-13).

[0104] Another exemplary method for preparing monoclonal antibodies includes screening protein expression libraries, such as phage display or ribosome display libraries. Phage display is described, for example, in Ladner et al., U.S. Patent No. 5,223,409; Smith (1985) Science 228: 1315-1317; Clackson et al., Nature 352: 624-628 (1991) and Marks et al., J. Mol. Biol. 222: 581-597 (1991).

[0105] In addition to using display libraries, related antigens can also be used to immunize non-human animals, such as rodents (such as mice, hamsters, rabbits or rats). In one embodiment, the non-human animal includes at least a portion of a human immunoglobulin gene. For example, it is possible to utilize large fragments of human Ig (immunoglobulin) loci to engineer mouse antibody production defective mouse strains. Using hybridoma technology, antigen-specific monoclonal antibodies derived from genes with desired specificity can be produced and selected. See, for example, XENOMOUSE TM ; Green et al. (1994) Nature Genetics 7: 13-21; US ​​2003-0070185; WO 96 / 34096 and WO 96 / 33735.

[0106] Monoclonal antibodies can also be obtained from non-human animals and then modified using recombinant DNA techniques known in the art, such as humanization, deimmunization, chimerization, etc. Examples of modified antigen-binding molecules include humanized variants of non-human antibodies, "affinity matured" antibodies (see, e.g., Hawkins et al. J. Mol. Biol. 254, 889-896 (1992) and Lowman et al. Biochemistry 30, 10832-10837 (1991)) and antibody mutants with altered effector function(s) (see, e.g., U.S. Pat. No. 5,648,260, Kontermann and Dübel (2010), loc. cit., and Little (2009), loc. cit.).

[0107] In immunology, affinity maturation is a process by which B cells produce antibodies with increased affinity for antigens during an immune response. After repeated exposure to the same antigen, the host will produce antibodies with greater affinity. Similar to the natural prototype, in vitro affinity maturation is based on the principle of mutation and selection. In vitro affinity maturation has been successfully used to optimize antibodies, antigen-binding molecules, and antibody fragments. Random mutations are introduced into CDRs using radiation, chemical mutagens, or error-prone PCR. In addition, genetic diversity can be increased by chain shuffling. Two or three rounds of mutations and selection are performed using display methods (such as phage display) to typically produce antibody fragments with an affinity in the low nanomolar range.

[0108] The amino acid substitution changes of the preferred type of antigen binding molecules include replacing one or more hypervariable region residues of the parent antibody (such as humanized or human antibody). Generally speaking, one or more obtained variants selected for further development will have improved biological properties relative to the parent antibody that produces them. A convenient way for producing such substitution variants relates to affinity maturation using phage display. In short, several hypervariable region side ends (such as 6-7 side ends) are mutated to produce all possible amino acid substitutions at each side end. The antibody variants thus produced are displayed as fusions of the gene III product of the M13 packaged in each particle from filamentous phage particles in a monovalent manner. The biological activity (such as binding affinity) of the variant displayed by phage display is then screened as disclosed herein. In order to identify the candidate hypervariable region side ends for modification, alanine scanning mutagenesis can be carried out to identify the hypervariable region residues that have a significant contribution to antigen binding. Alternatively or additionally, it may be beneficial to analyze the crystal structure of the antigen-antibody complex to identify contact points between the binding domain and, for example, human CS1, BCMA, CD20, CD22, FLT3, CD123, CDH3, MSLN, CLL1, or EpCAM. Such contact residues and adjacent residues are candidates for substitution according to the techniques described herein. Once such variants are generated, the panel of variants is screened as described herein, and antibodies with superior properties in one or more relevant assays can be selected for further development.

[0109] The monoclonal antibodies and antigen-binding molecules of the present invention particularly include "chimeric" antibodies (immunoglobulins) in which a portion of the heavy and / or light chains is identical or homologous to the corresponding sequence in an antibody derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of one or more chains is identical or homologous to the corresponding sequence in an antibody derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, as long as they exhibit the desired biological activity (U.S. Patent No. 4,816,567; Morrison et al., Proc. Natl. Acad. Sci. USA [Proceedings of the National Academy of Sciences of the United States of America] 81: 6851-6855 (1984)). Chimeric antibodies of interest herein include "primatized" antibodies, which comprise variable domain antigen-binding sequences derived from non-human primates (e.g., Old World monkeys, apes, etc.) and human constant region sequences. A variety of methods for preparing chimeric antibodies have been described. See, e.g., Morrison et al., Proc. Natl. Acad. ScL USA 81:6851, 1985; Takeda et al., Nature 314:452, 1985; Cabilly et al., U.S. Pat. No. 4,816,567; Boss et al., U.S. Pat. No. 4,816,397; Tanaguchi et al., EP 0171496; EP 0173494; and GB 2177096.

[0110] In some embodiments, the present invention relates to antibody, antigen binding molecules, antibody fragment or antibody variants.Antibody, antigen binding molecules, antibody fragment or antibody variants can also be modified by the method specifically lacking human t cell epitopes (being called " deimmunization " method) disclosed in for example WO 98 / 52976 or WO 00 / 34317.In brief, can be for the heavy chain and the light chain variable domain of the peptide analysis antibody combined with MHC II class; These peptides represent potential t cell epitopes (as defined in WO 98 / 52976 and WO 00 / 34317).In order to detect potential t cell epitopes, can apply the computer modeling method being called " peptide threading ", and in addition for the motif existing in VH and VL sequence, can search the database of people MHC I 1 class binding peptides, as described in WO 98 / 52976 and WO 00 / 34317.These motifs are combined with any one in 18 kinds of main MHC I 1 class DR allotypes, and therefore constitute potential t cell epitopes.The potential t cell epitopes detected can be by replacing a small amount of amino acid residues in the variable domains, or preferably eliminate by single amino acid replacement. Typically, conservative substitutions are made. Generally, but not exclusively, amino acids that are common to positions in human germline antibody sequences can be used. Human germline sequences are disclosed, for example, in Tomlinson et al. (1992) J. Mol. Biol. 227: 776-798; Cook, GP et al. (1995) Immunol. Today 16(5): 237-242; and Tomlinson et al. (1995) EMBO J. 14: 14: 4628-4638. The V BASE directory provides a comprehensive directory of human immunoglobulin variable region sequences (edited by Tomlinson, LA. et al., MRC Centre for Protein Engineering, Cambridge, UK). These sequences can be used as a source of human sequences, for example, for framework regions and CDRs. Common human framework regions can also be used, for example, as described in U.S. Patent No. 6,300,064.

[0111] "Humanized" antibodies, antigen-binding molecules, variants or fragments thereof (such as Fv, Fab, Fab', F(ab')2 or other antigen-binding subsequences of antibodies) are antibodies or immunoglobulins of most human sequences that contain one or more minimal sequences derived from non-human immunoglobulins. In most cases, humanized antibodies are human immunoglobulins (recipient antibodies) in which the residues from the hypervariable regions (also known as CDRs) of the receptor are replaced by residues from the hypervariable regions (donor antibodies) of non-human (e.g., rodent) species (such as mice, rats, hamsters or rabbits) with desired specificity, affinity and capacity. In some cases, the Fv framework region (FR) residues of the human immunoglobulin are replaced by corresponding non-human residues. In addition, as used herein, "humanized antibodies" may also include residues that are not found in either the recipient antibody or the donor antibody. These modifications are made to further improve and optimize antibody performance. Humanized antibodies may also comprise at least a portion of an immunoglobulin constant region (Fc), which is typically a human immunoglobulin. For more details, see Jones et al., Nature, 321:522-525 (1986); Reichmann et al., Nature, 332:323-329 (1988); and Presta, Curr. Op. Struct. Biol., 2:593-596 (1992).

[0112] Humanized antibodies or fragments thereof can be produced by replacing the sequences of the Fv variable domains that are not directly involved in antigen binding with equivalent sequences of human Fv variable domains. Exemplary methods for producing humanized antibodies or fragments thereof are provided by: Morrison (1985) Science [Science] 229: 1202-1207; Oi et al. (1986) BioTechniques [Biotechnology] 4: 214; and US 5,585,089; US 5,693,761; US ​​5,693,762; US 5,859,205; and US 6,407,213. Those methods include isolating, manipulating and expressing nucleic acid sequences encoding all or part of the immunoglobulin Fv variable domains from at least one of the heavy chain or light chain. Such nucleic acids can be obtained from hybridomas that produce antibodies against a predetermined target as described above, as well as other sources. The recombinant DNA encoding the humanized antibody molecule can then be cloned into an appropriate expression vector.

[0113] Humanized antibodies can also be produced using transgenic animals (such as mice expressing human heavy chain and light chain genes but not expressing endogenous mouse immunoglobulin heavy chain and light chain genes). Winter has described an exemplary CDR transplantation method (U.S. Patent number 5,225,539) that can be used for preparing humanized antibodies as herein described. The whole CDR of specific human antibodies can be replaced with at least a portion of non-human CDR, or only some CDR can be replaced with non-human CDR. Only need to replace the CDR quantity required for humanized antibodies and predetermined antigen.

[0114] Humanized antibodies can be optimized by introducing conservative substitutions, consensus sequence substitutions, germline substitutions and / or back mutations. Such altered immunoglobulin molecules can be prepared by any of several techniques known in the art (e.g., Teng et al., Proc. Natl. Acad. Sci. USA [Proceedings of the National Academy of Sciences of the United States], 80:7308-7312, 1983; Kozbor et al., Immunology Today [Today's Immunology], 4:7279, 1983; Olsson et al., Meth. Enzymol. [Enzymology Methods], 92:3-16, 1982, and EP 239 400).

[0115] The terms "human antibody", "human antigen binding molecules" and "human binding domain" include antibodies, antigen binding molecules and binding domains with antibody regions (such as variable regions and constant regions) or domains substantially corresponding to human germline immunoglobulin sequences known in the art, including those described by, for example, Kabat et al. (1991) (in the above citations). Human antibodies, antigen binding molecules or binding domains of the present invention may include, for example, amino acid residues not encoded by human germline immunoglobulin sequences in CDR and particularly in CDR3 (for example, mutations introduced by random or side-specific mutagenesis in vitro or by somatic mutations in vivo). Human antibodies, antigen binding molecules or binding domains may have at least one, two, three, four, five or more positions substituted by amino acid residues that are not encoded by human germline immunoglobulin sequences. However, the definition of human antibodies, antigen binding molecules and binding domains as used herein also encompasses "fully human antibodies", which only comprise non-artificial and / or genetically altered human antibody sequences, such as those derived using, for example, Xenomouse technology or systems. Preferably, a "fully human antibody" does not include amino acid residues not encoded by human germline immunoglobulin sequences.

[0116] In some embodiments, the antigen binding molecules of the present invention are "isolated" or "substantially pure" antigen binding molecules. When used to describe the antigen binding molecules disclosed herein, "isolated" or "substantially pure" means that the antigen binding molecules have been identified, separated and / or recovered from the components of their production environment. Preferably, the antigen binding molecules do not contain all other components from their production environment or are not substantially associated with all other components from their production environment. The contaminating components of their production environment, such as those produced by recombinant transfected cells, are substances that typically interfere with the diagnosis or therapeutic use of the polypeptide, and may include enzymes, hormones, and other proteins or non-protein solutes. The antigen binding molecules can, for example, account for at least about 5% or at least about 50% by weight of the total protein in a given sample. It should be understood that, depending on the circumstances, the isolated protein can account for 5% to 99.9% by weight of the total protein content. By using an inducible promoter or a high expression promoter, polypeptides can be prepared at significantly higher concentrations so that they are prepared at increased concentration levels. This definition includes producing antigen binding molecules in a variety of organisms and / or host cells known in the art. In preferred embodiments, the antigen-binding molecule is (1) purified to a sufficient extent to obtain at least 15 N-terminal or internal amino acid sequence residues by using a spinning cup sequenator, or (2) purified to homogeneity by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue or preferably silver stain. However, isolated antigen-binding molecules are typically prepared by at least one purification step.

[0117] The term "binding domain" in relation to the present invention characterizes a domain that (specifically) binds to / interacts with / recognizes a given target epitope or a given target flanking a target molecule (antigen) (e.g., CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, or EpCAM and CD3, respectively). Typically, the structure and function of the first and third or second and fourth binding domains (e.g., recognizing CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, or EpCAM) and preferably also the structure and / or function of the effector binding domain (usually the second and fourth or first and third binding domains recognizing CD3) are based on the structure and / or function of an antibody (e.g., a full-length or intact immunoglobulin molecule) and / or are extracted from the variable heavy chain (VH) and / or variable light chain (VL) domains of an antibody or a fragment thereof. Preferably, the one or more target cell surface antigen binding domains are characterized by the presence of three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). The effector (typically CD3) binding domain also preferably comprises the minimum structural requirements of an antibody that allows target binding. More preferably, the second binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). It is envisioned that the first binding domain and / or the second binding domain are produced or obtainable by phage display or library screening methods, rather than being produced or obtainable by transplanting CDR sequences from pre-existing (monoclonal) antibodies into a scaffold.

[0118] According to the present invention, the binding domain is in the form of one or more polypeptides. Such polypeptides may include a protein portion and a non-protein portion (e.g., a chemical linker or chemical cross-linking agent, such as glutaraldehyde). Proteins (including fragments thereof, preferably biologically active fragments, and peptides generally having less than 30 amino acids) comprise two or more amino acids coupled to each other via covalent peptide bonds (generating an amino acid chain).

[0119] As used herein, the term "polypeptide" describes a group of molecules that are generally composed of more than 30 amino acids. Polypeptides can further form multimers, such as dimers, trimers and higher-order oligomers, i.e., are composed of more than one polypeptide molecule. The polypeptide molecules that form such dimers, trimers, etc. can be identical or different. Therefore, the corresponding high-order structures of such multimers are called homo- or hetero-dimers, homo- or hetero-trimers, etc. An example of a hetero-multimer is an antibody molecule, the naturally occurring form of which is composed of two identical polypeptide light chains and two identical polypeptide heavy chains. The terms "peptide", "polypeptide" and "protein" also refer to naturally modified peptides / polypeptides / proteins, wherein the modification is achieved, for example, by post-translational modification (such as glycosylation, acetylation, phosphorylation, etc.). When mentioned in this article, "peptide", "polypeptide" or "protein" may also be chemically modified, such as pegylated. Such modifications are well known in the art and are described below.

[0120] Preferably, the binding domain that binds to any one of CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, CDH3, MSLN and EpCAM and / or the binding domain that binds CD3□ is a human binding domain. Antibodies and antigen-binding molecules comprising at least one human binding domain avoid some problems associated with antibodies or antigen-binding molecules with non-human, such as rodent (e.g., mouse, rat, hamster or rabbit) variable and / or constant regions. The presence of such rodent-derived proteins can result in rapid clearance of antibodies or antigen-binding molecules, or can result in the patient generating an immune response to antibodies or antigen-binding molecules. In order to avoid the use of rodent-derived antibodies or antigen-binding molecules, human or fully human antibodies / antigen-binding molecules can be produced by introducing human antibody functions into rodents so that rodents produce fully human antibodies.

[0121] The term "high molecular weight species" (HMWS) typically refers to product-related variants of the bispecific molecules of the present invention. HMWs may include dimers, trimers, tetramers, etc. formed from monomers that may be covalently or non-covalently linked. HMWS may be composed of misfolded monomers that typically would not be in monomeric form, wherein the surfaces of the monomers are exposed. They may or may not have an impact on safety and / or efficacy, but should generally be avoided from a regulatory perspective. Typically, HMWS in the context of the present invention are determined by size exclusion chromatography (SE-UHPLC) and may represent a fraction different from the "main peak" (i.e., the desired monomeric product).

[0122] The term "amino acid" or "amino acid residue" typically refers to an amino acid having its art-recognized definition, such as an amino acid selected from the group consisting of: alanine (Ala or A); arginine (Arg or R); asparagine (Asn or N); aspartic acid (Asp or D); cysteine ​​(Cys or C); glutamine (Gln or Q); glutamic acid (Giu or E); glycine (Giy or G); histidine (His or H); isoleucine (Ile or I); leucine (Leu or L); lysine (Lys or K); methionine (Met or M); phenylalanine (Phe or F); proline (Pro or P); serine (Ser or S); threonine (Thr or T); tryptophan (Trp or W); tyrosine (Tyr or Y); and valine (Val or V), although modified, synthetic, or rare amino acids may be used as desired. In general, amino acids can be grouped as having nonpolar side chains (e.g., Ala, Cys, Ile, Leu, Met, Phe, Pro, Val); having negatively charged side chains (e.g., Asp, Giu); having positively charged side chains (e.g., Arg, His, Lys); or having uncharged polar side chains (e.g., Asn, Cys, Gln, Giy, His, Met, Phe, Ser, Thr, Trp, and Tyr).

[0123] Amino acid modifications include, for example, deletions and / or insertions and / or substitutions of residues within the amino acid sequence of the antibody construct. Any combination of deletions, insertions, and substitutions is performed to arrive at the final construct, provided that the final construct possesses the desired characteristics. Amino acid changes can also alter post-translational processes of the antibody construct, such as altering the number or position of glycosylation sites.

[0124] For example, can insert, replace or lack 1,2,3,4,5 or 6 amino acid (of course, depend on its length) in each CDR, and can insert, replace or lack 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20 or 25 amino acid in each FR.Preferably, aminoacid sequence is inserted into the antibody construct and comprises that the length range of the polypeptide that contains 100 or more residues is 1,2,3,4,5,6,7,8,9 or 10 residue amino and / or carboxyl terminal fusions, and single or multiple amino acid residue sequence inserts.Also can carry out corresponding modification in the third structural domain of antibody construct of the present invention.The insertion variants of antibody construct of the present invention comprise with the N-terminal of the antibody construct of enzyme or the fusion of C-terminal or with the fusion of polypeptide.

[0125] The sites of greatest interest for substitution mutagenesis include, but are not limited to, CDRs of heavy and / or light chains, particularly hypervariable regions, but FR changes of heavy and / or light chains are also contemplated. Substitutions are preferably conservative substitutions as described herein. Preferably, 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids can be substituted in the CDRs, while 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 25 amino acids can be substituted in the framework regions (FRs), depending on the length of the CDRs or FRs. For example, if a CDR sequence encompasses 6 amino acids, it is envisioned that 1, 2 or 3 of these amino acids are substituted. Similarly, if a CDR sequence encompasses 15 amino acids, it is envisioned that 1, 2, 3, 4, 5 or 6 of these amino acids are substituted.

[0126] A useful method for identifying certain residues or regions of an antibody construct that are preferred locations for mutagenesis is called "alanine scanning mutagenesis," as described by Cunningham and Wells in Science, 244: 1081-1085 (1989). Here, residues or target residue groups within an antibody construct (e.g., charged residues such as arg, asp, his, lys, and glu) are identified and replaced with neutral or negatively charged amino acids (most preferably alanine or polyalanine) to affect the interaction of the amino acids with the epitope.

[0127] In some embodiments, the present invention provides the method for the present invention to replace the amino acid position that has functional sensitivity.Then by introducing further or other variants at the substitution site or for the substitution site to refine those demonstrations to replace the amino acid position that has functional sensitivity.Therefore, although the site or the region for introducing amino acid sequence variation is predetermined, the property of sudden change itself does not need to be predetermined.For example, in order to analyze or optimize the performance of given site sudden change, alanine scanning or random mutagenesis can be carried out at the target codon or the region, and the expressed antibody construct variant is screened to obtain the optimal combination of desired activity.The technology for carrying out substitution mutation at the predetermined site in the DNA with known sequence is well known, for example, M13 primer mutagenesis and PCR mutagenesis.The mensuration of antigen-binding activity (such as target cell surface antigen or CD3 combination) is used to screen mutants.

[0128] In general, if an amino acid is substituted in one or more or all CDRs of the heavy and / or light chain, it is preferred that the "substituted" sequence obtained thereafter has at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85%, and particularly preferably 90% or 95% identity with the "original" CDR sequence. This means that the substitution depends on the length of the CDR and the degree of identity with the "substituted" sequence. For example, a CDR with 5 amino acids is preferably 80% identical to its substituted sequence, so that at least one amino acid is substituted. Thus, the CDRs of an antibody construct may have different degrees of identity with the sequences they replace, for example, CDRL1 may have 80% identity, while CDRL3 may have 90%.

[0129] Preferred substitutions (or replacements) are conservative substitutions. However, as long as the antibody construct retains its ability to bind to the target cell surface antigen via the first domain and to CD3, respectively, CD3ε via the second domain and / or its CDRs are identical to the sequences subsequently substituted (at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85% and particularly preferably 90% or 95% identical to the "original" CDR sequences), any substitution (including non-conservative substitutions or one or more of the "exemplary substitutions" listed in Table 3 below) is contemplated.

[0130] Conservative substitutions are shown under the heading "Preferred Substitutions" in Table 3. If such substitutions result in a change in biological activity, then more substantial changes, denominated "Exemplary Substitutions" in Table 3, or as described further below with reference to amino acid classes, can be introduced and screened for the desired characteristics.

[0131] Table 1: Amino acid substitutions

[0132]

[0133]

[0134] Substantial modification of the biological properties of the antibody constructs of the invention is accomplished by selecting substitutions that differ significantly in their effect on maintaining: (a) the structure of the polypeptide backbone in the area of ​​the substitution, such as a sheet or helical conformation, (b) the charge or hydrophobicity of the molecule at the target site, or (c) the bulk of the side chain. Naturally occurring residues are grouped based on common side chain properties: (1) hydrophobic: norleucine, met, ala, val, leu, ile; (2) neutral hydrophilic: cys, ser, thr, asn, gln; (3) acidic: asp, glu; (4) basic: his, lys, arg; (5) residues affecting chain orientation: gly, pro; and (6) aromatic: trp, tyr, phe.

[0135] Non-conservative substitutions will entail exchanging a member of one of these categories for another. Any cysteine ​​residue that is not involved in maintaining the proper conformation of the antibody construct can generally be substituted with a serine to improve the oxidative stability of the molecule and prevent abnormal cross-linking. Conversely, one or more cysteine ​​bonds can be added to the antibody to improve its stability (particularly when the antibody is an antibody fragment (such as an Fv fragment)).

[0136] For amino acid sequences, sequence identity and / or similarity are determined using standard techniques known in the art, including, but not limited to, the local sequence identity algorithm of Smith and Waterman, 1981, Adv. Appl. Math. 2:482, the sequence identity alignment algorithm of Needleman and Wunsch, 1970, J. Mol. Biol. 48:443, the search by similarity method of Pearson and Lipman, 1988, Proc. Nat. Acad. Sci. USA 85:2444, computerized implementations of these algorithms (GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Avenue, Madison, WI). Drive, Madison, Wis.), Devereux et al., 1984, Nucl. Acid Res. 12:387-395, preferably using the default settings, or by inspection. Preferably, percent identity is calculated by FastDB based on the following parameters: mismatch penalty of 1; gap penalty of 1; gap size penalty of 0.33; and ligation penalty of 30, "Current Methods in Sequence Comparison and Analysis", Macromolecule Sequencing and Synthesis, Selected Methods and Applications, pp. 127-149 (1988), Alan R. Liss, Inc.

[0137] An example of a useful algorithm is PILEUP. PILEUP creates a multiple sequence alignment from a group of related sequences using progressive, pairwise alignments. It can also draw a dendrogram showing the clustering relationships used to create the alignment. PILEUP uses a simplification of the progressive alignment method of Feng and Doolittle, 1987, J. Mol. Evol. 35:351-360; this method is similar to the method described by Higgins and Sharp, 1989, CABIOS 5:151-153. Useful PILEUP parameters include a default gap weight of 3.00, a default gap length weight of 0.10, and weighted end gaps.

[0138] Another example of a useful algorithm is the BLAST algorithm, described in Altschul et al., 1990, J. Mol. Biol. 215:403-410; Altschul et al., 1997, Nucleic Acids Res. 25:3389-3402; and Karin et al., 1993, Proc. Natl. Acad. Sci. USA 90:5873-5787. A particularly useful BLAST program is the WU-BLAST-2 program obtained from Altschul et al., 1996, Methods in Enzymology 266:460-480. WU-BLAST-2 uses several search parameters, most of which are set to default values. The adjustable parameters are set to the following values: overlap interval = 1, overlap fraction = 0.125, word threshold (T) = II. The HSP S and HSP S2 parameters are dynamic values ​​and are established by the program itself based on the composition of the particular sequence and the composition of the particular database against which the sequence of interest is being searched; however, these values ​​can be adjusted to increase sensitivity.

[0139] Another useful algorithm is Gapped BLAST as reported by Altschul et al., 1993, Nucl. Acids Res. 25:3389-3402. Gapped BLAST uses BLOSUM-62 instead of scoring; the threshold T parameter is set to 9; a double-click method is used to trigger non-gap extensions, with a cost of 10+k for a gap length of k; Xu is set to 16, and Xg is set to 40 (for the database search phase) and 67 (for the output phase of the algorithm). Gapped alignments are triggered by a score corresponding to approximately 22 bits.

[0140] Generally speaking, the amino acid homology, similarity or identity between each variant CDR or VH / VL sequence and the sequence depicted herein is at least 60%, and more typically has at least 65% or 70%, more preferably at least 75% or 80%, even more preferably at least 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% and almost 100% homology or identity that is preferably increased. In a similar manner, "percent (%) nucleic acid sequence identity" relative to the nucleic acid sequence of the binding protein identified herein is defined as the percentage of nucleotide residues in the candidate sequence that are identical to the nucleotide residues in the coding sequence of the antibody construct. Specific methods utilize the BLASTN module of WU-BLAST-2 set to default parameters, and the overlap interval and overlap fraction are set to 1 and 0.125, respectively.

[0141] Generally, the nucleic acid sequence homology, similarity or identity between the nucleotide sequence encoding each variant CDR or VH / VL sequence and the nucleotide sequences depicted herein is at least 60%, and more typically has at least 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% and preferably increasing to almost 100% homology or identity. Thus, a "variant CDR" or "variant VH / VL region" has a specified homology, similarity or identity with a parent CDR / VH / VL of the invention and shares a biological function, including but not limited to at least 60%, 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% of the specificity and / or activity of the parent CDR or VH / VL.

[0142] In one embodiment, the antibody construct according to the present invention has a percent identity to the human germline of ≥70% or ≥75%, more preferably ≥80% or ≥85%, even more preferably ≥90%, and most preferably ≥91%, ≥92%, ≥93%, ≥94%, ≥95% or even ≥96%. Identicality to human antibody germline gene products is considered to be an important feature that reduces the risk of therapeutic proteins triggering an immune response to the drug in patients during treatment. Hwang and Foote ("Immunogenicity of engineered antibodies"; Methods 36 (2005) 3-10) demonstrated that the reduction of the non-human portion of the drug antibody construct leads to a reduced risk of inducing anti-drug antibodies in patients during treatment. By comparing numerous clinically evaluated antibody drugs and corresponding immunogenicity data, the following trend was shown: humanization of the V region of the antibody makes the protein immunogenicity (an average of 5.1% of patients) lower than antibodies carrying unchanged non-human V regions (an average of 23.59% of patients). Therefore, protein therapeutics based on V regions in the form of antibody constructs need to have a high degree of identity with human sequences. For the purpose of determining germline identity, the V region of VL can be aligned with the amino acid sequences of human germline V segments and J segments (http: / / vbase.mrc-cpe.cam.ac.uk / ) using Vector NTI software and the amino acid sequence (in percentage) calculated by dividing the identical amino acid residues by the total number of amino acid residues of VL. The same applies to the VH segment (http: / / vbase.mrc-cpe.cam.ac.uk / ), except that VH CDR3 can be excluded due to the high diversity of VH CDR3 and the lack of existing human germline VH CDR3 comparison partners. Recombinant techniques can then be used to increase sequence identity with human antibody germline genes.

[0143] In further embodiments, the bispecific antibody constructs of the present invention exhibit high monomer yields under standard research-scale conditions, for example, in a standard two-step purification process. Preferably, the monomer yield of the antibody constructs according to the present invention is ≥0.25 mg / L supernatant, more preferably ≥0.5 mg / L, even more preferably ≥1 mg / L, and most preferably ≥3 mg / L supernatant.

[0144] Similarly, the yield of the dimeric antibody construct isotype of the antibody construct can be determined, and the monomer percentage can be determined therefrom (i.e., monomer: (monomer + dimer)). The productivity of monomeric and dimeric antibody constructs and the calculated monomer percentage can be obtained, for example, in an SEC purification step of culture supernatants from standardized research-scale production in roller bottles. In one embodiment, the monomer percentage of the antibody construct is ≥80%, more preferably ≥85%, even more preferably ≥90%, and most preferably ≥95%.

[0145] In one embodiment, the preferred plasma stability of the antibody construct (ratio of EC50 with plasma to EC50 without plasma) is ≤5 or ≤4, more preferably ≤3.5 or ≤3, even more preferably ≤2.5 or ≤2, and most preferably ≤1.5 or ≤1. The plasma stability of the antibody construct can be determined by incubating the construct in human plasma at 37°C for 24 hours, followed by incubation at 4 °C for 24 hours. 51 The EC50 is measured in a chromium release cytotoxicity assay to test. The effector cells in the cytotoxicity assay can be stimulated enriched human CD8 positive T cells. The target cells can be, for example, CHO cells transfected with human target cell surface antigens. The effector cell to target cell (E:T) ratio can be selected to be 10:1. The human plasma pool for this purpose is derived from the blood of healthy donors collected by EDTA-coated syringes. Cellular components are removed by centrifugation, and the upper plasma phase is collected and subsequently pooled. As a control, the antibody construct is diluted immediately before the cytotoxicity assay in RPMI-1640 culture medium. Plasma stability is calculated as the ratio of EC50 (after plasma incubation) to EC50 (control).

[0146] It is also preferred that the conversion of monomers to dimers of the antibody constructs of the present invention is low. Conversion can be measured under different conditions and analyzed by high-performance size exclusion chromatography. For example, incubation of monomeric isoforms of the antibody construct can be carried out in an incubator at 37°C and, for example, a concentration of 100 μg / ml or 250 μg / ml for 7 days. Under these conditions, it is preferred that the antibody constructs of the present invention show a dimer percentage of ≤5%, more preferably ≤4%, even more preferably ≤3%, even more preferably ≤2.5%, even more preferably ≤2%, even more preferably ≤1.5% and most preferably ≤1% or ≤0.5% or even 0%.

[0147] It is also preferred that the bispecific antibody constructs of the present invention exist with very low dimer conversion after multiple freeze / thaw cycles. For example, it is 250 μg / ml for the antibody construct monomer to be adjusted to a concentration of, for example, 250 μg / ml in a general formulation buffer, and three freeze / thaw cycles are performed (freezing 30 min at -80°C, subsequently thawing 30 min at room temperature), followed by high performance SEC to determine the percentage of the initial monomer antibody construct that has been converted into a dimeric antibody construct. Preferably, for example, after three freeze / thaw cycles, the dimer percentage of the bispecific antibody construct is ≤5%, more preferably ≤4%, even more preferably ≤3%, even more preferably ≤2.5%, even more preferably ≤2%, even more preferably ≤1.5%, and most preferably ≤1% or even ≤0.5%.

[0148] The bispecific antibody constructs of the present invention preferably exhibit favorable thermal stability with an aggregation temperature of ≥45°C or ≥50°C, more preferably ≥52°C or ≥54°C, even more preferably ≥56°C or ≥57°C, and most preferably ≥58°C or ≥59°C. The thermal stability parameter can be determined based on the antibody aggregation temperature as follows: an antibody solution with a concentration of 250 μg / ml is transferred to a disposable cuvette and placed in a dynamic light scattering (DLS) device. The sample is heated from 40°C to 70°C at a heating rate of 0.5°C / min, and the measured radius is constantly acquired. The aggregation temperature of the antibody is calculated using the increase in radius indicating the melting of the protein and aggregates.

[0149] Alternatively, the temperature melting curve can be determined by differential scanning calorimetry (DSC) to determine the intrinsic biophysical protein stability of the antibody construct. These experiments were performed using a VP-DSC device from MicroCal LLC (Northampton, MA, USA). Compared to samples containing only formulation buffer, the energy uptake of samples containing the antibody construct was recorded from 20°C to 90°C. For example, the antibody construct was adjusted to a final concentration of 250 μg / ml in SEC running buffer. In order to record the corresponding melting curve, the entire sample temperature was gradually increased. At each temperature T, the energy uptake of the sample and formulation buffer reference was recorded. The energy uptake Cp (kcal / mole / °C) of the sample minus the difference in reference was plotted for the corresponding temperature. The melting temperature was defined as the temperature at the time of the first maximum energy uptake.

[0150] It is also envisaged that the target cell surface antigen xCD3 bispecific antibody constructs of the invention have a turbidity of ≤0.2, preferably ≤0.15, more preferably ≤0.12, even more preferably ≤0.1 and most preferably ≤0.08 (as measured by OD340 after concentration of the purified monomeric antibody construct to 2.5 mg / ml and overnight incubation).

[0151] It is further envisaged that the bispecific antibody constructs of the invention exhibit therapeutic efficacy or anti-tumour activity. This can be assessed, for example, in studies as described in the following examples of advanced human tumour xenograft models:

[0152] Those skilled in the art will know how to modify or adjust certain parameters of the study, such as the number of tumor cells injected, the injection site, the number of transplanted human T cells, the amount of bispecific antibody construct to be administered, and the timeline, while still obtaining meaningful and reproducible results. Preferably, the tumor growth inhibition T / C [%] is ≤70 or ≤60, more preferably ≤50 or ≤40, even more preferably ≤30 or ≤20, and most preferably ≤10 or ≤5 or even ≤2.5.

[0153] In a preferred embodiment of the antibody construct of the present invention, the antibody construct is a single-chain antibody construct.

[0154] In addition, in a preferred embodiment of the antibody construct of the present invention, the third domain comprises, in amino to carboxyl order:

[0155] Hinge-CH2-CH3-Joint-Hinge-CH2-CH3.

[0156] In addition, in one embodiment of the present invention, the CH2 domain of one or preferably each (two) polypeptide monomers of the third domain comprises an intradomain cysteine ​​disulfide bridge. As known in the art, the term "cysteine ​​disulfide bridge" refers to a functional group having the general structure R-S-S-R. This connection is also known as an SS bond or a disulfide bridge and is derived by coupling the two thiol groups of a cysteine ​​residue. It is particularly preferred for the antibody construct of the present invention that the cysteines that form the cysteine ​​disulfide bridge in the mature antibody construct are introduced into the amino acid sequence of the CH2 domain corresponding to 309 and 321 (Kabat numbering).

[0157] In one embodiment of the present invention, the glycosylation site in Kabat position 314 of the CH2 domain is removed. Preferably, the removal of the glycosylation site is achieved by a N314X substitution, wherein X is any amino acid except Q. Preferably, the substitution is an N314G substitution. In a more preferred embodiment, the CH2 domain further comprises the following substitutions (according to Kabat positions): V321C and R309C (these substitutions introduce intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321).

[0158] It is assumed that preferred features of the antibody constructs of the present invention, compared to, for example, bispecific heterologous Fc antibody constructs known in the art ( FIG. 1 b ), may relate in particular to the introduction of the above-mentioned modifications in the CH2 domain. Thus, it is preferred for the constructs of the present invention that the CH2 domain in the third domain of the antibody constructs of the present invention comprises intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321 and / or that the glycosylation site at Kabat position 314 is removed by the above-mentioned N314X substitution, preferably by the N314G substitution.

[0159] In another preferred embodiment of the present invention, the CH2 domain in the third domain of the antibody construct of the present invention comprises intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321 and the glycosylation site at Kabat position 314 is removed by N314G substitution.

[0160] In one embodiment, the present invention provides an antibody construct, wherein:

[0161] (i) the first domain comprises two antibody variable domains, and the second domain comprises two antibody variable domains;

[0162] (ii) the first domain comprises one antibody variable domain, and the second domain comprises two antibody variable domains;

[0163] (iii) the first domain comprises two antibody variable domains, and the second domain comprises one antibody variable domain; or

[0164] (iv) the first domain comprises an antibody variable domain, and the second domain comprises an antibody variable domain.

[0165] In some embodiments, the first and second domains can be binding domains that comprise two antibody variable domains (such as VH and VL domains). Examples of such binding domains that comprise two antibody variable domains have been described above and include Fv fragments, scFv fragments or Fab fragments as described above. Alternatively, one or both of these binding domains can only comprise a single variable domain. Examples of such single domain binding domains have been described above and include nano antibodies or single variable domain antibodies that only comprise a variable domain, which can be VHH, VH or the VL that are independent of other V districts or domain-specific binding antigens or epi-positions.

[0166] In a preferred embodiment of the antibody construct of the present invention, the first domain and the second domain are fused to the third domain via a peptide linker. Preferred peptide linkers have been described above and are characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e., Gly4Ser (SEQ ID NO: 187), or a polymer thereof, i.e., (Gly4Ser) x, wherein x is an integer of 1 or greater (e.g., 2 or 3). A particularly preferred linker for fusion of the first domain, the second domain, and the third domain is depicted in SEQ ID No: 1.

[0167] In a preferred embodiment, the antibody construct of the present invention is characterized in that it comprises, in order from amino to carboxyl:

[0168] (a) the first domain;

[0169] (b) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID Nos: 187-189;

[0170] (c) the second domain;

[0171] (d) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NO: 187, 188, 189, 195, 196, 197 and 198;

[0172] (e) a first polypeptide monomer of the third domain;

[0173] (f) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID Nos: 191, 192, 193 and 194; and

[0174] (g) a second polypeptide monomer of the third domain.

[0175] In one aspect of the present invention, the target cell surface antigen bound by the first domain is a tumor antigen, an antigen specific for an immune disorder, or a viral antigen. As used herein, the term "tumor antigen" may be understood as referring to those antigens presented on tumor cells. These antigens may be presented on the surface of cells having an extracellular portion that is typically combined with the transmembrane and cytoplasmic portions of the molecule. These antigens can sometimes only be presented by tumor cells, but never by normal cells. Compared to normal cells, tumor antigens may only be expressed on tumor cells or may represent tumor-specific mutations. In this case, they are referred to as tumor-specific antigens. More common are antigens presented by tumor cells and normal cells, and they are referred to as tumor-associated antigens. Compared to normal cells, these tumor-associated antigens may be overexpressed, or because the structure of tumor tissue is less tight than that of normal tissue, they are accessible to antibody binding in tumor cells. Non-limiting examples of tumor antigens as used herein are CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, MUC17, CLDN18.2, CDH3, CD70, BCMA, and PSMA.

[0176] In a preferred embodiment of the antibody construct of the present invention, the tumor antigen, preferably the tumor antigen is selected from the group consisting of CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, MUC17, CLDN18.2, CDH3, CD70, BCMA and PSMA.

[0177] In one aspect of the invention, the antibody construct comprises, in amino to carboxyl order:

[0178] (a) a first domain having an amino acid sequence selected from the group consisting of: SEQ ID No: 7, 8, 17, 27, 28, 37, 38, 39, 40, 41, 48, 49, 50, 51, 52, 59, 60, 61, 62, 63, 64, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 89, 90, 91, 92, 93, 100, 101, 102, 103, 104, 1 13, 114, 121, 122, 123, 124, 125, 131, 132, 133, 134, 135, 136, 143, 144, 145, 146, 147, 148, 149, 150, 151, 158, 159, 160, 161, 162, 163, 164, 165, 166, 173, 174, 175, 176, 177, 178, 179, 180, 181, 223, 235 and 246,

[0179] (b) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID Nos: 187-189;

[0180] (c) a second domain having an amino acid sequence selected from the group consisting of SEQ ID No: 23, 25, 41, 43, 59, 61, 77, 79, 95, 97, 113, 115, 131, 133, 149, 151, 167, 169, 185 or 187 of WO 2008 / 119567 or SEQ ID NO: 202;

[0181] (d) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID No: 187, 188, 189, 195, 196, 197 and 198;

[0182] (e) a first polypeptide monomer of the third domain, the first polypeptide monomer having a polypeptide sequence selected from the group consisting of: SEQ ID No: 249-256;

[0183] (f) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID Nos: 191, 192, 193 and 194; and

[0184] (g) a second polypeptide monomer of the third domain, the second polypeptide monomer having a polypeptide sequence selected from the group consisting of SEQ ID Nos: 249-256.

[0185] In one aspect, the bispecific antibody construct of the invention is characterized by having an amino acid sequence selected from the group consisting of:

[0186] (a) SEQ ID Nos: 27, 28, 37 to 41; CD33

[0187] (b) Each of SEQ ID Nos: 48 to 52; EGFRvIII

[0188] (c) Each of SEQ ID Nos: 59 to 64; MSLN

[0189] (d) Each CDH19 in SEQ ID No: 71 to 82

[0190] (e) Each DLL3 in SEQ ID Nos: 100 to 104

[0191] (f) SEQ ID No: 7, 8, 17, 113 and 114 CD19

[0192] (g) Each FLT3 in SEQ ID Nos: 89 to 93

[0193] (h) Each CDH3 in SEQ ID Nos: 121 to 125

[0194] (i) Each BCMA in SEQ ID No: 132 to 136

[0195] (j) PSMA of each of SEQ ID Nos: 143 to 151, 158 to 166, and 173 to 181

[0196] (k) SEQ ID NO 213MUC17

[0197] (1) Each of CLDN18.2 and SEQ ID NO: 225 and 237

[0198] (m)SEQ ID No:248CD70

[0199] In some embodiments, the bispecific antibody construct comprises a first binding domain that binds to DLL3 that comprises an anti-DLL3 variable light chain domain.

[0200] In some embodiments, the bispecific antibody construct comprises a first binding domain that binds BCMA that comprises an anti-BCMA variable light chain domain.

[0201] In some embodiments, the bispecific antibody construct comprises a first binding domain that binds CD33 that comprises an anti-CD33 variable light chain domain.

[0202] In some aspects of the present disclosure, the protein (such as an antibody or bispecific antibody construct) is present in the liquid formulation (before lyophilization) in an amount ranging from about 0.1 mg / mL to about 100 mg / mL (or about 0.1 mg / mL, 0.5 mg / mL, 1 mg / mL, 5 mg / mL, 10 mg / mL, 15 mg / mL, 20 mg / mL, 25 mg / mL, 30 mg / mL, 35 mg / mL, 40 mg / mL, 45 mg / mL, 50 mg / mL, 55 mg / mL, 60 mg / mL, 65 mg / mL, 70 mg / mL, 75 mg / mL, 80 mg / mL, 85 mg / mL, 90 mg / mL, 95 mg / mL, or 100 mg / mL). For example, the protein is optionally present in the liquid formulation in an amount ranging from about 0.1 mg / mL to about 70 mg / mL. In some cases, the protein is present in the liquid formulation in an amount ranging from about 0.5 mg / mL to about 30 mg / mL (or about 0.5 mg / mL, 0.6 mg / mL, 0.7 mg / mL, 0.8 mg / mL, 0.9 mg / mL, 1 mg / mL, 2 mg / mL, 3 mg / mL, 4 mg / mL, 5 mg / mL, 6 mg / mL, 7 mg / mL, 8 mg / mL, 9 mg / mL, 10 mg / mL, 11 mg / mL, 12 mg / mL, 13 mg / mL, 14 mg / mL, 15 mg / mL, 16 mg / mL, 17 mg / mL, 18 mg / mL, 19 mg / mL, 20 mg / mL, 21 mg / mL, 22 mg / mL, 23 mg / mL, 24 mg / mL, 25 mg / mL, 26 mg / mL, 27 mg / mL, 28 mg / mL, 29 mg / mL, or 30 mg / mL). In each case, the protein is present at a concentration of from about 1 mg / mL to about 20 mg / mL (or about 1 mg / mL, 1.5 mg / mL, 2 mg / mL, 2.5 mg / mL, 3 mg / mL, 3.5 mg / mL, 4 mg / mL, 4.5 mg / mL, 5 mg / mL, 5.5 mg / mL, 6 mg / mL, 6.5 mg / mL, 7 mg / mL, 7.5 mg / mL, 8 mg / mL, 8.5 mg / mL, 9 mg / mL, 9.5 mg / mL, 10 mg / mL, 10.5 mg / mL). g / mL, 11 mg / mL, 11.5 mg / mL, 12 mg / mL, 12.5 mg / mL, 13 mg / mL, 13.5 mg / mL, 14 mg / mL, 14.5 mg / mL, 15 mg / mL, 15.5 mg / mL, 16 mg / mL, 16.5 mg / mL, 17 mg / mL, 17.5 mg / mL, 18 mg / mL, 18.5 mg / mL, 19 mg / mL, 19.5 mg / mL, or 20 mg / mL) is present in the liquid formulation.In some aspects, the protein is present in the liquid formulation in an amount of about 21 mg / mL.

[0203] The protein formulations of the present disclosure optionally include sugar. In certain embodiments, sugar is a monosaccharide or a disaccharide. Suitable sugars include, for example, glucose, galactose, fructose, xylose, sucrose, lactose, maltose, trehalose, or any combination thereof. In some cases, sugar includes trehalose.

[0204] In some aspects, liquid preparation (before lyophilization) comprises concentration and it is about 1% to about 15%w / v or about 4% to about 13%w / v or about 6% to about 12%w / v sugar.In certain embodiments, liquid preparation comprises concentration and it is at least 1%, at least 2%, at least 3%, at least 4%, at least 5%, at least 6%, at least 7%, at least 8%, at least 9%, at least 10%, at least 11%, at least 12%, at least 13% or at least 14%w / v sugar.In certain embodiments, liquid preparation comprises concentration and it is about 1%, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, about 10%, about 11%, about 12%, about 13%, about 14% or about 15%w / v sugar. In certain embodiments, liquid preparation comprises the sugar that concentration is about 7%, about 7.5%, about 8%, about 8.5%, about 9%, about 9.5%, about 10%, about 10.5%, about 11%, about 11.5% or about 12%w / v. In certain embodiments, liquid preparation comprises the sugar that concentration is about 7% to about 12%w / v. In some aspects, liquid preparation comprises the sugar that concentration is about 9%w / v. In certain embodiments, sugar is sucrose, and is present in liquid preparation with the concentration from about 6% to about 12%w / v scope.

[0205] The protein formulation of the present disclosure optionally comprises a surfactant. Suitable surfactant comprises polysorbate, poloxamer, polyoxyethylene or any combination thereof. Expected surfactant comprises polysorbate 20, polysorbate 40, polysorbate 60, polysorbate 80, poloxamer 188, poloxamer 407, Triton X-100, polyoxyethylene, PEG3350, PEG 4000 and any combination thereof. In some aspects, the surfactant comprises polysorbate. In some cases, the surfactant is polysorbate 20.

[0206] Protein formulations described herein can include a surfactant or a mixture of surfactants (although this is not required). In some aspects, the liquid formulation (before lyophilization) includes a surfactant at a concentration of about 0.001% to about 5% w / v (or about 0.001% to about 0.5%, or about 0.004% to about 0.5% w / v, or about 0.001% to about 0.01% w / v or about 0.004% to about 0.01% w / v). In some aspects, liquid preparation comprises concentration and is at least 0.001%, at least 0.002%, at least 0.003%, at least 0.004%, at least 0.005%, at least 0.007%, at least 0.01%, at least 0.05%, at least 0.1%, at least 0.2%, at least 0.3%, at least 0.4%, at least 0.5%, at least 0.6%, at least 0.7%, at least 0.8%, at least 0.9%, at least 1.0%, at least 1.5%, at least 2.0%, at least 2.5%, at least 3.0%, at least 3.5%, at least 4.0% or at least 4.5%w / v surfactant. In some aspects, liquid preparation comprises concentration and is about 0.001% to about 0.5%w / v surfactant. In some aspects, liquid preparation comprises concentration and is about 0.001% to about 0.01%w / v surfactant. In some respects, liquid preparation comprises a surfactant that concentration is about 0.001% to about 0.01% w / v. In some respects, liquid preparation comprises a surfactant that concentration is about 0.001%, about 0.002%, about 0.003%, about 0.004%, about 0.005%, about 0.006%, about 0.007%, about 0.008%, about 0.009%, about 0.01%, about 0.05%, about 0.1%, about 0.2%, about 0.3%, about 0.4% to about 0.5% w / v. In some respects, liquid preparation comprises a surfactant that concentration is about 0.001% to about 0.01% w / v.

[0207] The protein formulations of this disclosure optionally include buffer. Suitable buffers include acetate buffer, glutamate buffer, citrate buffer, lactate buffer, succinate buffer, tartrate buffer, fumarate buffer, maleate buffer, histidine buffer, phosphate buffer, 2-(N-morpholino) ethanesulfonate buffer or any combination thereof. In some cases, the buffer includes histidine.

[0208] Buffers are typically used to control the pH in a formulation. In some aspects, buffer is added at a concentration to maintain the pH of the liquid formulation at about 3 to about 7, or about 4 to about 6, about 4 to 5, or about 5 to about 6, or about 6 to about 6.5. The effect of pH on the formulation can be characterized using any one or more of several methods, such as accelerated stability studies and calorimetric screening studies (Remmele RL Jr., et al., Biochemistry, 38(16):5241-7(1999)).

[0209] The buffer system (when present in a protein formulation) is selected to be physiologically compatible and to maintain the desired pH. The buffer can be present in the liquid formulation (before lyophilization) at a concentration between about 0.1 mM and about 1000 mM (1 M), or between about 5 mM and about 200 mM, or between about 5 mM and about 100 mM, or between about 10 mM and about 50 mM. Suitable buffer concentrations encompass about 200 mM or lower concentrations. In some aspects, the buffer in the liquid protein formulation (before lyophilization) is present at a concentration of about 190 mM, about 180 mM, about 170 mM, about 160 mM, about 150 mM, about 140 mM, about 130 mM, about 120 mM, about 110 mM, about 100 mM, about 80 mM, about 70 mM, about 60 mM, about 50 mM, about 40 mM, about 30 mM, about 20 mM, about 10 mM, or about 5 mM. 70, 80, or 90 mM and 100 mM. In some aspects, the concentration of the buffer is between 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, or 40 mM and 50 mM.

[0210] As another aspect, a test kit is provided, which comprises a freeze-dried protein composition as described herein packaged in a manner that is beneficial to administering to a subject. In one aspect, the test kit includes a freeze-dried protein composition as described herein, which is packaged in a container (e.g., a sealed bottle, a vial, a single-use or multiple-use vial, a pre-filled device (e.g., a syringe) or a pre-filled injection device), and the test kit optionally has a label attached to the container or included in the packaging, which describes the purposes of the freeze-dried protein composition. In one aspect, the pharmaceutical composition is packaged as a unit dosage form. The test kit may include a device suitable for administering the reconstructed protein composition according to a specific route of administration, although this is not essential. For example, the present disclosure provides a dual-chamber device for delivering the reconstructed protein composition disclosed herein to a subject in need. The dual-chamber device is a combination product containing the freeze-dried protein composition disclosed herein and a diluent in two separate chambers of the device. The pre-filled dual-chamber device is a combination product containing freeze-dried medicine and a diluent in two separate chambers of the device. The art describes a suitable dual-chamber device for use with the present disclosure. See, e.g., Ingle R., Fang W. (2021). Int. Journal of Pharmaceutics 597, 12031.

[0211] The formulations described herein can be used as pharmaceutical formulations for treating or ameliorating cancer in subjects in need thereof. The terms "subject in need thereof" or "those in need of treatment" include subjects already suffering from the disorder, as well as those in whom the disorder is to be prevented. "Subject in need thereof" or "patient" include humans and other mammalian subjects receiving prophylactic or therapeutic treatment. "Treatment" does not require complete remission or eradication of the disease; any improvement in the disease and / or improvement in symptoms associated with the disease are contemplated. For example, a therapeutic response refers to one or more of the following improvements in the disease: (1) a decrease in the number of neoplastic cells; (2) an increase in neoplastic cell death; (3) inhibition of neoplastic cell survival; (4) inhibition of tumor growth or the appearance of new lesions (i.e., slowing to some extent, preferably stopping); (5) slowing of disease progression; (6) improvement in patient survival; (7) downgrading of cancer stage (e.g., from stage 2 to stage 1); and / or (8) relief of one or more symptoms associated with the disease or condition. The compositions can also be administered to achieve disease prevention or to slow the onset of disease, such as avoiding the occurrence or recurrence of a tumor or cancer. Disease status is monitored by, for example, clinical examination, X-ray, computed tomography (CT, such as spiral CT), magnetic resonance imaging (MRI), positron emission tomography (PET), ultrasound, endoscopy and laparoscopy, tumor marker levels (e.g., carcinoembryonic antigen (CEA)), cytology, histology, tumor biopsy sampling, and / or counting of circulating tumor cells. These methods are also typically used to diagnose and stage cancer.

[0212] Present disclosure provides a method for treating cancer, the method including administering a therapeutically effective amount of a reconstructed composition based on a lyophilized formulation as described herein to a subject in need. In certain embodiments, the subject is a person. In some aspects, cancer is a solid tumor. In certain embodiments, cancer is brain cancer, bladder cancer (bladder cancer), breast cancer (e.g., triple-negative breast cancer), clear cell renal carcinoma, cervical cancer, colon and rectal cancer, endometrial cancer, gastric cancer (gastric cancer), head / neck squamous cell carcinoma, lip and oral cancer, liver cancer, lung squamous cell carcinoma, melanoma, mesothelioma, non-small cell lung cancer (NSCLC), non-melanoma skin cancer, ovarian cancer, oral cancer, pancreatic cancer, prostate cancer, neuroendocrine prostate cancer, renal cell carcinoma, sarcoma, small cell lung cancer (SCLC), head and neck squamous cell carcinoma (SCCHN) or thyroid cancer. In some aspects, cancer is acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL) or chronic myeloid leukemia (CML). In some aspects, the cancer is diffuse large B-cell lymphoma (DLBCL), follicular lymphoma, Hodgkin lymphoma (HL), mantle cell lymphoma (MCL), multiple myeloma (MM), myelodysplastic syndrome (MDS), non-Hodgkin lymphoma (NHL) or small lymphocytic lymphoma (SLL). The present disclosure also provides the use of a reconstituted composition based on a lyophilized formulation in a method for treating cancer, and the use of a lyophilized formulation in the preparation of a medicament for treating cancer.

[0213] Preferably, the pharmaceutical formulation is administered parenterally, e.g., intravenously, subcutaneously, intratumorally, or intramuscularly. Parenteral administration can be by injection, e.g., bolus injection, or by infusion, e.g., continuous infusion. Administration can be achieved via a reservoir to achieve long-term release. In some embodiments, the formulation is administered intravenously by an initial bolus followed by continuous infusion to maintain therapeutic circulating levels of the drug product. In some embodiments, the formulation is administered in a disposable dose. The pharmaceutical formulation can be administered using a medical device. Examples of medical devices for administering pharmaceutical formulations are described in U.S. Patent Nos. 4,475,196, 4,439,196, 4,447,224, 4,447,233, 4,486,194, 4,487,603, 4,596,556, 4,790,824, 4,941,880, 5,064,413, 5,312,335, 5,312,335, 5,383,851, and 5,399,163.

[0214] The following examples illustrate representative features of the present disclosure. From the description of these aspects, other aspects of the present invention can be accomplished and / or practiced based on the description provided below. These methods involve the use of molecular biology techniques, which are described in, for example, the following monographs: Molecular Cloning: A Laboratory Manual [Molecular Cloning: Laboratory Manual] 2nd Edition, Volumes 1-3, Sambrook et al., eds., Cold Spring Harbor Laboratory Press [Cold Spring Harbor Laboratory Press], Cold Spring Harbor, New York, 2001; and Current Protocols in Molecular Biology [Molecular Biology Modern Methods], Ausubel et al., eds., Greene Publishing and Wiley-Interscience [Greene Publishing and Wiley Interscience Publishers], New York. The examples are merely illustrative of the present invention and are not intended to limit the scope of the present invention in any way.

[0215] Examples

[0216] This example describes methods for preparing lyophilized bispecific molecule compositions according to an alternative method including an annealing step and according to the present invention including a CIN step, the latter resulting in desirable product properties of a low percentage of HMWS and desired process times.

[0217] Annealing and controlled ice nucleation are two approaches to addressing the challenges associated with freeze-drying, including aggregation and long processing times. Both annealing and controlled ice nucleation address this issue by producing larger ice crystals to reduce ice heterogeneity, resulting in a porous filter cake matrix with lower mass transfer resistance during sublimation.

[0218] Annealing involves holding the product between the glass transition temperature and the eutectic melting temperature after random ice nucleation (Figure 1A). This melts the ice crystals, with the smallest ice crystals disappearing completely, leaving behind larger ice crystals. Subsequent freezing further allows the large crystals to grow. Larger ice crystals result in larger pores in the freeze-dried filter cake, which corresponds to lower resistance to vapor passage through the drying layer and faster primary drying. However, holding the product above the glass transition temperature can promote the formation of high molecular weight species (HMWS) for some products due to increased protein mobility and accessibility. For bispecific molecules with protein concentrations above 5 mg / mL (e.g. In addition, annealing prolongs the freezing phase of lyophilization, which partially offsets the reduction in primary drying time.

[0219] Controlled ice nucleation (CIN) uses different methods to produce larger ice crystals. For example, a CIN technique involves injecting crystal seeds to trigger ice nucleation (Figure 1B). This triggers nucleation when the sample is only slightly supercooled, and the isothermal hold after nucleation can be used to further grow the crystals. Compared to the ice crystals produced in random nucleation, a lower degree of supercooling produces larger ice crystals, and compared to annealing, a shorter isothermal hold can reduce the risk of HMWS formation. In this example, single-chain bispecific T cell engager molecule freeze-drying cycles without a CIN step, with and without annealing were compared with CIN freeze-drying to evaluate bispecific molecule drug products with "high protein concentration" (defined as 15 mg / mL or higher in this example). BCMAxCD3 bispecific molecules (20 mg / mL) and DLL3xCD3 bispecific molecules (15 mg / mL) are used as model high-concentration bispecific molecule products. Annealing is currently a typical choice for freeze-dried bispecific molecule drug products. Therefore, such a freeze-drying cycle is used as a representative annealing freeze-drying cycle herein. The annealing step was removed from the lyophilization cycle to create an "anneal-free cycle." The CIN lyophilization cycle, originally developed for the BCMAxCD3 bispecific molecule 1 mg 6R SKU7, was adapted for high-concentration bispecific formulations. Lyophilization cycles were evaluated based on total cycle time and product quality.

[0220] Materials and Methods

[0221] BCMAxCD3 bispecific molecules (20 mg / mL) and DLL3xCD3 bispecific molecules (15 mg / mL) were used in these experiments. Table 1 contains details of the drug substances (DS) used.

[0222] Table 1. Bispecific Drug Substances

[0223]

[0224] The BCMAxCD3 bispecific and DLL3xCD3 bispecific DS were stored at -70°C and -30°C, respectively. Thaw the DS at room temperature for 72 hours before use. Remove a 150 mL aliquot and mix by inversion for 2 minutes. The DS was then filtered through a 0.22 μm polyvinylidene fluoride (PVDF) filter and filled into vials.

[0225] Tables 2 and 3 contain the equipment and vial components used in this study. 6R vials were depyrogenated prior to use. The sterile stoppers were stored in the autoclaved packaging prior to use but did not maintain the sterile seal after opening the packaging.

[0226] Table 2. Equipment used in the experiment

[0227]

[0228] Table 3. Vial Components

[0229]

[0230] Experimental Summary

[0231] The purpose of these experiments was to compare the effects of conventional lyophilization cycles with and without annealing versus CIN lyophilization cycles on model high concentration BCMAxCD3 bispecific molecules (20 mg / mL) and DLL3xCD3 bispecific molecules (15 mg / mL). The bispecific molecule lyophilization cycles were used as conventional lyophilization cycles with and without annealing. The CIN lyophilization cycles were based on past data from the BCMAxCD3 bispecific 1 mg 6R SKU. 7 A summary of all three lyophilization cycles can be found in Table 4.

[0232] Table 4. Summary of freeze-drying cycles

[0233]

[0234] Freeze-drying procedure

[0235] Before filling, 6R vials were used to completely fill the Millrock lyophilization tray. A total of 100 6R vials were filled with 1.3 mL of 20 mg / mL BCMAxCD3 bispecific molecule DS, and another 100 6R vials were filled with 1.3 mL of 15 mg / mL DLL3xCD3 bispecific molecule DS. The 1.3 mL fill solution presented in the 6R vial was selected as a high concentration representative of the BiTE, consisting of 1.28 mL of fill solution of ≤5 mg / mL bispecific molecule formulation in the 6R vial. The two sets of vials were placed side by side in a single tray for lyophilization. Thermocouples were placed in front, middle, and back of the fill sample.

[0236] The freeze-drying cycle parameters for freeze-drying cycles with and without annealing, as well as CIN cycles, are given in Table 5. BCMAxCD3 bispecific molecules. The primary and secondary drying parameters remained unchanged. However, the nucleation conditions were adjusted to minimize the risk of HMWS formation. According to the present invention, this involved lowering the nucleation temperature from a less favorable -7°C to a more favorable -12°C, -15°C, or -17°C, and setting the post-nucleation time to 90 minutes or even reducing the hold time from one hour to 30 or 20 minutes (Figure 3). These changes create a CIN freezing stage, reducing the time the product is in a cryogenically concentrated "slushy" state, thereby reducing the chance of HMWS formation.

[0237] Table 5. Lyophilization cycle parameters

[0238]

[0239] Freeze-drying data analysis

[0240] Freeze-drying process data was exported from the instrument and analyzed for primary drying time, total cycle time, and product temperature. Primary drying was considered to begin when the primary drying temperature ramp began. The end of the primary drying time was determined by observing the instantaneous slope of the Pirani gauge approaching zero. The total freeze-drying time was the sum of the freezing, primary drying, and secondary drying stages of the freeze-drying cycle.

[0241] Determine the maximum product temperature for each vial containing a thermocouple as the highest observed temperature while the thermocouple was within the freeze matrix during the primary drying period. Do this for each thermocouple to obtain the product temperature range for the entire lyophilization tray.

[0242] Assays - BCMAxCD3 Bispecific and DLL3xCD3 Bispecific

[0243] Product quality analysis was performed on lyophilized samples of the BCMAxCD3 bispecific and DLL3xCD3 bispecific molecules for 13 weeks. Samples were analyzed at the initial time point (time zero) immediately after lyophilization. The remaining samples were then randomly sorted (with respect to the lyophilization tray position) and placed at 4°C, 25°C and 40% relative humidity (25°C / 40RH), and 40°C and 65% relative humidity (40°C / 65RH). Samples were removed from storage at designated time points for analytical testing (Tables 5 and 7).

[0244] Table 5. BCMAxCD3 Bispecific Molecular Quality Standards*

[0245]

[0246] BCMAxCD3 bispecific molecules BCMAxCD3 bispecific molecules DLL3xCD3 bispecific molecules

[0247] Results and Discussion

[0248] Freeze-drying results: cycle time and maximum observed temperature

[0249] All freeze-drying cycles were completed successfully. The chamber pressure (as measured by capacitance manometer and Pirani gauge), shelf temperature and product temperature during each freeze-drying cycle are given in FIG2 .

[0250] During the primary drying period, the highest BCMAxCD3 bispecific molecule product temperature for the non-CIN lyophilization cycle (Figure 2A) was -33.3°C, and the highest DLL3xCD3 bispecific molecule product temperature was -33.1°C. At the same shelf temperature set point during the primary drying period, the BCMAxCD3 bispecific molecule product temperature (-31.3°C) for the non-CIN lyophilization cycle without annealing (Figure 2B) was 2°C higher than the product temperature in the non-CIN lyophilization cycle, and the DLL3xCD3 bispecific molecule product temperature (-31.5°C) was 1.6°C higher than the product temperature in the non-CIN lyophilization cycle. This is consistent with the increased resistance to sublimation due to the removal of annealing. However, the non-CIN lyophilization cycles without annealing had similar primary drying endpoints compared to the non-CIN lyophilization cycles with annealing. This indicates that annealing does not provide a time-saving benefit for these high-concentration bispecific molecules using the non-CIN lyophilization cycle.

[0251] Compared to the freeze-drying cycle, the CIN cycle (Figure 2C) with a primary drying shelf set point of -8°C reduced the primary drying time by 50%. CIN can produce faster primary drying times because CIN completely eliminates random freezing to form larger, more uniform ice crystals, which promote faster sublimation through a more porous filter cake structure. 2 However, when lyophilization with CIN was used, the maximum product temperatures observed for the BCMAxCD3 bispecific molecule and the DLL3xCD3 bispecific molecule (-27.8°C and -28.7°C, respectively) were also higher than the previously reported critical temperatures of -30°C for the 1 mg / mL BCMAxCD3 bispecific molecule and the 5 mg / mL DLL3xCD3 bispecific molecule.

[0252] Table 6 lists the shortest primary drying times and shortest total cycle times for the lyophilization cycles tested. Since the primary drying time is determined when the instantaneous slope of the Pirani gauge approaches zero, it is considered the shortest primary drying time without the addition of the safety buffer typically included in lyophilization formulations. Therefore, the total cycle time calculated using the shortest primary drying time is considered the shortest total cycle time. The lyophilization non-CIN cycle with annealing had a longer shortest total cycle time than the standard non-CIN cycle without annealing. This is because both cycles have similar primary and secondary drying times, but the annealing step increases the freezing time of the standard cycle by 7 hours.

[0253] The primary drying endpoint for the CIN cycle is 13 hours, which reduces the minimum total cycle time by 32%. While this may change based on optimization of the primary drying shelf temperature to reduce the maximum product temperature, CIN has the potential to produce a faster primary drying time by forming larger, more uniform ice crystals, resulting in a more porous filter cake structure. This, combined with the shorter freezing phase, makes CIN a better choice in terms of total cycle time.

[0254] Table 6. Laboratory scale primary drying time and total cycle time

[0255]

[0256] HMWS product quality results

[0257] After lyophilization, samples were reconstituted and analyzed for % HMWS by SE-UHPLC, and the results were compared to the pre-lyophilization samples. The difference between the % HMWS of the reconstituted sample and the % HMWS of the pre-lyophilization sample was defined as Δ% HMWS. While all results were within the specification limit of ≤10.0% HMWS, for both the BCMAxCD3 and DLL3xCD3 bispecific molecules, Δ% HMWS upon lyophilization was greatest in samples from the standard cycle with annealing. Samples from the standard cycle without annealing had the least HMWS growth upon lyophilization.

[0258] For both the BCMAxCD3 and DLL3xCD3 bispecific molecules, the Δ% HMWS of the CIN lyophilized samples was slightly higher than that of the standard cycle without annealing, but still approximately 80% lower than that of the standard cycle with annealing. Because the CIN samples spent less time in the partially frozen matrix, the risk of aggregation and HMWS formation was lower in the CIN lyophilization cycle compared to the annealing cycle.

[0259] After storage of lyophilized samples at 4°C, 25°C / 40RH, and 40°C / 65RH, samples were removed at designated time points for % HMWS analysis. The DLL3xCD3 bispecific molecule samples were measured at time zero, 2 weeks, and 4 weeks, and the BCMAxCD3 bispecific molecule samples were measured at time zero and 13 weeks. Although the BCMAxCD3 bispecific molecule samples were initially planned to have time points of 2 and 4 weeks, instrument issues resulted in the measurement time point being changed to 13 weeks.

[0260] None of the lyophilized groups showed an increase in % HMWS over time (Δ% HMWS was not used when comparing stability time points). All BCMAxCD3 bispecific molecule samples (Figure 4A) and DLL3xCD3 bispecific molecule samples (Figure 4B) maintained % HMWS over time. Lyophilization cycles with annealing consistently had higher % HMWS than lyophilization cycles without annealing and CIN cycles.

[0261] Water content of BCMAxCD3 and DLL3xCD3 bispecific molecules

[0262] All moisture results were ≤ 3.0%. Three samples were measured at each time point, with samples from the front, middle, and back of the freeze-drying shelf measured at time zero, and no significant differences (p>0.05) in moisture content were found between samples from the three freeze-drying conditions. Figure 5 shown.

[0263] The CIN cycle in this example used the same primary drying temperature set point as the CIN cycle for the 1 mg 6R BCMAxCD3 bispecific molecule. This resulted in higher product temperatures during primary drying for the high-concentration bispecific molecule in this study than observed in previous studies with the 1 mg / mL product due to higher cake resistance in the high-concentration formulation. As discussed above, the product temperature of the CIN sample exceeded the critical temperature of 30°C for previously characterized CIN lyophilized bispecific molecules. Exceeding this temperature during primary drying resulted in an increase in moisture content over time, as observed in previous studies using 1 mg / mL and 5 mg / mL BiTE standard products, and this was also observed in this example. Optimizing the primary drying temperature for the CIN lyophilization cycle may alleviate this issue.

[0264] CEX results for BCMAxCD3 and DLL3xCD3 bispecific molecules

[0265] CEX-HPLC analysis of the BCMAxCD3 and DLL3xCD3 bispecific molecules showed minimal differences between samples from each lyophilization cycle. A slight decrease in the main peak area and a corresponding slight increase in the acidic and basic peak areas were observed in the BCMAxCD3 bispecific molecule samples stored at 40°C / 65RH for 13 weeks, but this phenomenon was observed regardless of the lyophilization cycle (Figures 6A-6C). At 4 weeks, the same behavior was observed for the DLL3xCD3 bispecific molecule samples (Figures 6D-6F). Therefore, no differences were observed between lyophilization cycles with and without annealing and CIN cycles.

[0266] BCMAxCD3 bispecific and DLL3xCD3 bispecific HIAC particle quantification results

[0267] Particle quantification of the BCMAxCD3 bispecific molecules (Table 7) and DLL3xCD3 bispecific molecules (Table 8) showed that the particle amounts were similar between samples from lyophilization cycles with and without annealing and samples from CIN cycles. All particle counts were well below the quality standard limits of 6000 particles / vial ≥10 μm and 600 particles / vial ≥25 μm.

[0268] Table 7. BCMAxCD3 Bispecific Molecule HIAC Particle Quantification

[0269]

[0270]

[0271] Table 8. DLL3xCD3 bispecific molecule HIAC particle quantification

[0272]

[0273] pH and osmotic pressure results for BCMAxCD3 and DLL3xCD3 bispecific molecules

[0274] At time zero, pH and osmolality were measured for reconstituted samples from standard cycles with and without annealing and CIN cycles. Osmolality and pH were tested to confirm product formulation and reconstitution. No differences in pH or osmolality were observed between pre-lyophilized and post-lyophilized reconstituted samples for the BCMAxCD3 bispecific molecules (Table 9) and DLL3xCD3 bispecific molecules (Table 10). Therefore, pH and osmolality were not monitored at any other time points during stability testing.

[0275] Table 9. pH and osmolarity measurements of BCMAxCD3 bispecific molecules

[0276]

[0277] Table 10. pH and osmolarity measurements of DLL3xCD3 bispecific molecules

[0278]

[0279] in conclusion

[0280] Controlled ice-nucleated lyophilization of high-concentration products was evaluated using BCMAxCD3 bispecific (20 mg / mL) and DLL3xCD3 bispecific (15 mg / mL) in 1.3 mL of fill solution in 6R vials as model proteins. CIN lyophilization avoids the high levels of HMWS typically observed when using lyophilized non-CIN cycles with annealing for these products. During 13 weeks of product quality testing, CIN lyophilized samples also had lower HMWS than annealed lyophilized samples, regardless of storage conditions.

[0281] CIN lyophilization also reduced the minimum total cycle time by 32% compared to lyophilized non-CIN cycles of the bispecific molecule. Optimization of primary drying parameters resulted in smaller reductions in the minimum total cycle time. However, CIN lyophilization primarily offers the benefit of reducing HMWS formation in highly concentrated bispecific molecules, along with the potential benefit of reducing lyophilization cycle time, making it an attractive option compared to lyophilized non-CIN cycles with annealing.

[0282] All references cited herein (including patents, patent applications, literature publications, etc.) are hereby incorporated by reference in their entirety.

[0283] Although the present invention has been described with emphasis on preferred embodiments, it will be apparent to those skilled in the art that variations of the preferred compounds and methods may be used, and that the invention is intended to be practiced in ways other than as specifically described herein. Therefore, the present invention includes all modifications encompassed within the spirit and scope of the invention as defined by the following claims.

[0284] Table 13. Sequence Listing

[0285]

[0286]

[0287]

[0288]

[0289]

[0290]

[0291]

[0292]

[0293]

[0294]

[0295]

[0296]

[0297]

[0298]

[0299]

[0300]

[0301]

[0302]

[0303]

[0304]

[0305]

[0306]

[0307]

[0308]

[0309]

[0310]

[0311]

[0312]

[0313]

[0314]

[0315]

[0316]

[0317]

[0318]

[0319]

[0320]

[0321]

[0322]

[0323]

[0324]

[0325]

[0326]

[0327]

[0328]

[0329]

[0330]

[0331]

[0332]

[0333]

[0334]

[0335]

[0336]

[0337]

[0338]

Claims

1. A method for preparing a lyophilized bispecific molecule composition, the method comprising (a) inducing ice nucleation in a liquid bispecific molecule composition having a bispecific molecule concentration of at least about 10 mg / ml in a vial exposed to a first temperature of about -18°C to about -10°C for a first time period of about 60 minutes to about 270 minutes; (b) exposing the vial of step (a) to a second temperature of about -25°C to -50°C for a second period of about 1 hour to about 5 hours; (c) drying the composition of step (b) at a third temperature of about -5°C to about -25°C for a third period of time of about 25 hours to about 70 hours; and (d) further drying the composition of step (c) at a fourth temperature of about 25° C. to about 50° C. for a fourth period of about 4 hours to about 12 hours to provide a vial comprising a lyophilized bispecific molecule composition having a high molecular weight species (HMWS) percentage content of less than or equal to 1.5% (m / V), wherein the bispecific molecule comprises at least three domains, wherein: The first domain binds to tumor antigens on target cells; The second domain binds to an extracellular epitope of the human and / or macaque CD3ε chain; and The third domain is fused to the second domain via a peptide linker, the third domain comprising two polypeptide monomers, each polypeptide monomer comprising a hinge, a CH2 domain and a CH3 domain, wherein the two polypeptide monomers are fused to each other via a peptide linker.

2. The method of claim 1, wherein step (a) comprises exposing the vial to the first temperature for about 90 minutes to about three hours.

3. The method of claim 2, wherein step (a) comprises exposing the vial to the first temperature for about 90 minutes to about two hours, preferably 110 minutes.

4. The method of any one of claims 1 to 3, wherein the first temperature in step (a) is from about -12°C to about -17°C.

5. The method of claim 4, wherein the first temperature in step (a) is about -15°C.

6. The method of any one of claims 1-5, wherein step (a) further comprises maintaining the ice-nucleated protein composition at the first temperature for a post-nucleation period of up to 90 minutes. The method of claim 6 , wherein the post-nucleation time period is about 20 minutes to about 90 minutes. The method of claim 7 , wherein the post-nucleation time period is about 30 minutes.

9. The method of any one of claims 1-8, wherein step (b) comprises exposing the vial to a second temperature of about -45°C.

10. The method of any one of claims 1-9, wherein the second period of time is about two hours to about four hours. The method of claim 10 , wherein the second period of time is approximately three hours.

12. The method of any one of claims 1-11, wherein the first temperature is transitioned to the second temperature at a rate of about 0.01°C per minute to about 0.5°C per minute.

13. The method of claim 12, wherein the first temperature is transitioned to the second temperature at a rate of about 0.2°C per minute.

14. The method of any one of claims 1 to 13, wherein the third temperature of step (c) is from about 0°C to about -20°C.

15. The method of claim 14, wherein the third temperature of step (c) is about -5°C to about -10°C.

16. The method of claim 15, wherein the third temperature of step (c) is about -8°C.

17. The method of any one of claims 1-9, wherein step (c) comprises exposing the vial containing the bispecific molecule composition to a temperature increase at a rate of about 0.01°C per minute to about 0.5°C per minute.

18. The method of any one of claims 1-17, wherein transitioning from step (b) to step (c) comprises increasing the temperature at a rate of about 0.2°C per minute to about 0.7°C per minute and maintaining the vial at a temperature of about -40°C to about -30°C for about 15 minutes to about one hour.

19. The method of any one of claims 1 to 18, wherein the liquid bispecific molecule composition is exposed in step (a) to a first temperature of about -15°C for about 110 minutes plus about 30 minutes after nucleation, and in step (b) to a second temperature of about -45°C for about 3 hours; in step (c) the composition of step (b) is dried at a third temperature of about -8°C for about 50 hours; and the composition of (c) is further dried at a fourth temperature of about 40°C for about 8 hours.

20. The method of any one of claims 1-18, wherein the bispecific molecule is a single chain molecule.

21. The method of any one of claims 1-18, wherein the bispecific molecule is present in the composition at a concentration of about 10 mg / ml to about 30 mg / ml or about 20 mg / ml to about 30 mg / ml, preferably about 15 mg / ml to about 25 mg / ml.

22. The method of any one of claims 1-18, wherein the method results in a vial comprising a lyophilized bispecific molecule composition that exhibits aggregation of less than or equal to about 1.5% (m / V) or preferably less than or equal to about 11% (m / V) in terms of high molecular weight species (HMSW) formation.

23. The method of any one of claims 1 to 22, wherein the third domain comprises, in amino to carboxyl order: Hinge-CH2-CH3-Joint-Hinge-CH2-CH3.

24. The method of any one of claims 1 to 23, wherein each of the polypeptide monomers of the third domain has an amino acid sequence that is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 249-256, or an amino acid sequence that is identical to a sequence selected from the group consisting of SEQ ID NOs: 249-256.

25. The method of any one of claims 1-24, wherein the first domain binds to CD33, CDH19, MSLN, FLT3, BCMA, CD19, MUC17, CDH3, CLDN18.2, CD70, EGFRviii, EpCAM, DLL3, and / or PSMA.

26. The method of any one of claims 1-25, wherein the first binding domain of the bispecific molecule comprises a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of: and a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from the group consisting of: (a) CDR-H1 as depicted in SEQ ID NO: 4, CDR-H2 as depicted in SEQ ID NO: 5, CDR-H3 as depicted in SEQ ID NO: 6, CDR-L1 as depicted in SEQ ID NO: 1, CDR-L2 as depicted in SEQ ID NO: 2, and CDR-L3 as depicted in SEQ ID NO: 3, (b) CDR-H1 as depicted in SEQ ID NO: 29, CDR-H2 as depicted in SEQ ID NO: 30, CDR-H3 as depicted in SEQ ID NO: 31, CDR-L1 as depicted in SEQ ID NO: 34, CDR-L2 as depicted in SEQ ID NO: 35, and CDR-L3 as depicted in SEQ ID NO: 36, (c) CDR-H1 as depicted in SEQ ID NO:42, CDR-H2 as depicted in SEQ ID NO:43, CDR-H3 as depicted in SEQ ID NO:44, CDR-L1 as depicted in SEQ ID NO:45, CDR-L2 as depicted in SEQ ID NO:46, and CDR-L3 as depicted in SEQ ID NO:47, (d) CDR-H1 as depicted in SEQ ID NO: 53, CDR-H2 as depicted in SEQ ID NO: 54, CDR-H3 as depicted in SEQ ID NO: 55, CDR-L1 as depicted in SEQ ID NO: 56, CDR-L2 as depicted in SEQ ID NO: 57, and CDR-L3 as depicted in SEQ ID NO: 58, (e) CDR-H1 as depicted in SEQ ID NO: 65, CDR-H2 as depicted in SEQ ID NO: 66, CDR-H3 as depicted in SEQ ID NO: 67, CDR-L1 as depicted in SEQ ID NO: 68, CDR-L2 as depicted in SEQ ID NO: 69, and CDR-L3 as depicted in SEQ ID NO: 70, (f) CDR-H1 as depicted in SEQ ID NO: 83, CDR-H2 as depicted in SEQ ID NO: 84, CDR-H3 as depicted in SEQ ID NO: 85, CDR-L1 as depicted in SEQ ID NO: 86, CDR-L2 as depicted in SEQ ID NO: 87, and CDR-L3 as depicted in SEQ ID NO: 88, (g) CDR-H1 as depicted in SEQ ID NO: 94, CDR-H2 as depicted in SEQ ID NO: 95, CDR-H3 as depicted in SEQ ID NO: 96, CDR-L1 as depicted in SEQ ID NO: 97, CDR-L2 as depicted in SEQ ID NO: 98, and CDR-L3 as depicted in SEQ ID NO: 99, (h) CDR-H1 as depicted in SEQ ID NO: 105, CDR-H2 as depicted in SEQ ID NO: 106, CDR-H3 as depicted in SEQ ID NO: 107, CDR-L1 as depicted in SEQ ID NO: 109, CDR-L2 as depicted in SEQ ID NO: 110, and CDR-L3 as depicted in SEQ ID NO: 111, (i) CDR-H1 as depicted in SEQ ID NO: 115, CDR-H2 as depicted in SEQ ID NO: 116, CDR-H3 as depicted in SEQ ID NO: 117, CDR-L1 as depicted in SEQ ID NO: 118, CDR-L2 as depicted in SEQ ID NO: 119, and CDR-L3 as depicted in SEQ ID NO: 120, (j) CDR-H1 as depicted in SEQ ID NO: 126, CDR-H2 as depicted in SEQ ID NO: 127, CDR-H3 as depicted in SEQ ID NO: 128, CDR-L1 as depicted in SEQ ID NO: 129, CDR-L2 as depicted in SEQ ID NO: 130, and CDR-L3 as depicted in SEQ ID NO: 131, (k) CDR-H1 as depicted in SEQ ID NO: 137, CDR-H2 as depicted in SEQ ID NO: 138, CDR-H3 as depicted in SEQ ID NO: 139, CDR-L1 as depicted in SEQ ID NO: 140, CDR-L2 as depicted in SEQ ID NO: 141, and CDR-L3 as depicted in SEQ ID NO: 142, (1) CDR-H1 as depicted in SEQ ID NO: 152, CDR-H2 as depicted in SEQ ID NO: 153, CDR-H3 as depicted in SEQ ID NO: 154, CDR-L1 as depicted in SEQ ID NO: 155, CDR-L2 as depicted in SEQ ID NO: 156, and CDR-L3 as depicted in SEQ ID NO: 157, (m) CDR-H1 as depicted in SEQ ID NO: 167, CDR-H2 as depicted in SEQ ID NO: 168, CDR-H3 as depicted in SEQ ID NO: 169, CDR-L1 as depicted in SEQ ID NO: 170, CDR-L2 as depicted in SEQ ID NO: 171, and CDR-L3 as depicted in SEQ ID NO: 172, (n) CDR-H1 as depicted in SEQ ID NO: 203, CDR-H2 as depicted in SEQ ID NO: 204, CDR-H3 as depicted in SEQ ID NO: 205, CDR-L1 as depicted in SEQ ID NO: 206, CDR-L2 as depicted in SEQ ID NO: 207, and CDR-L3 as depicted in SEQ ID NO: 208; (o) CDR-H1 as depicted in SEQ ID NO: 214, CDR-H2 as depicted in SEQ ID NO: 215, CDR-H3 as depicted in SEQ ID NO: 216, CDR-L1 as depicted in SEQ ID NO: 217, CDR-L2 as depicted in SEQ ID NO: 218, and CDR-L3 as depicted in SEQ ID NO: 219; (p) CDR-H1 as depicted in SEQ ID NO: 226, CDR-H2 as depicted in SEQ ID NO: 227, CDR-H3 as depicted in SEQ ID NO: 228, CDR-L1 as depicted in SEQ ID NO: 229, CDR-L2 as depicted in SEQ ID NO: 230, and CDR-L3 as depicted in SEQ ID NO: 231; and (q) CDR-H1 as depicted in SEQ ID NO: 238, CDR-H2 as depicted in SEQ ID NO: 239, CDR-H3 as depicted in SEQ ID NO: 240, CDR-L1 as depicted in SEQ ID NO: 241, CDR-L2 as depicted in SEQ ID NO: 242, and CDR-L3 as depicted in SEQ ID NO:

243.

27. The method of any one of claims 1-26, wherein the first domain has an amino acid sequence selected from the group consisting of: SEQ ID NO: 7, 8, 17, 27, 28, 37, 38, 39, 40, 41, 48, 49, 50, 51, 52, 59, 60, 61, 62, 63, 64, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 89, 90, 91, 92, 93, 100, 101, 102, 103, 104, 113, 114, 121, 122, 123, 124, 1 174, 175, 176, 177, 178, 179, 180, 181, 223, 235 and 246, preferably 100 to 104.

28. The method of any one of claims 1 to 27, wherein ice nucleation is induced via an ice mist method or a reduced pressure method.

29. A lyophilized bispecific molecule composition prepared by the method of any one of claims 1-28.

Images