Controlled ice nucleation freeze-drying process for bispecific molecules

Abstract

This disclosure provides a method for preparing a lyophilized bispecific molecular composition, comprising: (a) inducing ice nucleation in a liquid protein composition having bispecific molecules at a 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 about 60 minutes to about 270 minutes; (b) exposing the vial from (a) to a second temperature of about -25°C to -50°C for about 1 hour to about 5 hours; (c) drying the composition from (b) at a third temperature of about -5°C to -25°C for about 25 hours to about 70 hours; and (d) further drying the composition from (c) at a fourth temperature of about 25°C to about 50°C for about 4 hours to about 12 hours to obtain a vial containing a lyophilized bispecific molecular composition having high molecular weight species (HMWS) in a percentage content of about 1.5% (m / V) or less. This disclosure further provides lyophilized bispecific molecular compositions prepared by the method described herein.

Description

Technical Field

[0001] The present disclosure relates to a method for lyophilizing a composition containing a bispecific molecule and the resulting lyophilized composition.

Background Art

[0002] Bispecific T cell engaging molecules such as BiTE® 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 target cells and the second binding domain is specific for CD3, a subunit of the T cell receptor complex on T cells. Due to their special design, BiTE® molecules are uniquely suited to transiently bind T cells to target cells and simultaneously potently activate the intrinsic cytolytic ability of T cells against target cells. Prior to storage, bispecific T cell engaging molecules are typically lyophilized. Lyophilization is a manufacturing process widely used to increase the stability of pharmaceutical products by removing the water content from liquid products by freeze - drying. Lyophilization is a batch process, typically time - consuming and accounting for over 50% of the drug product processing time. In addition, manufacturing - scale lyophilizers are limited in capacity. The number of vials that can be loaded in one batch is only a specific amount, and the amount of liquid composition that can be dispensed into vials is limited. Due to these constraints, 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 the lyophilization process for biopharmaceutical products such as bispecific T cell engaging molecules can be hampered in manufacturing if consistency of product attributes cannot be ensured, such as the formation of excessive high - molecular - weight species (HMWS), especially when the product concentration is high, resulting in potentially costly product losses.

Prior Art Documents

[0003] [Non-Patent Document 1] Awotwe-Otoo et al., International Journal of Pharmaceutics, 450(2013), 70-78 [Non-Patent Document 2] Esfandiary et al.,J.Pharm.Sci.,105(2016),1427-1433 [Overview of the project] [Problems that the invention aims to solve]

[0004] The freezing step in conventional freeze-drying processes involves uncontrolled or stochastic ice nucleation occurring at temperatures significantly lower than the normal freezing point. Consequently, such "supercooling" results in finer ice crystals, requiring longer drying times. One method to shorten the overall freeze-drying time is to add an annealing step, which increases the average size of ice crystals through a process called Ostwald aging. However, as we have found, such an annealing step can typically lead to aggregation, i.e., the formation of high molecular weight species (HMWS), at high concentrations of bispecific molecular formulations, e.g., at least 5 mg / ml or even at least 10 mg / ml, which is thought to be due to a cryolysis mechanism. Therefore, there is a need for an improved freeze-drying process that provides bispecific T-cell engagement molecular products with a low proportion of aggregation products, even when starting with high-concentration formulations. [Means for solving the problem]

[0005] In consideration of the above unmet needs, an object of the present invention is to provide a resource-saving method for providing bispecific molecules, preferably bispecific (T cell engagement) molecules containing an extended half-life Fc domain, at concentrations greater than 10 mg / ml, where undesirable aggregates such as HMWS are present at very low levels (e.g., less than 1.5% of the total molecule). This problem is solved by introducing a controlled ice nucleation (CIN) step into the freeze-drying process for freeze-drying high concentrations of bispecific molecules having an Fc domain, thereby reducing the overall process time compared to a standard freeze-drying process. The reduction in process time reduces energy consumption and quickly frees up the equipment for the next run. In a first aspect, a method for preparing a freeze-dried bispecific molecular composition, relating to the present invention, (a) A step of inducing ice nucleation (CIN) in a liquid bispecific molecular composition having a concentration of bispecific molecules of at least 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, which preferably includes holding after nucleation for a maximum of 90 minutes, preferably about 30 minutes. (b) A freezing step in which the vial is exposed to a second temperature of approximately -25°C to -50°C for approximately 1 to 5 hours, A drying step in which the composition of (c)(b) is dried at a third temperature of approximately -5°C to approximately -25°C for approximately 25 to 70 hours, (d) A step of further drying the composition of (c) at a fourth temperature of about 25°C to 50°C for about 4 to 12 hours to obtain a vial containing a lyophilized bispecific molecular composition having a percentage content of high molecular weight species (HMWS) of 1.5% (m / V) or less, wherein the bispecific molecule comprises at least three domains, The first domain binds to tumor antigens on target cells. The second domain binds to the extracellular epitope of the human and / or macaque CD3ε chain, and Preferably, a third domain resulting in an extended half-life of the bispecific molecule is fused to the second domain by a peptide linker, and the third domain comprises two polypeptide monomers, each containing a hinge, CH2, and CH3 domain, and the two polypeptide monomers are fused to each other via a peptide linker, and It is envisioned that a method including this will be provided.

[0006] In the above embodiments, it is also conceivable that the present invention may provide a method comprising step (a) of exposing a vial to a first temperature for about 90 minutes to about 3 hours (CIN step).

[0007] In the above embodiments, it is also conceivable that the present invention may provide a method comprising step (a) of exposing a vial to a first temperature for about 90 minutes to about 2 hours, preferably 110 minutes.

[0008] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the first temperature in step (a) is about -12°C to about -17°C.

[0009] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the first temperature in step (a) is about -15°C.

[0010] In the above embodiments, it is also conceivable that the present invention may provide a method, the method comprising step (a) further comprising holding the ice nucleation bispecific molecular composition at its temperature for a period of up to 90 minutes after nucleation.

[0011] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the period after nucleation is about 20 minutes to about 90 minutes.

[0012] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the period after nucleation is about 30 minutes.

[0013] In the above aspect, in relation to the present invention, a method is also contemplated, wherein step (b) includes exposing the vial to a second temperature of about -45°C (freezing step).

[0014] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the second period is from about 2 hours to about 4 hours.

[0015] In the above aspect, in relation to the present invention, a method as recited in claim 10 is also contemplated, wherein the second period is about 3 hours.

[0016] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the first temperature is shifted to the second temperature at a rate of about 0.01°C to about 0.5°C per minute.

[0017] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the first temperature is shifted to the second temperature at a rate of about 0.2°C per minute.

[0018] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the third temperature in step (c) is from about 0°C to about -20°C (drying step), preferably about 70 to 120 mTorr, preferably 100 mTorr.

[0019] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the third temperature in step (c) is from about -5°C to about -10°C.

[0020] In the above aspect, in relation to the present invention, a method is also contemplated, wherein the third temperature in step (c) is about -8°C.

[0021] In the above aspect, in relation to the present invention, there is also contemplated a method, wherein step (c) comprises exposing a vial containing a bispecific molecule composition to an increase in temperature at a rate of from about 0.01 °C per minute to about 0.5 °C per minute.

[0022] In the above aspect, in relation to the present invention, there is also contemplated a method, wherein the transition from step (b) to step (c) comprises increasing the temperature at a rate of from about 0.2 °C per minute to about 0.7 °C per minute and holding the vial at a temperature of from about -40 °C to about -30 °C for from about 15 minutes to about 1 hour.

[0023] In the above aspect, in relation to the present invention, there is also contemplated 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, exposed in step (b) to a second temperature of about -45 °C for about 3 hours, the composition of (b) is dried in step (c) 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, and step (d) is preferably carried out at about 70 - 120 mTorr, preferably 100 mTorr.

[0024] In the above aspect, in relation to the present invention, there is also contemplated a method, wherein the bispecific molecule is a single-chain molecule.

[0025] In the above aspect, in relation to the present invention, there is also contemplated a method, wherein the bispecific molecule is present in the composition at a concentration of from about 10 mg / ml to 30 mg / ml or from about 20 mg / ml to about 30 mg / ml or preferably from about 15 mg / ml to 25 mg / ml.

[0026] In the above embodiments, it is also conceivable that the present invention may provide a method for obtaining a vial containing a lyophilized bispecific molecular composition that exhibits aggregation of about 1.5% (m / w) or less, or preferably about 1.2, 1, 0.75, or even more 0.5% or less (with respect to total bispecific molecular weight) with respect to high molecular weight species (HMSW) formation.

[0027] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the third domain comprises hinge-CH2-CH3-linker-hinge-CH2-CH3 in the order of amino to carboxyl.

[0028] In the above embodiments, it is also conceivable that the present invention may provide a method wherein each of the polypeptide monomers of the third domain is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 249 to 256, or has an amino acid sequence that is identical to a sequence selected from the group consisting of SEQ ID NOs: 249 to 256.

[0029] In the above embodiments, it is also conceivable that the present invention may provide a method relating to the present invention, wherein the first domain is coupled to CD33, CDH19, MSLN, FLT3, BCMA, CD19, MUC17, CDH3, CLDN18.2, CD70, EGFRviii, EpCAM, DLL3 and / or PSMA, preferably DLL3.

[0030] In the above embodiments, it is also conceivable that the present invention may provide a method wherein the first binding domain of the construct comprises a VH region including CDR-H1, CDR-H2, and CDR-H3 and a VL region including CDR-L1, CDR-L2, and CDR-L3, selected from the group consisting of: (a) CDR-H1 shown in SEQ ID NO: 4, CDR-H2 shown in SEQ ID NO: 5, CDR-H3 shown in SEQ ID NO: 6, CDR-L1 shown in SEQ ID NO: 1, CDR-L2 shown in SEQ ID NO: 2, and CDR-L3 shown in SEQ ID NO: 3 (b) CDR-H1 shown in SEQ ID NO: 29, CDR-H2 shown in SEQ ID NO: 30, CDR-H3 shown in SEQ ID NO: 31, CDR-L1 shown in SEQ ID NO: 34, CDR-L2 shown in SEQ ID NO: 35, and CDR-L3 shown in SEQ ID NO: 36 (c) CDR-H1 shown in SEQ ID NO: 42, CDR-H2 shown in SEQ ID NO: 43, CDR-H3 shown in SEQ ID NO: 44, CDR-L1 shown in SEQ ID NO: 45, CDR-L2 shown in SEQ ID NO: 46, and CDR-L3 shown in SEQ ID NO: 47 (d) CDR-H1 shown in SEQ ID NO: 53, CDR-H2 shown in SEQ ID NO: 54, CDR-H3 shown in SEQ ID NO: 55, CDR-L1 shown in SEQ ID NO: 56, CDR-L2 shown in SEQ ID NO: 57, and CDR-L3 shown in SEQ ID NO: 58 (e) CDR-H1 shown in SEQ ID NO: 65, CDR-H2 shown in SEQ ID NO: 66, CDR-H3 shown in SEQ ID NO: 67, CDR-L1 shown in SEQ ID NO: 68, CDR-L2 shown in SEQ ID NO: 69, and CDR-L3 shown in SEQ ID NO: 70 (f) CDR-H1 shown in SEQ ID NO: 83, CDR-H2 shown in SEQ ID NO: 84, CDR-H3 shown in SEQ ID NO: 85, CDR-L1 shown in SEQ ID NO: 86, CDR-L2 shown in SEQ ID NO: 87, and CDR-L3 shown in SEQ ID NO: 88 (g) CDR-H1 shown in SEQ ID NO: 94, CDR-H2 shown in SEQ ID NO: 95, CDR-H3 shown in SEQ ID NO: 96, CDR-L1 shown in SEQ ID NO: 97, CDR-L2 shown in SEQ ID NO: 98, and CDR-L3 shown in SEQ ID NO: 99 (h) CDR-H1 shown in SEQ ID NO: 105, CDR-H2 shown in SEQ ID NO: 106, CDR-H3 shown in SEQ ID NO: 107, CDR-L1 shown in SEQ ID NO: 109, CDR-L2 shown in SEQ ID NO: 110, and CDR-L3 shown in SEQ ID NO: 111 (i) CDR-H1 shown in SEQ ID NO: 115, CDR-H2 shown in SEQ ID NO: 116, CDR-H3 shown in SEQ ID NO: 117, CDR-L1 shown in SEQ ID NO: 118, CDR-L2 shown in SEQ ID NO: 119, and CDR-L3 shown in SEQ ID NO: 120 (j) CDR-H1 shown in SEQ ID NO: 126, CDR-H2 shown in SEQ ID NO: 127, CDR-H3 shown in SEQ ID NO: 128, CDR-L1 shown in SEQ ID NO: 129, CDR-L2 shown in SEQ ID NO: 130, and CDR-L3 shown in SEQ ID NO: 131 (k) CDR-H1 shown in SEQ ID NO: 137, CDR-H2 shown in SEQ ID NO: 138, CDR-H3 shown in SEQ ID NO: 139, CDR-L1 shown in SEQ ID NO: 140, CDR-L2 shown in SEQ ID NO: 141, and CDR-L3 shown in SEQ ID NO: 142 (l) CDR-H1 shown in SEQ ID NO: 152, CDR-H2 shown in SEQ ID NO: 153, CDR-H3 shown in SEQ ID NO: 154, CDR-L1 shown in SEQ ID NO: 155, CDR-L2 shown in SEQ ID NO: 156, and CDR-L3 shown in SEQ ID NO: 157 (m) CDR-H1 shown in SEQ ID NO: 167, CDR-H2 shown in SEQ ID NO: 168, CDR-H3 shown in SEQ ID NO: 169, CDR-L1 shown in SEQ ID NO: 170, CDR-L2 shown in SEQ ID NO: 171, and CDR-L3 shown in SEQ ID NO: 172 (n) CDR-H1 shown in SEQ ID NO: 203, CDR-H2 shown in SEQ ID NO: 204, CDR-H3 shown in SEQ ID NO: 205, CDR-L1 shown in SEQ ID NO: 206, CDR-L2 shown in SEQ ID NO: 207, and CDR-L3 shown in SEQ ID NO: 208 (o) CDR-H1 shown in SEQ ID NO: 214, CDR-H2 shown in SEQ ID NO: 215, CDR-H3 shown in SEQ ID NO: 216, CDR-L1 shown in SEQ ID NO: 217, CDR-L2 shown in SEQ ID NO: 218, and CDR-L3 shown in SEQ ID NO: 219 (p) CDR-H1 shown in SEQ ID NO: 226, CDR-H2 shown in SEQ ID NO: 227, CDR-H3 shown in SEQ ID NO: 228, CDR-L1 shown in SEQ ID NO: 229, CDR-L2 shown in SEQ ID NO: 230, and CDR-L3 shown in SEQ ID NO: 231, (q) CDR-H1 shown in Sequence ID No. 238, CDR-H2 shown in Sequence ID No. 239, CDR-H3 shown in Sequence ID No. 240, CDR-L1 shown in Sequence ID No. 241, CDR-L2 shown in Sequence ID No. 242, and CDR-L3 shown in Sequence ID No. 243.

[0031] In the above embodiment, a method relating to the present invention, wherein the first domain is sequence numbers 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, 1 It is also conceivable to provide a method having an amino acid sequence selected from the group consisting of 31, 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, preferably 100 to 104.

[0032] In the above embodiments, it is also conceivable that the present invention may provide a method wherein ice nucleation is induced via ice fog or decompression.

[0033] In a further aspect of the present invention, a freeze-dried bispecific molecular composition is prepared by the method described herein.

[0034] While various embodiments described herein are expressed using the word “includes” in various contexts, it should be understood that relevant embodiments may also be described using “consist of” or “essentially from.” This disclosure intends that embodiments described as “including” a certain feature include embodiments that “consist of” or “substantially from” that feature. The terms “one (a)” or “one (an)” refer to one or more. Thus, the terms “one (a)” (or “one (an)”), “one or more” and “at least one” may be used interchangeably herein. The term “or” should be understood to include alternative or joint items unless the context clearly indicates otherwise.

[0035] When describing a range of values, it should be understood that this disclosure intends to refer to the individual values ​​found within that range. For example, "pH of approximately pH 4 to approximately pH 6" could be, but not limited to, pH 4.2, 4.6, 5.2, 5.5, and any value between such values. In any range described herein, the endpoints of the range are included within the range. However, this description also intends the same range in which the smaller and / or larger endpoints are excluded. Where the term "approximately" is used, it means ±5%, 10%, or more than the listed number. The intended actual variation is determined from the context.

[0036] Further features and variations of the present invention will be apparent to those skilled in the art from the whole of this application, including the drawings and detailed description, and all such features are intended as embodiments of the present invention. Similarly, the features of the present invention described herein can be rearranged to form further embodiments, which are also intended as embodiments of the present invention, regardless of whether the combination of features is specifically described as an embodiment or aspect of the present invention. The entire specification is intended to be related as an integrated disclosure, and it should be understood that all combinations of features described herein are intended, even if they are not described together in the same sentence, paragraph, or section of this specification (even if they are described in separate sections). Furthermore, only the limitations described herein as essential to the present invention should be considered such, and variations of the present invention lacking limitations not described herein as essential are intended as embodiments of the present invention. [Brief explanation of the drawing]

[0037] [Figure 1A-B] (A) A standard freeze-drying cycle for bispecific molecules without a CIN step involves annealing during the freezing stage and a longer primary drying time. (B) In the absence of an annealing step, the CIN freeze-drying cycle relies on inducing ice nucleation at warmer temperatures to form larger ice crystals, resulting in shorter primary drying and total cycle times. [Figure 2A-C]Lyophilization profiles of lyophilization cycles for both BCMAxCD3 and DLL3xCD3 bispecific molecular samples. The same lyophilization tray configuration was used for each lyophilization cycle. (A) In a standard non-CIN cycle with annealing, the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecular molecule was -33.3°C, and the maximum Tp for the DLL3xCD3 bispecific molecular molecule was -32.8°C. (B) In a standard non-CIN cycle without annealing, the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecular molecule was -31.3°C, and the maximum Tp for the DLL3xCD3 bispecific molecular molecule was -31.5°C. (C) In a CIN cycle, the maximum product temperature (Tp) for the BCMAxCD3 bispecific molecular molecule was -27.8°C, and the maximum Tp for the DLL3xCD3 bispecific molecular molecule was -28.7°C. [Figure 3A-B] 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 retention. Based on this, -15°C and 0.5 hours post-nucleation were selected as the nucleation parameters for CIN cycle 2, and the experiment proceeded. (A) BCMAxCD3 bispecific molecule, (B) DLL3xCD3 bispecific molecule. [Figure 4A-B] The %HMWS time-course stability results for the BCMAxCD3 bispecific molecule and the DLL3xCD3 bispecific molecule for the standard non-CIN cycle, the standard non-CIN cycle without annealing, and the CIN cycle are the %HMWS recorded immediately after the investigation at each time point. The lines overlap in the graph. (A) BCMAxCD3 bispecific molecule, (B) DLL3xCD3 bispecific molecule. [Figure 5] Zero-time water content of BCMAxCD3 bispecific molecule and DLL3xCD3 bispecific molecule samples after freeze-drying. There were no significant differences in water content between the groups (p>0.05). [Figure 6A-F]CEX-HPLC results of BCMAxCD3 bispecific molecules 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. [Figure 7] Average cake resistance at a depth of 1 mm in a 6R vial with a filling volume of 1.3 ml. Cake resistance is reduced by >50% when using CIN compared to a standard lyophilization process, and is also significantly reduced compared to a standard process with additional annealing. In the "CIN1" process using CIN process parameters not according to the present invention, the nucleation temperature was -7°C, while in contrast, the "CIN2" process according to the present invention had a nucleation temperature of -15°C. [Modes for carrying out the invention]

[0038] This disclosure provides a method for preparing lyophilized bispecific molecular compositions, particularly compositions comprising bispecific T cell engagement molecules containing an extended half-life Fc domain, the method comprising a controlled ice nucleation (CIN) step. Surprisingly, CIN has been found to be a superior alternative to introducing an annealing step into the lyophilization of bispecific molecules according to the present invention, compared to conventional processes without an annealing step or a CIN step, in order to achieve both desirable product quality (significantly reduced HMWS compared to lyophilization with a conventional annealing step) and reduced process time. The reduction in process time results from the shortening of steps (c) and (d) described herein compared to conventional lyophilization processes without annealing. For example, while a conventional lyophilization process may take up to 100 hours for the primary drying step (c), the process according to the present invention has a shorter drying time, preferably only about 50 hours for step (c). Considering the secondary drying step (d), the reduction in overall process time compared to conventional lyophilization processes is at least 30%, or even more, for example, 40% or more. Consequently, the consumption of resources such as energy is also reduced. For bispecific molecules, exemplary conventional or, as referred herein, “standard” freeze-drying processes, i.e., freeze-drying processes generally without a CIN step, are understood herein to be those without a CIN or annealing step, e.g., a freezing step of about 2 hours at about -45°C (b), a primary drying step of about 100 hours at about -25°C and about 70 mTorr (c), and a secondary drying step of about 8 hours at about 40°C and about 70 mTorr (d).

[0039] While not bound by theory, introducing a CIN step using the process parameters disclosed herein results in larger ice crystals and therefore shorter cycle times. CIN introduces externally formed ice crystals into the drug product vial, thus increasing the size of the ice crystals after freezing. The present invention demonstrates that, in contrast to an annealing step operated under selected conditions tailored to bispecific molecules, for example, aggregation is efficiently mitigated in terms of HMSW without affecting other product quality attributes such as the appearance and moisture content of the freeze-dried cake. Furthermore, cake resistance was advantageously reduced by the method of the present invention, including the CIN step, by at least 30%, typically about 50%, compared to standard non-CIN freeze-drying with or without the annealing step. Freeze-drying cycles with annealing typically cause undesirable aggregation, typically at least 3% HMSW, while freeze-drying processes including the CIN step described herein maintain aggregation, i.e., HMSW levels below 1.5%, which is close to 1% of the pre-freeze-drying value or even lower, as further demonstrated in the examples.

[0040] The CIN step according to the present invention involves incubating the composition for about 1 to 5 hours, typically 90 to 120 minutes, e.g., 90, 110, or 120 minutes, preferably in the range of -18 to -10°C, to cool it so that it does not melt when ice crystals are introduced in the next step (pre-nucleation). After ice crystals are introduced, the composition is incubated for a further period of time, preferably in the range of -18 to -10°C, to allow the ice crystals to grow (post-nucleation). An advantage of the present invention is to provide a method that includes a CIN step based on “sweet spot” conditions, which promotes both the quality of beneficial bispecific molecular products (preferably without HMWS with respect to product uniformity and at higher product concentrations) and makes the process more efficient in terms of shortening the length of the drying process, as described and demonstrated herein. Higher ice nucleation temperatures above -10°C, such as -7°C, result in bispecific molecular products that are typically of lower quality in terms of aggregation, despite the advantage of faster drying times. In contrast, lower ice nucleation temperatures below -18°C typically result in longer drying times, potentially negating the benefits in terms of process length, resources, and energy savings. Some process parameters may have less impact on product quality and process economics than the temperature and duration of the CIN process in terms of drying time. For example, it is known in the art that secondary drying pressure has little effect on the drying process and therefore may not be particularly helpful in saving energy, and thus cannot be set over a wide range of temperatures.

[0041] The method preferably comprises inducing ice nucleation in a liquid bispecific molecular composition in a vial exposed to a first temperature in the range of about -18°C to about -10°C (e.g., about -18, -17, -16, -15, -14, -13, -12, -11, or -10°C) for about 1 to about 5 hours, preferably 90 to 120 minutes; exposing the vial to a second temperature of about -25°C to -50°C for a second period; and drying the composition at a third temperature of about -5°C to about -25°C. The method yields a vial containing a lyophilized bispecific molecular composition with favorable product parameters (low HMWS despite higher commercially beneficial product concentrations) and having an aspect ratio of about 0.75 or greater (e.g., 0.8, 0.9, or 0.95 or greater). In various embodiments, the method yields vials containing lyophilized bispecific molecular compositions having an aspect ratio of approximately 1 or greater (e.g., 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, or 2 or greater). Notably, the processes disclosed herein allow for remarkably high fill volumes (i.e., aspect ratios) for a given vial while minimizing the risk of vial breakage and maintaining reasonable drying times. The "aspect ratio" is the fill height of the vial (the height of the bispecific molecular composition in the vial) divided by the inner diameter of the vial (aspect ratio = fill height / inner diameter of vial). Previous methods required reducing the amount introduced into the vial (lowering the aspect ratio) to prevent vial breakage and collapse of the product cake during the lyophilization process. The methods described herein allow for higher aspect ratios during the lyophilization process and offer numerous potential advantages. For example, for bispecific molecular therapeutics administered at high doses, fewer vials are required for dosing. Alternatively (or in addition), this process allows for the use of smaller vials compared to other freeze-drying methods while maintaining the same fill volume, thereby resulting in more units per batch and thus increased yield. The use of smaller vials also offers the benefit of reducing storage space in manufacturing plants and clinics.These are just a few examples of the advantages of the instant method, achieved while generating a product cake suitable for use in biopharmaceuticals.

[0042] This disclosure describes various conditions for use in a freeze-drying process to produce freeze-dried bispecific molecular compositions. Generally, vials suitable for pharmaceutical compositions (e.g., glass vials) are filled with the liquid bispecific molecular composition and exposed to different temperatures and pressures to achieve a freeze-dried product. Vials can be of any size or shape suitable for use in a freeze-drying process and can be formed from various materials such as glass, metal, or plastic (e.g., polycarbonate, polystyrene, polypropylene, or polyolefin). For example, vials can be glass, glass-like, and tubular in shape. Molded glass vials are commercially available in various different 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). Vials may be configured to include suitable stoppers, such as commercially available elastomer stoppers, which are available from, for example, Daikyo Seiko, Ltd. or West Pharmaceutical Services, Inc. The processes described herein are carried out in many embodiments in a freeze-drying chamber or an ice nucleation system. The freeze-drying chamber may be operated according to the instructions of a manufacturer suitable for the pharmaceutical composition. In various embodiments of this disclosure, the freeze-drying method does not include an annealing step. "Annealing" refers to the process of circulating the temperature of the formulation (e.g., from low to high and then back to low). Various embodiments of the disclosed methods enable the production of a freeze-dried product without such an annealing step.

[0043] The method of this disclosure comprises inducing ice nucleation in a liquid bispecific molecular composition in a vial (referred to herein as the “ice nucleation step”). Ice nucleation may be initiated using any of a number of methods, including but not limited to ice fog, emergency decompression / rapid decompression, and vacuum-induced evaporative cooling. Other methods for controlling ice nucleation include, for example, ultrasound, gap freezing, electrofreezing, temperature quenching, use of pre-cooling shelves, and mechanical stirring.

[0044] In various aspects of this disclosure, ice nucleation is induced in a liquid bispecific molecular composition by ice fog. Ice fog involves "seeding" externally produced ice crystals into a supercooled solution in a vial. An ice fog generator is used to produce a suspension of fine ice crystals, which is then injected into a freeze-drying chamber. The ice fog crystals act as ice seeds for the supercooled liquid product in the vial. When the ice crystals from the fog enter the partially capped vial and come into contact with the surface of the supercooled liquid, ice nucleation occurs instantaneously in the vial at a specified shelf temperature. This occurs simultaneously in all vials, improving uniformity within the batch. Ice fog systems are available from IMA Life (Tonawanda, NY) and Millrock Technologies (Kingston, NY). Ice fog technology is further described, for example, in Azzarella et al., BioPharm.Int., 29(12)(2017), 36-41.

[0045] In various aspects of this disclosure, ice nucleation is induced in liquid bispecific molecular compositions by reduced pressure. Rapid depressurization generally involves initially pressurizing a freeze-drying chamber to 1.5–2 atmospheres (approximately 20–30 psig) using an inert gas such as nitrogen, and then rapidly releasing the pressure (e.g., less than 3 seconds) to a pressure slightly above the ambient pressure. The rapid change in pressure induces nucleation in the vial. Rapid depressurization systems are available from SP Scientific (Gardiner, NY) and are further described, for example, in Luoma et al., “Controlled Ice Nucleation Using ControLyo® Pressurization-Depressurization 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.

[0046] Vacuum induction evaporative cooling generally involves reducing the pressure in the freeze-drying chamber to slightly above the boiling point of the solution, thereby enhancing the evaporative cooling effect on the liquid surface and enabling nucleation. Vacuum induction evaporative cooling systems are available from HOF Sonderanlagenbau GmbH (Lohra, Germany).

[0047] Ice nucleation is induced in a liquid bispecific molecular composition related to 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, etc.) for a period of about 60 minutes to about 5 hours (e.g., by ice fog). Optionally, the ice nucleation step includes exposing the vial to the first temperature for about 30 minutes to about 2 hours, for example, about 90 minutes to about 2 hours. In various embodiments, the vial is exposed to a first temperature of about -15°C. Higher temperatures, such as -7°C, result in significantly higher percentage levels of HMWS, which is generally undesirable in the production of therapeutic bispecific molecules related to the present invention. Therefore, careful selection of a specific temperature range for the CIN step is not arbitrary. Optionally, this step of the method further comprises holding the ice nucleation protein composition at that temperature (optionally the same temperature) for a post-nucleation period of up to 2 hours. For example, the post-nucleation holding time may be about 30 minutes to about 90 minutes (e.g., about 45 minutes to about 75 minutes, e.g., 50 minutes, 55 minutes, 60 minutes, 65 minutes, 70 minutes, or 75 minutes). In various embodiments, the post-nucleation holding time is about 60 minutes.

[0048] The method further includes exposing the vial to a second temperature of about -25°C to about -50°C for a second period of time (referred to herein as the “freezing step”). The second temperature may 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 embodiments, the freezing step includes exposing the vial to a second temperature of about -45°C. The second period is optionally about 1 hour to about 5 hours, for example, about 2 hours to about 4 hours. The second period may be approximately 60 minutes, 90 minutes, 120 minutes, 150 minutes, 180 minutes, 210 minutes, or 240 minutes (or any range including these endpoints). In various embodiments, the second period is approximately 3 hours.

[0049] In various embodiments of this disclosure, the cooling rate of the vial can be controlled. For example, the first temperature is optionally transitioned to a 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 embodiments, the first temperature is transitioned to a second temperature at a rate of about 0.2°C per minute.

[0050] The method further comprises a step of drying the composition obtained from the freezing step at a third temperature of about 0°C to about 40°C (referred to herein as the “drying step”), resulting in a vial containing a freeze-dried bispecific composition having an aspect ratio of about 0.75 or greater. In various embodiments, 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. Optionally, the third temperature is about 0°C to about 25°C, for example, about 10°C to about 25°C (e.g., about 25°C).

[0051] Optionally, the transition from the freezing process to the drying process includes increasing the temperature at a rate of approximately 0.2°C to 0.7°C per minute and maintaining the vial at a temperature of approximately -40°C to -30°C for approximately 15 minutes to approximately 1 hour. For example, this transition may include increasing the temperature at a rate of approximately 0.2°C, 0.3°C, 0.4°C, 0.5°C, 0.6°C, or 0.7°C. This transition may occur over a period of approximately 15 minutes, 30 minutes, 45 minutes, or 60 minutes.

[0052] In various aspects of the present disclosure, the drying step includes exposing a vial containing the bispecific molecular composition to a temperature increase 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 present disclosure, the drying step includes (1) holding the vial at a temperature of about -5°C to about 5°C for about 8 hours to about 12 hours, and (2) holding the vial at a temperature of about 20°C to about 30°C for about 20 hours to about 50 hours. In exemplary embodiments of the present disclosure, drying step 1 includes holding the vial at a temperature in any range having endpoints 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 endpoints of these. In various embodiments, the duration of drying step 1 is about 8 hours to about 10 hours, about 9 hours to about 11 hours, or 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 endpoints of these. In exemplary embodiments of the present disclosure, drying step 2 includes holding the vial at a temperature of about 25°C to about 30°C, about 20°C to about 25°C, or 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 duration of drying step 2 is optionally about 20 to 45 hours, about 20 to 40 hours, about 20 to 35 hours, about 20 to 30 hours, about 25 to 50 hours, about 30 to 50 hours, about 35 to 50 hours, about 40 to 50 hours, or about 45 to 50 hours (for example, 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 holding the vial at a temperature of about 0°C for about 10 hours. Drying step (2) optionally includes holding the vial at a temperature of about 25°C for about 40 hours.

[0053] The method of this disclosure allows for the use of a larger volume of the liquid composition in the vial before freeze-drying. In this regard, the liquid bispecific molecular composition may fill at least 50% of the vial volume before carrying out the ice nucleation step. The method may include filling at least 50% of the vial volume with the liquid bispecific molecular composition before the ice nucleation step. The liquid bispecific molecular composition may fill at least 55%, 60%, 65%, or 75% of the vial volume.

[0054] This disclosure further provides lyophilized bispecific molecular compositions prepared by the methods described herein. The lyophilized bispecific molecular formulations optionally further comprise sugars, surfactants and / or buffers. The formulations optionally also have 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 examples, the pH is about 4 to about 6. In some preferred examples, the pH of the formulation is about 4 or about 4.2. In various examples, the pH of the formulation is about 5. In some embodiments, the pH of the formulation is about 6.

[0055] In some embodiments, the bispecific molecule of the lyophilized formulation is an antigen-binding protein. An "antigen-binding protein" is a protein containing a domain that binds to a designated target antigen (such as HER2 or CD3 and / or DLL3, BCMA or CD33). The antigen-binding protein includes a backbone or framework portion that allows the antigen-binding domain to adopt a three-dimensional structure that facilitates the binding of the antigen-binding protein to the antigen.

[0056] The term "half-life" as used herein refers to the period during which the concentration of a biological substance (such as the bispecific molecule of the present invention) reaches the maximum concentration in plasma (C). max ) from C max It is understood as the time it takes for the amount to decrease to half. The extended half-life is preferably at least 40 hours, more preferably 50, 60, 70, 80, 90, or 100 hours, compared to non-half-life extended bispecific molecules that typically exhibit a half-life of 24 hours or less.

[0057] In this specification, the term "polypeptide" is understood to mean an organic polymer comprising at least one continuous, unbranched amino acid chain. In connection with the present invention, polypeptides comprising two or more amino acid chains are also envisioned. The amino acid chains of a polypeptide typically comprise at least 50 amino acids, preferably at least 100, 200, 300, 400, or 500 amino acids. In connection with the present invention, it is also envisioned that the amino acid chains of a polymer are linked to entities not composed of amino acids.

[0058] The term “antigen-binding polypeptide” in the present invention is preferably a polypeptide that binds immunospecifically to its target or antigen. Antigen-binding polypeptides typically include domains comprising or derived from the heavy chain variable region (VH) and / or light chain variable region (VL) of an antibody. Polypeptides according to the present invention include the minimum structural requirements of an antibody that enable immunospecific target binding. These minimum requirements can be defined, for example, 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), and preferably by the presence of all six CDRs. The antigen-binding molecule of the present invention is therefore preferably a T-cell engagement polypeptide characterized by the presence of three or six CDRs in one or both binding domains, and those skilled in the art will know the location (and order) of these CDRs within the binding domain. Preferably, “antigen-binding molecule” is understood as “antigen-binding polypeptide” in the context of the present invention. In alternative embodiments, the antigen-binding polypeptide of the present invention may be an aptamer.

[0059] Instead, in relation to the present invention, a molecule is an antigen-binding polypeptide corresponding to an “antibody construct,” and typically refers to a molecule whose structure and / or function are based on the structure / function of an antibody (e.g., a full-length immunoglobulin molecule or a whole immunoglobulin molecule). Thus, an antigen-binding molecule can bind to its specific target or antigen and / or originates from the variable heavy chain (VH) and / or variable light chain (VL) domains of an antibody or a fragment thereof. Furthermore, a domain that binds to a binding partner according to the present invention is understood herein as the binding domain of the antigen-binding molecule according to the present invention. Typically, the binding domain according to the present invention includes the minimum structural requirements of an antibody that enable target binding. These minimum requirements can be defined, for example, 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. Alternative methods for defining the minimum structural requirements of an antibody include defining the protein domain (epitope cluster) of the target protein that contains the antibody's epitope or epitope region within the structure of a particular target, or by referencing a specific antibody that competes with the defined antibody's epitope. Examples of antibodies that form the basis of the construct according to the present invention include monoclonal antibodies, recombinant antibodies, chimeric antibodies, deimmunized antibodies, humanized antibodies, and human antibodies.

[0060] In relation to the present invention, the polypeptides of the present invention bind to their respective target structures in a specific manner. Preferably, the polypeptides according to the present invention contain one paratope for each binding domain that "specifically or immunospecifically binds to," "recognizes (specifically or immunospecifically)," or "reacts (specifically or immunospecifically) with" each target structure. This means that, according to the present invention, the polypeptide or its binding domain interacts with, or (immunologically)-specifically interacts with, a given target molecule (antigen) and a given epitope on CD3, respectively. This interaction or association occurs more frequently, more rapidly, more persistently, more favorably, or in some combination of these parameters, with respect to the epitope on a particular target compared to an alternative substance (non-target molecule). However, due to sequence similarities between homologous proteins in different species, a binding domain that (immunologically)-specifically binds to its target (e.g., a human target) may cross-react with homologous target molecules from different species (e.g., non-human primates). Therefore, the term “specific / immunospecific binding” may include binding of a binding domain to epitopes of two or more species and / or structurally related epitopes. The term “(immuno)selective binding” excludes binding to structurally related epitopes.

[0061] The binding domain of the antigen-binding molecule according to the present invention may include, for example, the group of CDRs referenced above. Preferably, these CDRs are contained within the framework of an antibody light chain variable region (VL) and an antibody heavy chain variable region (VH), but do not necessarily have to include both. An Fd fragment, for example, has two VH regions and often retains the antigen-binding function of a portion of the intact antigen-binding domain. Further examples of antibody fragment, antibody variant, or binding domain formats include: (1) Fab fragment, a monovalent fragment having VL, VH, CL, and CH1 domains; (2) F(ab')2 fragment, a bivalent fragment having two Fab fragments linked by disulfide crosslinks at the hinge region; (3) Fd fragment having two VH domains and a CH1 domain; (4) Fv fragment having a single arm of the antibody with VL and VH domains; (5) dAb fragment having a VH domain (Ward et al., (1989) Nature 341:544-546); (6) isolated complementarity-determining regions (CDRs); and (7) single-stranded Fv (scFv), the latter of which is preferred (e.g., derived from scFV libraries). Examples of embodiments of the antigen-binding molecule according to the present invention are described, for example, in International Publication Nos. 00 / 006605, 2005 / 040220, 2008 / 119567, 2010 / 037838, 2013 / 026837, 2013 / 026833, U.S. Patent Application Publication Nos. 2014 / 0308285, 2014 / 0302037, International Publication Nos. 2014 / 144722, 2014 / 151910, and 2015 / 048272.

[0062] The definition of “binding domain” or “domain that binds” also includes fragments of full-length antibodies, such as VH, VHH, VL, (s)dAb, Fv, Fd, Fab, Fab', F(ab')2, or “r IgG” (“half-antibody”). The antigen-binding molecules according to the present invention may also include modified fragments of antibodies, also called antibody variants, such as scFv, di-scFv or bi(s)-scFv, scFv-Fc, scFv-zipper, scFab, Fab2, Fab3, diabody, single-stranded diabody, tandem diabody (Tandab's), tandem di-scFv, tandem tri-scFv, “multibody,” such as triabody or tetrabody, and single-domain antibodies, such as nanobody, or single variable-domain antibodies containing a single variable domain that may be VHH, VH, or VL that specifically binds to an antigen or epitope independent of other V regions or domains. Typically, the binding domain of the present invention includes a paratope that facilitates binding to its binding partner.

[0063] As used herein, the terms “single-chain Fv,” “single-chain antibody,” or “scFv” refer to an antibody fragment of a single polypeptide chain that contains variable regions derived from both the heavy and light chains but lacks a constant region. Typically, single-chain antibodies further include a polypeptide linker between the VH and VL domains, which enables the formation of a desired structure that allows binding to an antigen. Single-chain antibodies are discussed in detail by Pluckthun in 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 Publication No. 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; and Skerra et al. (1988) Science 242:1038-1041. In certain embodiments, single-chain antibodies may be bispecific, multispecific, human and / or humanized and / or synthetic.

[0064] In relation to the present invention, a paratope is understood to be a part of a polypeptide, as described herein, that recognizes and binds to an antigen as an antigen-binding site. A paratope is typically a small region of at least about five amino acids. Paratopes as understood herein typically include a portion of the heavy chain (VH) and light chain (VL) sequences derived from an antibody. Each binding domain of the molecule according to the present invention comprises a paratope containing a set of six complementarity-determining regions (CDR loops), three of which are contained within the antibody-derived VH and VL sequences.

[0065] Furthermore, the definition of the term “antigen-binding molecule” preferably includes a polyvalent / multivalent construct, and therefore a bispecific molecule, where bispecificity means that it specifically binds to two cell types containing different antigenic structures (i.e., target cells and effector cells). The antigen-binding molecules of the present invention are preferably multitarget, and therefore they are usually polyvalent / multivalent molecules, that is, they specifically bind to three or more antigenic structures, preferably four different binding domains related to the present invention, which are two target-binding domains and two CD3-binding domains. The term “multitarget bispecific antigen-binding molecule” includes the terms “multitarget bispecific T cell engager molecule” and “multitarget bispecific T cell engager polypeptide (MBiTEP)”. Preferred “multitarget bispecific antigen-binding molecules” are “multitarget bispecific T cell engager molecules” or “multitarget bispecific T cell engager polypeptide (MBiTEP)”. The term “multitarget bispecific T cell engager molecule” is understood to include the term “multitarget bispecific T cell engager polypeptide.” Furthermore, the definition of the term “antigen-binding molecule” includes molecules containing only one peptide chain and molecules consisting of two or more polypeptide chains (these chains may be identical (homodimer, homotrimer, or homooligomer) or different (heterodimer, heterotrimer, or heterooligomer)). Such molecules containing two or more polypeptide chains (i.e., typically two chains) have these chains typically linked to each other as heterodimers by charge pair bonds within a hetero-Fc entity that functions as a spacer and half-life extension portion between two bispecific entities, as described herein.Examples of antigen-binding molecules (e.g., antibody-based molecules) and their variants or derivatives identified above are described, in particular, in Harlow and Lane, Antibodies: A Laboratory Manual, CSHL Press (1988) and Using Antibodies: A Laboratory Manual, CSHL Press (1999), Kontermann and Duebel, Antibody Engineering, Springer, 2nd ed. 2010, and Little, Recombinant Antibodies for Immunotherapy, Cambridge University Press 2009.

[0066] The term “bispecificity,” as used herein, refers to an antigen-binding molecule that is “at least bispecific,” i.e., addresses two different cell types, i.e., target cells and effector cells, and comprises at least a first and third binding domain and a second and fourth binding domain, wherein at least two binding domains preferably bind to two antigens or targets selected from CD20, CD22, FLT3, MSLN, CDH3, CLL1, and EpCAM, and the other two binding domains of the same molecule bind to another antigen (CD3 in this specification) on effector cells, typically T cells. Thus, the antigen-binding molecule according to the present invention comprises specificity to at least two different antigens or targets. For example, it is preferable that the two domains do not bind to one or more extracellular epitopes of CD3e species as described herein.

[0067] The term “target cell surface antigen” refers to an antigenic structure expressed by a cell and present on the cell surface so that antigen-binding molecules, as described herein, can reach it. In relation to the present invention, a preferred target cell surface antigen is a tumor-associated antigen (TAA). A TAA may 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). The TAA is preferably a tumor antigen. The term “bispecific antigen-binding molecule” in the present invention also includes bispecific multitarget antigen-binding molecules, such as a tritarget antigen-binding molecule (the latter containing three binding domains) or a construct having four or more (e.g., four, five, etc.) specificities.

[0068] Molecules that are “multitarget” and “typically target at least two targets per molecule of the present invention (e.g., TAAs) for each target cell” are preferred in relation to the present invention. In this regard, multitarget molecules such as antigen-binding molecules are specific to two (typically identical) effector structures on effector cells, such as CD3, more preferably CD3 epsilon (CD3e included whenever “CD3” is referenced in the present invention), and at least two target cell surface antigens. The specificity is conferred by each binding domain, as defined herein. Typically, “multitarget” refers to a molecule that is specific to at least two (preferably different) target cell surface antigens (e.g., TAAs), thereby conferring preferred properties of multitarget antigen-binding molecules according to the present invention, in other words, reduced antigen deficiency and increased selectivity, i.e., the selectivity to kill target cells that co-express the target because the molecule of the present invention has a binding domain and the target cells are disease-related. Thus, the therapeutic range of the molecule of the present invention is increased compared to single-target bispecific molecules, resulting in higher drug tolerability, typically as demonstrated herein.

[0069] The T cell engagement antigen-binding molecule according to the present invention, for example, a single-chain polypeptide, is preferably bispecific, which is understood herein to typically include one domain that binds to at least one target antigen and another domain that binds to CD3. Therefore, it does not exist in nature, and its function differs significantly from that of naturally occurring products. Accordingly, the polypeptide according to the present invention is an artificial "hybrid" polypeptide comprising at least two distinct binding domains with different specificities, and is therefore bispecific. Bispecific antigen-binding molecules can be produced by various methods, including hybridoma fusion or Fab' fragment linking. See, for example, Songsivilai & Lachmann, Clin. Exp. Immunol. 79:315-321 (1990).

[0070] The antigen-binding molecule of the present invention typically comprises four binding domains and variable domains (VH / VL) that include a peptide linker (spacer peptide). The term "peptide linker," according to the present invention, includes an amino acid sequence that links the amino acid sequences of one (variable and / or binding) domain of the antigen-binding molecule of the present invention to each other. The peptide linkers between the first and second domains and the third and fourth domains (the first and third domains can preferably bind simultaneously to two targets, preferably different targets (e.g., TAA1 and TAA2) on the same cell) are preferably mobile and have a limited length (e.g., 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, or 18 amino acids). The peptide linker can also be used to fuse spacers to other domains of the antigen-binding molecule of the present invention. An essential technical feature of such peptide linkers is that they do not contain polymerization activity. Suitable peptide linkers are those described in U.S. Patent Nos. 4,751,180 and 4,935,233 or International Publication No. 88 / 09344. Peptide linkers can also be used to attach other domains, modules, or regions (such as half-life extension domains) to the antigen-binding molecule of the present invention. However, typically, the linker between the first and second target-binding domains differs from the in-conjugate linker that links the VH and VL within the target-binding domains. This difference is that the linker between the first and second binding domains has one more amino acid than the in-conjugate linker (e.g., 6 and 5 amino acids, such as GGGGS compared to SGGGGS). Surprisingly, this simultaneously confers mobility and stability to certain antigen-binding molecule forms, as described herein.Since a spacer also functions as a linker, linking two bispecific entities and contributing to the construction of at least one continuous polypeptide chain containing preferably four binding domains or portions thereof, a spacer (or spacer entity as a synonym) between two bispecific entities as described herein is a particular embodiment of a linker. However, in addition, a spacer functions as an entity that sterically separates two bispecific entities. Therefore, in relation to the present invention, a spacer is a particular embodiment of a linker, thereby contributing to linking two binding domains (of two different bispecific entities) together with two even shorter, mobile linkers at both ends, but above all, it separates the bispecific entities so that they can act advantageously as described herein, for example, to exhibit a remarkably high selectivity gap.

[0071] The antigen-binding molecule of the present invention is preferably an "in vitro-generated antigen-binding molecule." This term refers to an antigen-binding molecule as defined above, in which all or part of the variable region (e.g., at least one CDR) is generated by any other method that allows for the selection of non-immune cells, e.g., in vitro phage display, protein chip, or any other method that allows for the testing of candidate sequences with respect to antigen-binding ability. Accordingly, this term preferably excludes sequences generated solely by genomic rearrangement in animal immune cells. A "recombinant antibody" is an antibody produced by the use of recombinant DNA technology or genetic engineering.

[0072] The term “monoclonal antibody” (mAb), or monoclonal antibodies derived from antigen-binding molecules as used herein, refers to antibodies obtained from a substantially homogeneous population of antibodies, i.e., individual antibodies that are identical except for possible naturally occurring mutations and / or post-translational modifications (e.g., isomerization, amidation) that may be present in small amounts. Monoclonal antibodies are highly specific and induced to a single antigenic site or determinant on an antigen, in contrast to conventional (polyclonal) antibody preparations, which typically contain different antibodies induced to different determinants (or epitopes). In addition to their specificity, monoclonal antibodies have the advantage of being synthesized by hybridoma culture and therefore not being contaminated by other immunoglobulins. The modifier “monoclonal” indicates the characteristics of an antibody as obtained from a substantially homogeneous population of antibodies and should not be interpreted as requiring the antibody to be produced by any particular method.

[0073] Any technique that yields antibodies produced by continuous cell line culture can be used to prepare monoclonal antibodies. For example, the monoclonal antibodies used may be produced by the hybridoma method first described by Koehler et al., Nature, 256:495 (1975), or by the recombinant DNA method (see, for example, U.S. Patent No. 4,816,567). Further examples of techniques for producing human monoclonal antibodies include the trioma technique, the human B-cell hybridoma technique (Kozbor, Immunology Today 4 (1983), 72), and the EBV-hybridoma technique (Cole et al., Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc. (1985), 77-96).

[0074] Next, hybridomas can be screened using standard methods such as enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance analysis, e.g., Biacore®, to identify one or more hybridomas that produce antibodies that specifically bind to a particular antigen. For example, any form of the relevant antigen, such as recombinant antigen, naturally occurring form, any variant or fragment thereof, and its antigenic peptide, can be used as an immunogen. Using surface plasmon resonance employed in the Biacore system, the efficiency of phage antibody binding to epitopes of surface antigens on target cells can be increased (Schier, Human Antibodies Hybridomas 7(1996), 97-105; Malmborg, J.Immunol.Methods 183(1995), 7-13).

[0075] Another exemplary method for producing monoclonal antibodies involves 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).

[0076] In addition to using display libraries, relevant antigens can be used to immunize non-human animals, such as rodents (e.g., mice, hamsters, rabbits, or rats). In one embodiment, the non-human animal contains at least a portion of the human immunoglobulin gene. For example, a mouse strain that is defective in producing mouse antibodies can be modified using a large fragment of the human Ig (immunoglobulin) locus. Using hybridoma technology, antigen-specific monoclonal antibodies derived from genes with desired specificity can be produced and selected. See, for example, XENOMOUSE®, Green et al. (1994) Nature Genetics 7:13-21, U.S. Patent Application Publication No. 2003-0070185, International Publication Brochures No. 96 / 34096 and 96 / 33735.

[0077] Monoclonal antibodies can be obtained from non-human animals and then modified using recombinant DNA techniques known in the art, such as humanization, deimmunization, and chimerization. Examples of modified antigen-binding molecules include humanized variants of non-human antibodies, "affinity-mature" 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 variants with modified effector function (see, e.g., U.S. Patent No. 5,648,260, Kontermann and Dubel (2010), loc.cit. and Little (2009), loc.cit.).

[0078] In immunology, affinity maturation is the process by which B cells produce antibodies with increased affinity for an antigen during an immune response. Repeated exposure to the same antigen causes the host to continuously produce antibodies with higher affinity. Similar to natural prototypes, in vitro affinity maturation is based on the principles of mutation and selection. In vitro affinity maturation is widely used to optimize antibodies, antigen-binding molecules, and antibody fragments. Random mutations within the CDR are introduced using radiation, chemical mutagens, or error-prone PCR. In addition, chain shuffling can increase genetic diversity. Two or three rounds of mutation and selection using display methods such as phage display typically yield antibody fragments with affinities in the low nanomolar range.

[0079] A preferred type of amino acid substitution variant of an antigen-binding molecule involves substitution of one or more hypervariable region residues of the parent antibody (e.g., a humanized antibody or a human antibody). Generally, the resulting variants, selected for further development, have improved biological properties compared to the parent antibody from which they were generated. A convenient method for generating such substitution variants involves affinity maturation using phage display. Briefly, several hypervariable region sites (e.g., 6-7 sites) are mutated to generate all possible amino acid substitutions at each site. The antibody variants thus generated are then presented in a monovalent form from filamentous phage particles as fusions with the M13 gene III product packaged within each particle. The phage-displayed variants are then screened for their biological activity (e.g., binding affinity) as disclosed herein. To identify candidate hypervariable region sites to modify, alanine scanning mutagenesis can be performed to identify hypervariable region residues that significantly contribute to antigen binding. Alternatively, or in addition, it may be beneficial to analyze the crystalline 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 by the techniques detailed herein. Once such variants are generated, a screening process as described herein may be performed on a panel of variants, and antibodies exhibiting superior properties in one or more relevant assays may be selected for further development.

[0080] The monoclonal antibodies and antigen-binding molecules of the present invention particularly include “chimeric” antibodies (immunoglobulins) in which a portion of the heavy chain and / or light chain is identical or homologous to a corresponding sequence of an antibody derived from a particular species or belonging to a particular antibody class or subclass, and the remainder of the chain is identical or homologous to a corresponding sequence of an antibody derived from another species or belonging to another antibody class or subclass, and to the extent that they exhibit the desired biological activity (U.S. Patent No. 4,816,567; Morrison et al., Proc. Natl. Acad. Sci. USA, 81:6851-6855 (1984)). The chimeric antibodies of interest herein include “primatized” antibodies that include a variable domain antigen-binding sequence derived from a non-human primate (e.g., Old World monkeys, apes, etc.) and a human constant region sequence. Various methods for producing chimeric antibodies are described. For example, see Morrison et al., Proc. Natl. Acad. ScL USA 81:6851, 1985; Takeda et al., Nature 314:452, 1985; U.S. Patent No. 4,816,567 by Cabilly et al.; U.S. Patent No. 4,816,397 by Boss et al.; European Patent No. 0171496 and No. 0173494 by Tanaguchi et al.; and British Patent No. 2177096.

[0081] Antibodies, antigen-binding molecules, antibody fragments, or antibody variants can also be modified by specific deletion of human T cell epitopes (a method called "deimmunization"), for example, by the methods disclosed in International Publication No. 98 / 52976 or International Publication No. 00 / 34317. Briefly, the heavy and light chain variable domains of an antibody can be analyzed to determine whether they are peptides that bind to MHC class II, and these peptides indicate that they are potential T cell epitopes (as defined in International Publication No. 98 / 52976 and International Publication No. 00 / 34317). To detect potential T cell epitopes, a computer modeling technique called "peptide threading" can be applied, as described in International Publication No. 98 / 52976 and International Publication No. 00 / 34317, and in addition, a database of human MHC class II-binding peptides can be searched for motifs present in the VH and VL sequences. These motifs bind to any of the 18 major MHC class II DR allotypes and therefore become potential T cell epitopes. Detected potential T cell epitopes can be eliminated by substituting a few amino acid residues in the variable domain, or preferably by substituting a single amino acid. Typically, conservative substitutions are performed. In many, though not all, amino acids common to the 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 Vol.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 (compiled 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.For example, a consensus human framework area, such as that described in U.S. Patent No. 6,300,064, can also be used.

[0082] A “humanized” antibody, antigen-binding molecule, variant, or fragment thereof (e.g., Fv, Fab, Fab', F(ab')2, or other antigen-binding subsequences of an antibody) is an antibody or immunoglobulin that is predominantly human, containing minimal sequences derived from non-human immunoglobulin. In most cases, a humanized antibody is a human immunoglobulin (recipient antibody) in which residues from the recipient’s hypervariable region (also known as the CDR) are replaced with residues from the hypervariable region of a non-human (e.g., rodent) species such as mouse, rat, hamster, or rabbit (donor antibody) having the desired specificity, affinity, and capability. In some cases, Fv framework region (FR) residues of the human immunoglobulin are replaced with corresponding non-human residues. Furthermore, “humanized antibodies” as used herein may also include residues not found in either the recipient antibody or the donor antibody. These modifications are made to further refine and optimize the performance of the antibody. A humanized antibody may also include the immunoglobulin constant region (Fc), typically at least a portion of the constant region of human immunoglobulin. For further 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).

[0083] Humanized antibodies or fragments thereof can be produced by replacing the sequence of the Fv variable domain, which is not directly involved in antigen binding, with an equivalent sequence derived from the human Fv variable domain. Exemplary methods for producing humanized antibodies or fragments thereof are provided by Morrison (1985) Science 229:1202-1207; Oi et al. (1986) BioTechniques 4:214 and U.S. Patents No. 5,585,089; U.S. Patents No. 5,693,761; U.S. Patents No. 5,693,762; U.S. Patents No. 5,859,205 and U.S. Patents No. 6,407,213. These methods involve isolating, manipulating, and expressing nucleic acid sequences encoding all or part of the immunoglobulin Fv variable domain derived from at least one of the heavy or light chains. Such nucleic acids can be obtained from hybridomas and other sources that produce antibodies against a given target, as described above. Next, recombinant DNA encoding a humanized antibody molecule can be cloned into a suitable expression vector.

[0084] Humanized antibodies can also be produced using transgenic animals, such as mice that express human heavy and light chain genes but cannot express endogenous mouse immunoglobulin heavy and light chain genes. Winter describes exemplary CDR transplantation methods that can be used for the preparation of humanized antibodies described herein (U.S. Patent No. 5,225,539). All of the CDRs of a particular human antibody may be replaced with at least some of the non-human CDRs, or only some of the CDRs may be replaced with non-human CDRs. Only the number of CDRs necessary for the humanized antibody to bind to a given antigen needs to be replaced.

[0085] Humanized antibodies can be optimized by introducing conservative substitutions, consensus sequence substitutions, germline substitutions, and / or reverse mutations. Such modified immunoglobulin molecules can be prepared by any of several techniques known in the art (e.g., Teng et al., Proc. Natl. Acad. Sci. USA, 80:7308-7312, 1983; Kozbor et al., Immunology Today, 4:7279, 1983; Olsson et al., Meth. Enzymol., 92:3-16, 1982; and European Patent No. 239400).

[0086] The terms “human antibody,” “human antigen-binding molecule,” and “human binding domain” include antibodies, antibody-binding molecules, and binding domains having antibody regions such as variable regions and constant regions or domains that substantially correspond to human germline immunoglobulin sequences known in the art, including those described in Kabat et al. (1991) (cited above). The human antibody, antigen-binding molecule, or binding domain of the present invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by in vitro random mutagenesis or site-directed mutagenesis or in vivo somatic mutation), for example, in the CDR, particularly CDR3. The human antibody, antigen-binding molecule, or binding domain may have at least one, two, three, four, five, or more positions replaced by amino acid residues not encoded by human germline immunoglobulin sequences. The definitions of human antibodies, antigen-binding molecules, and binding domains, as used herein, also include fully human antibodies that consist only of human sequences of non-artificially and / or genetically modified antibodies, such as those that can be obtained by using technologies or systems such as Xenomouse. Preferably, “fully human antibodies” do not contain amino acid residues not encoded by human germline immunoglobulin sequences.

[0087] In some embodiments, the antigen-binding molecules of the present invention are “isolated” or “substantially pure” antigen-binding molecules. When “isolated” or “substantially pure” is used in the description of the antigen-binding molecules disclosed herein, it means antigen-binding molecules identified, separated and / or recovered from components of their production environment. Preferably, the antigen-binding molecules do not associate with or substantially associate with all other components from their production environment. Contaminating components of their production environment, such as components arising from recombinant transfected cells, are typically materials that interfere with diagnostic or therapeutic applications relating to polypeptides and may include enzymes, hormones, and other proteolytic or non-proteolytic solutes. The antigen-binding molecules may constitute, for example, at least about 5% by weight or at least about 50% by weight of the total protein in a given sample. It is understood that isolated proteins may constitute 5% to 99.9% by weight of the total protein content, depending on the context. Polypeptides can be produced at significantly higher concentrations by using inducible promoters or high-expression promoters so that they are produced at increased concentration levels. This definition includes the generation of antigen-binding molecules in a wide variety of organisms and / or host cells known in the art. In preferred embodiments, the antigen-binding molecule is purified (1) by using a spinning cup sequencer to a degree sufficient to obtain at least 15 residues of the N-terminus or internal amino acid sequence, or (2) by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue or preferably silver staining until homogeneous. However, typically, the isolated antigen-binding molecule is prepared by at least one purification step.

[0088] In relation to the present invention, the term "binding domain" is considered to be a domain that (specifically) binds to / interacts with / recognizes a given target epitope or target site on a target molecule (antigen), for example, CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, or EpCAM and CD3, respectively. Typically, the structure and function of the first and third binding domains or the second and fourth binding domains (e.g., recognizing CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, or EpCAM), and preferably the structure and / or function of the effector binding domain (typically the second and fourth binding domains recognizing CD3 or the first and third binding domains), are based on the structure and / or function of an antibody, for example, a full-length or whole immunoglobulin molecule, and / or derived from the variable heavy chain (VH) and / or variable light chain (VL) domains of the antibody or a fragment thereof. Preferably, the target cell surface antigen-binding domain is 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 preferably also includes the minimum structural requirements of the antibody that enable target binding. More preferably, the second binding domain includes 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). The first and / or second binding domains are expected to be prepared or obtained by phage display or library screening, in addition to transplanting CDR sequences derived from existing (monoclonal) antibodies onto a scaffold.

[0089] According to the present invention, the binding domain is in the form of one or more polypeptides. Such polypeptides may comprise a protein portion and a non-protein portion (e.g., a chemical linker or a chemical crosslinking agent such as glutaraldehyde). Proteins (including fragments thereof, preferably biologically active fragments and peptides having typically fewer than 30 amino acids) comprise two or more amino acids linked to one another via covalent peptide bonds (resulting in a chain of amino acids).

[0090] As used herein, the term “polypeptide” typically refers to a group of molecules consisting of more than 30 amino acids. Polypeptides can further form multimers such as dimers, trimers, and higher-order oligomers, i.e., two or more polypeptide molecules. The polypeptide molecules forming such dimers, trimers, etc., may or may not be identical. The corresponding higher-order structures of such multimers are therefore referred to as homodimers or heterodimers, homotrimers or heterotrimers, etc. An example of a heteromultimer is an antibody molecule that, in its natural form, consists of two identical polypeptide light chains and two identical polypeptide heavy chains. The terms “peptide,” “polypeptide,” and “protein” also refer to naturally occurring modified peptides / polypeptides / proteins that have been modified by post-translational modifications such as glycosylation, acetylation, and phosphorylation. “Peptides,” “polypeptides,” or “proteins,” as referred herein, may also be chemically modified, such as pegylation. Such modifications are known in the art and are described below herein.

[0091] Preferably, the binding domain that binds to any of CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, CDH3, MSLN, and EpCAM, and / or the binding domain that binds to CD3ε, is a human binding domain. Antibodies and antigen-binding molecules containing at least one human binding domain avoid some of the problems associated with antibodies or antigen-binding molecules having non-human variable regions and / or constant regions, such as those from rodents (e.g., mice, rats, hamsters, or rabbits). The presence of such rodent-derived proteins can lead to rapid clearance of the antibody or antigen-binding molecule or to an immune response by the patient to the antibody or antigen-binding molecule. To avoid using rodent-derived antibodies or antigen-binding molecules, human or fully human antibody / antigen-binding molecules can be generated by introducing human antibody function into rodents so that the rodents produce fully human antibodies.

[0092] 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 can be covalently or non-covalently bonded. HMWS may typically consist of misfolded monomers, where the surface of the monomer that does not exist in monomeric form is exposed. These should generally be avoided from a regulatory standpoint, although they may or may not affect safety and / or efficacy. Typically, in relation to the present invention, HMWS may be determined by size exclusion chromatography (SE-UHPLC) and may represent a "main peak," i.e., a fraction different from the desired monomeric product.

[0093] The terms “amino acid” or “amino acid residue” typically refer to amino acids with a definition recognized in the art, such as those 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 (Glu or E), glycine (Gly 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), but modified amino acids, synthetic amino acids, or rare amino acids may be used as needed. Generally, amino acids can be classified by the presence of nonpolar side chains (e.g., Ala, Cys, Ile, Leu, Met, Phe, Pro, Val), negatively charged side chains (e.g., Asp, Glu), positively charged side chains (e.g., Arg, His, Lys), or uncharged polar side chains (e.g., Asn, Cys, Gln, Gly, His, Met, Phe, Ser, Thr, Trp, and Tyr).

[0094] Amino acid modification includes, for example, deletions from and / or insertions of residues 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 has the desired properties. Amino acid changes may also alter the post-translational processes of the antibody construct, such as changes in the number or location of glycosylation sites.

[0095] For example, (naturally, depending on their lengths) one, two, three, four, five, or six amino acids may be inserted, substituted, or deleted in each CDR, while one, two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, three, four, five, six, seven, eight, nine, ten, ten, ten, ten, ten, three, three, four, five, six, seven, eight, nine, ten, six, seven, eight, nine, ten, six, seven, eight, eight, nine, or ten residues, and one or more intrasequence insertions of amino acid residues may be performed. Corresponding modifications may also be carried out within a third domain of the antibody construct of the present invention. Insertion variants of the antibody construct of the present invention include fusion of an enzyme to the N-terminus or C-terminus of the antibody construct or fusion to a polypeptide.

[0096] The most important sites for substitutional mutagenesis are the CDRs of the heavy chain and / or light chain, particularly the hypervariable region (HDR), but modifications of the FRs in the heavy chain and / or light chain are also intended. Substitutions are preferably conservative substitutions as described herein. Preferably, depending on the length of the CDR or FR, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acids may be substituted in the CDR, 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 may be substituted in the framework region (FR). For example, if the CDR sequence contains 6 amino acids, 1, 2, or 3 of these amino acids are expected to be substituted. Similarly, if the CDR sequence contains 15 amino acids, 1, 2, 3, 4, 5, or 6 of these amino acids are expected to be substituted.

[0097] A useful method for identifying specific residues or regions in an antibody construct that are favorable for mutagenesis is called the "alanine scanning mutagenesis method," as described in Cunningham and Wells in Science, 244:1081-1085 (1989). This method identifies residues or target residue groups within the antibody construct (e.g., charged residues such as arg, asp, his, lys, and glu) and replaces them with neutral or negatively charged amino acids (most preferably alanine or polyalanine) that influence the interaction between the amino acid and the epitope.

[0098] Next, by introducing further or other variants at the substitution site, i.e., in place of the substitution site, the range of amino acid positions that are functionally sensitive to the substitution is carefully selected. Thus, although the site or region to introduce the amino acid sequence mutation is predetermined, the nature of the mutation itself does not need to be predetermined. For example, to analyze or optimize the performance of a mutation at a given site, alanine scanning or random mutagenesis may be performed at the target codon or target region, and variants of the expressed antibody construct are screened for the optimal combination of desired activity. Techniques for introducing substitutional mutations at predetermined sites in DNA with known sequences are well known, such as M13 primer mutagenesis and PCR mutagenesis. Screening of variants is performed using antigen-binding activity assays, such as target cell surface antigen or CD3 binding.

[0099] Generally, when one or more or all of the amino acids in the heavy chain and / or light chain CDRs are substituted, the resulting "substituted" sequence is preferably 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 sequence. This means that the degree of identity with the "substituted" sequence depends on the length of the CDR. For example, for a CDR with five amino acids to have at least one substituted amino acid, it is preferable that it be 80% identical to the substituted sequence. Therefore, the CDRs of an antibody construct may have varying degrees of identity with respect to their substituted sequences; for example, CDRL1 may have 80% identity, while CDRL3 may have 90%.

[0100] A preferred substitution (or replacement) is a conservative substitution. However, any substitution (including non-conservative substitutions or one or more of the “exemplary substitutions” listed in Table 3 below) is conceivable, as long as the antibody construct retains its ability to bind to the surface antigen of the target cell via the first domain and to CD3 or CD3 epsilon via the second domain, and / or its CDR is identical to the substituted sequence (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 sequence).

[0101] Conservative substitutions are shown under the heading “Preferred Substitutions” in Table 3. If such substitutions alter biological activity, they are referred to as “Exemplary Substitutions” in Table 3, or substantial changes may be introduced, as further described below by referring to classes of amino acids, and the product may be screened for desired characteristics.

[0102] [Table 1]

[0103] Substantial modification of the biological properties of the antibody construct of the present invention is achieved by selecting substitutions that have a significantly different effect on (a) the structure of the polypeptide backbone of the substitution region, for example, as a sheet-like or helical three-dimensional structure, (b) the charge or hydrophobicity of the molecule at the target site, or (c) the maintenance of the bulkiness of the side chain. Naturally occurring residues are classified into the following groups based on common side chain properties: (1) hydrophobic: norleucine, met, ala, val, leu, ile, (2) neutral hydrophobic: cys, ser, thr, asn, gln, (3) acidic: asp, glu, (4) basic: his, lys, arg, (5) residues that affect chain orientation: gly, pro, and (6) aromatic: trp, tyr, phe.

[0104] Non-conservative substitutions involve exchanging one member of one class for another. To avoid abnormal crosslinking, any cysteine ​​residue that does not contribute to maintaining the proper conformation of the antibody construct can generally be substituted with serine to improve the oxidative stability of the molecule. Conversely, adding cysteine ​​bonds to an antibody can improve its stability (especially if the antibody is an antibody fragment such as an Fv fragment).

[0105] With respect to amino acid sequences, sequence identity and / or similarity are determined using standard techniques known in the art, such as, 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 similarity search method of Pearson and Lipman, 1988, Proc. Nat. Acad. Sci. USA 85:2444, computer execution of these algorithms (GAP, BESTFIT, FASTA, and TFASTA from Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Dr., Madison, Wis), or by visual inspection using the Best Fit sequence program, preferably with default settings, as described by Devereux et al., 1984, Nucl. Acid Res. 12:387-395. Preferably, the identity percentage is calculated by FastDB based on the following parameters: a mismatch penalty of 1, a gap penalty of 1, a gap size penalty of 0.33, and a join 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.)

[0106] One example of a useful algorithm is PILEUP. PILEUP uses progressive pairwise alignment to create multiple sequence alignments from a group of related sequences. This also allows plotting a tree showing the clustering relationships used to generate the alignments. PILEUP uses a simplified progressive alignment method from Feng & Doolittle, 1987, J.Mol.Evol.35:351-360, and the method is similar to that described in 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 a weighted end gap.

[0107] 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.USA90:5873-5787. 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 their default values. The adjustable parameters are set to the following values: overlap span = 1, overlap fraction = 0.125, word threshold (T) = II. The HSP S-parameter and HSP S2-parameter are dynamic values, constructed by the program itself depending on the composition of a particular sequence and the composition of a particular database from which the target sequence is searched, but their values ​​can be adjusted to increase sensitivity.

[0108] Another useful algorithm is GapBLAST, reported by Altschul et al., 1993, Nucl. Acids Res. 25:3389-3402. GapBLAST uses BLOSUM-62 substitution scores, with the threshold T parameter set to 9. In the two-hit method that yields gapless extension, the cost of the gap length k is 10+k, Xu is set to 16, and Xg is set to 40 during the database search phase and 67 during the algorithm's output phase. Gap alignment is provided by a score corresponding to approximately 22 bits.

[0109] Generally, the amino acid homology, similarity, or identity between individual variant CDR or VH / VL sequences is at least 60% with respect to the sequences shown herein, and more typically, the homology or identity increases to at least 65% or 70%, more preferably at least 75% or 80%, and even more preferably at least 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, and nearly 100%. Similarly, the “percentage of nucleic acid sequence identity (%)” for the nucleic acid sequences of the binding proteins identified herein is defined as the percentage of nucleotide residues in the candidate sequence that are identical to nucleotide residues in the coding sequence of the antibody construct. Specifically, the BLASTN module of WU-BLAST-2 is used with default parameters where the overlap span and overlap fraction are set to 1 and 0.125, respectively.

[0110] Generally, the nucleic acid sequence homology, similarity, or identity between the nucleotide sequences encoding individual variant CDR or VH / VL sequences and the nucleotide sequences shown herein is at least 60%, and more typically, it is preferable that the homology or identity increases to 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 nearly 100%. Accordingly, the "variant CDR" or "variant VH / VL region" has specific homology, similarity or identity with respect to the parent CDR / VH / VL of the present invention and shares a biological function that includes, but is not limited to, 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.

[0111] In one embodiment, the percentage of identity of the antibody construct according to the present invention to human germline is ≥70% or ≥75%, more preferably ≥80% or ≥85%, even more preferably ≥90%, most preferably ≥91%, ≥92%, ≥93%, ≥94%, ≥95%, or even further ≥96%. Identity to human antibody germline gene products is considered an important feature for reducing the risk that therapeutic proteins will induce an immune response to a drug in a patient undergoing treatment. Hwang & Foote ("Immunogenicity of engineered antibodies"; Methods 36(2005)3-10) have demonstrated that reducing the non-human portion of a drug-antibody construct reduces the risk of inducing anti-drug antibodies in a patient undergoing treatment. By comparing a vast number of clinically evaluated antibody drugs and their respective immunogenicity data, humanization of the V region of antibodies tends to result in lower protein immunogenicity (average 5.1% of patients) compared to antibodies possessing the unmodified, non-human V region (average 23.59% of patients). Therefore, for V region-based protein therapeutics in the form of antibody constructs, a high degree of identity with the human sequence is desirable. To determine this germline identity, Vector NTI software can be used to align the V region of VL with the amino acid sequences of human germline V and J segments (http: / / vbase.mrc-cpe.cam.ac.uk / ), and the amino acid sequence can be calculated as a percentage by dividing the number of identical amino acid residues by the total number of amino acid residues in VL. A similar method is possible for the VH segment (http: / / vbase.mrc-cpe.cam.ac.uk / ), however, VH CDR3 may be excluded due to its high diversity and the lack of existing human germline VH CDR3 alignment partners. Next, recombinant technology can be used to increase sequence identity for human antibody germline genes.

[0112] In further embodiments, the bispecific antibody constructs of the present invention exhibit high monomer yields under standard research-scale conditions (e.g., a standard two-step purification process). Preferably, the monomer yield of the antibody construct according to the present invention is ≥0.25 mg per L of supernatant, more preferably ≥0.5 mg per L, even more preferably ≥1 mg per L, and most preferably ≥3 mg per L of supernatant.

[0113] Similarly, the yield of isoforms of the dimeric antibody construct, and therefore the percentage of monomers of the antigen-binding molecule (i.e., monomer / (monomer + dimer)), can be determined. The productivity of monomer and dimeric antibody constructs, and the calculated monomer percentages, can be obtained, for example, in an SEC purification step of the culture supernatant derived from standardized study-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%.

[0114] In one embodiment, the antibody construct has plasma stability (ratio of EC50 in the presence of plasma to EC50 in the absence of plasma) preferably ≤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 is determined by incubating the construct in human plasma at 37°C for 24 hours, followed by 51This can be verified by determining the EC50 using a chromium-releasing cytotoxicity assay. Effector cells in the cytotoxicity assay may be stimulated, enriched human CD8-positive T cells. Target cells may be, for example, CHO cells transfected with the surface antigen of human target cells. An effector cell to target cell (E:T) ratio of 10:1 can be selected. The human plasma pool used for this purpose is derived from the blood of healthy donors collected using EDTA-coated syringes. Cellular components are removed by centrifugation, the upper plasma phase is collected, and then pooled. As a control, the antibody construct is diluted in RPMI-1640 medium immediately before the cytotoxicity assay. Plasma stability is calculated as the ratio of EC50 (after plasma incubation) to EC50 (control).

[0115] It is even more preferable that the monomer-to-dimer conversion rate of the antibody construct of the present invention is low. The conversion rate can be measured under different conditions and analyzed by high-speed size exclusion chromatography. For example, incubation of the monomer isoform of the antibody construct can be carried out in an incubator at a concentration of, for example, 100 μg / ml or 250 μg / ml at 37°C for 7 days. Under these conditions, it is preferable that the antibody construct of the present invention exhibits a dimer ratio 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%, ≤0.5%, or even more preferably 0%.

[0116] The bispecific antibody construct of the present invention may also preferably exhibit a very low dimerization rate after several freeze / thaw cycles. For example, the antibody construct monomer is adjusted to a concentration of 250 μg / ml in a general formulation buffer, subjected to three freeze / thaw cycles (freezing at -80°C for 30 minutes, followed by thawing at room temperature for 30 minutes), and then a fast SEC is performed to determine the percentage of the initial monomer antibody construct converted to a dimer antibody construct. Preferably, the dimerization rate of the bispecific antibody construct is, for example, ≤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 further ≤0.5% after three freeze / thaw cycles.

[0117] The bispecific antibody construct of the present invention preferably exhibits good 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. From the viewpoint of antibody aggregation temperature, the thermal stability parameter can be determined as follows: A 250 μg / ml antibody solution is transferred to a single-use cuvette and placed in a dynamic light scattering (DLS) apparatus. The sample is heated from 40°C to 70°C at a heating rate of 0.5°C / min while continuously acquiring the measured radius. An increase in radius indicates protein melting and aggregation, which is used to calculate the antibody aggregation temperature.

[0118] Alternatively, the melting temperature curve can be determined by differential scanning calorimetry (DSC) to determine the intrinsic biophysical protein stability of the antibody construct. These experiments are performed using a MicroCal LLC (Northampton, MA, USA) VP-DSC instrument. Energy uptake of a sample containing the antibody construct is recorded from 20°C to 90°C and compared with a sample containing only the formulation buffer. The antibody construct is adjusted to a final concentration of 250 μg / ml in, for example, SEC running buffer. The temperature of the entire sample is gradually increased to record each melting curve. Energy uptake of the sample and the formulation buffer standard is recorded at each temperature T. The difference in energy uptake Cp (kcal / mole / °C) obtained by subtracting the standard from the sample is plotted against each temperature. The melting temperature is defined as the temperature at the first maximum value of energy uptake.

[0119] The target cell surface antigen × CD3 bispecific antibody construct of the present invention may also have a turbidity of ≤0.2, preferably ≤0.15, more preferably ≤0.12, even more preferably ≤0.1, and most preferably ≤0.08 (measured by OD340 after concentrating the purified monomer antibody construct to 2.5 mg / ml and incubating overnight).

[0120] The bispecific antibody constructs of the present invention are further expected to exhibit therapeutic efficacy or antitumor activity. This can be evaluated, for example, in the tests disclosed in the following examples of advanced-stage human tumor xenograft models.

[0121] Those skilled in the art know how to obtain meaningful and reproducible results while modifying or adapting specific parameters of the present test, such as the number of tumor cells injected, the injection site, the number of human T cells transplanted, the amount of bispecific antibody construct administered, and the schedule. Preferably, the tumor growth inhibitory T / C [%] is ≤70 or ≤60, more preferably ≤50 or ≤40, even more preferably ≤30 or ≤20, most preferably ≤10 or ≤5, or even further ≤2.5.

[0122] In a preferred embodiment of the antibody construct of the present invention, the antibody construct is a single-chain antibody construct.

[0123] In a preferred embodiment of the antibody construct of the present invention, the third domain is arranged in the order of amino to carboxyl, Hinge-CH2-CH3-Linker-Hinge-CH2-CH3 This also includes.

[0124] In one embodiment of the present invention, the CH2 domain of one or preferably each (both) of the third domains of the polypeptide monomer also includes an intradomain cysteine ​​disulfide crosslink. As is well known in the art, the term "cysteine ​​disulfide crosslink" refers to a functional group having the general structure RSSR. This linkage is also called an SS bond or disulfide crosslink and is obtained by the coupling of two thiol groups of a cysteine ​​residue. With respect to the antibody construct of the present invention, it is particularly preferable that the cysteine ​​forming the cysteine ​​disulfide crosslink in the mature antibody construct of the present invention be introduced into the amino acid sequence of the CH2 domain corresponding to 309 and 321 (Kabat numbering).

[0125] In one embodiment of the present invention, the glycosylation site at Kabat position 314 of the CH2 domain is removed. This removal of the glycosylation site is preferably achieved by an N314X substitution, where X is any amino acid other than Q. The substitution is preferably an N314G substitution. In a more preferred embodiment, the CH2 domain further comprises the following substitutions (positions according to Kabat): V321C and R309C (these substitutions introduce intradomain cysteine ​​disulfide crosslinks at Kabat positions 309 and 321).

[0126] For example, a preferred feature of the antibody construct of the present invention compared to a bispecific hetero-Fc antibody construct known in the art (Figure 1b) may be related, in particular, to the introduction of the above-mentioned modification in the CH2 domain. Accordingly, with respect to the construct of the present invention, it is preferable that the CH2 domain in the third domain of the antibody construct of the present invention contains intradomain cysteine ​​disulfide crosslinks at Kabat positions 309 and 321, and / or that the glycosylation site at Kabat position 314 is removed by N314X substitution, preferably N314G substitution, as described above.

[0127] In a more preferred embodiment of the present invention, the CH2 domain in the third domain of the antibody construct of the present invention includes intradomain cysteine ​​disulfide crosslinks at Kabat positions 309 and 321, and the glycosylation site at Kabat position 314 is removed by N314G substitution.

[0128] In one embodiment, the present invention is an antibody construct, (i) The first domain contains two antibody variable domains, and the second domain contains two antibody variable domains, (ii) The first domain contains one antibody variable domain, and the second domain contains two antibody variable domains, (iii) The first domain contains two antibody variable domains and the second domain contains one antibody variable domain, or (iv) Provide an antibody construct in which the first domain contains one antibody variable domain and the second domain contains one antibody variable domain.

[0129] Accordingly, the first and second domains may be binding domains containing two antibody-variable domains, such as a VH domain and a VL domain. Examples of such binding domains containing two antibody-variable domains as described above in this specification include, for example, the Fv fragment, scFv fragment, or Fab fragment as described above in this specification. Alternatively, one or both of these binding domains may contain only a single variable domain. Examples of such single-domain binding domains as described above in this specification include, for example, nanobody or single-variable-domain antibodies containing only one variable domain, which may be VHH, VH, or VL, that specifically binds to an antigen or epitope independently of other V regions or domains.

[0130] In preferred embodiments of the antibody construct of the present invention, the first and second domains are fused to a third domain via a peptide linker. Preferred peptide linkers are described above herein 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 (where x is an integer of 1 or more (e.g., 2 or 3)). A particularly preferred linker for the fusion of the first and second domains to the third domain is shown in SEQ ID NO: 1.

[0131] In a preferred embodiment, the antibody construct of the present invention is composed of amino to carboxyl molecules in that order. (a) First domain, (b) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs. 187-189, (c) Second domain, (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198. (e) The first polypeptide monomer of the third domain, (f) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 191, 192, 193, and 194, and (g) Second polypeptide monomer of the third domain It is characterized by including.

[0132] In one aspect of the present invention, the surface antigen of a target cell bound by the first domain is a tumor antigen, an antigen specific to an immunodeficiency, or a viral antigen. As used herein, the term “tumor antigen” may be understood as those antigens presented on tumor cells. These antigens may be presented on the cell surface along with their extracellular components, often comprising both transmembrane and cytoplasmic portions of the molecule. These antigens may, in some cases, be presented only by tumor cells and never by normal cells. Tumor antigens may be expressed exclusively on tumor cells or may exhibit tumor-specific mutations compared to normal cells. In this case, they are called tumor-specific antigens. More general antigens are those presented by both tumor cells and normal cells, and these are called tumor-associated antigens. These tumor-associated antigens may be overexpressed compared to normal cells or, due to the less compact structure of tumor tissue compared to normal tissue, are accessible to tumor cells for antibody binding. Non-exclusive examples of tumor antigens used herein include CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, MUC17, CLDN18.2, CDH3, CD70, BCMA, and PSMA.

[0133] In preferred embodiments of the antibody construct of the present invention, the tumor antigen is preferably selected from the group consisting of CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, MUC17, CLDN18.2, CDH3, CD70, BCMA, and PSMA.

[0134] In one aspect of the present invention, the antibody construct is composed of amino acids followed by carboxyl molecules. (a) Sequence numbers 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, 131, 1 A first domain having an amino acid sequence selected from the group consisting of 32, 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, (b) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs. 187-189, (c) A second domain having an amino acid sequence selected from the group consisting of SEQ ID NOs. 23, 25, 41, 43, 59, 61, 77, 79, 95, 97, 113, 115, 131, 133, 149, 151, 167, 169, 185 or 187 or SEQ ID NO. 202 from International Publication No. 2008 / 119567, (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198. (e) A first polypeptide monomer of a third domain having a polypeptide sequence selected from the group consisting of SEQ ID NOs: 249-256, (f) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 191, 192, 193, and 194, and (g) A second polypeptide monomer of a third domain having a polypeptide sequence selected from the group consisting of SEQ ID NOs. 249-256. Includes.

[0135] In one embodiment, the bispecific antibody construct of the present invention comprises the following, and is characterized by having an amino acid sequence selected from the group that is induced by the surface antigen of each target cell. (a) Sequence IDs 27, 28, 37-41; CD33 (b) Each of sequence numbers 48-52; EGFRvIII (c) Each of sequence numbers 59-64; MSLN (d) Each of sequence numbers 71-82; CDH19 (e) Each of sequence numbers 100 to 104; DLL3 (f) Sequence IDs 7, 8, 17, 113 and 114; CD19 (g) Each of sequence numbers 89-93; FLT3 (h) Each of sequence numbers 121-125; CDH3 (i) Each of sequence numbers 132-136; BCMA (j) Sequence IDs 143-151, 158-166 and 173-181 respectively; PSMA (k) Sequence ID 213; MUC17 (l) Sequence IDs 225 and 237, respectively; CLDN18.2, and (m) Sequence ID 248; CD70

[0136] In some embodiments, the bispecific antibody construct includes a first binding domain that binds to DLL3, which includes an anti-DLL3 variable light domain.

[0137] In some embodiments, the bispecific antibody construct includes a first binding domain that binds to BCMA, which includes an anti-BCMA variable light domain.

[0138] In some embodiments, the bispecific antibody construct includes a first binding domain that binds to CD33, which includes an anti-CD33 variable light domain.

[0139] In some aspects of this disclosure, proteins, such antibodies, or bispecific antibody constructs are present in the liquid formulation (before lyophilization) in amounts 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, proteins are optionally present in the liquid formulation in amounts ranging from about 0.1 mg / mL to about 70 mg / mL. In some cases, the protein is present in the liquid formulation in amounts ranging from approximately 0.5 mg / mL to approximately 30 mg / mL (or approximately 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 various cases, the protein is present in liquid formulations in a range of approximately 1 mg / mL to approximately 20 mg / mL (or approximately 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). It is present in amounts of L, 10.5 mg / 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). In some embodiments, the protein is present in the liquid formulation in an amount of approximately 21 mg / mL.

[0140] The protein formulations of this disclosure optionally contain sugars. In some embodiments, the sugars are monosaccharides or disaccharides. Preferred sugars include, for example, glucose, galactose, fructose, xylose, sucrose, lactose, maltose, trehalose, or any combination thereof. In some examples, the sugars include trehalose.

[0141] In some embodiments, the liquid formulation (before freeze-drying) contains sugars at a concentration of about 1% to about 15% w / v, or about 4% to about 13% w / v, or about 6% to about 12% w / v. In some embodiments, the liquid formulation contains sugars at a concentration of 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. In some embodiments, the liquid formulation contains sugars at a concentration of 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. In some embodiments, the liquid formulation contains sugars at concentrations of 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 some embodiments, the liquid formulation contains sugars at concentrations of about 7% to about 12% w / v. In some embodiments, the liquid formulation contains sugars at a concentration of about 9% w / v. In some embodiments, the sugar is sucrose and is present in the liquid formulation at concentrations ranging from about 6% to about 12% w / v.

[0142] The protein formulations of this disclosure optionally include a surfactant. Preferred surfactants include polysorbate, poloxamer, polyoxyethylene, or any combination thereof. Intended surfactants include polysorbate 20, polysorbate 40, polysorbate 60, polysorbate 80, poloxamer 188, poloxamer 407, Triton X-100, polyoxyethylene, PEG3350, PEG4000, and any combination thereof. In some embodiments, the surfactant includes polysorbate. In some examples, the surfactant is polysorbate 20.

[0143] The protein formulations described herein may contain one surfactant or a mixture of surfactants (but this is not required). In some embodiments, the liquid formulation (before lyophilization) contains 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 embodiments, the liquid formulation contains a surfactant at a concentration of 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. In some embodiments, the liquid formulation contains a surfactant at a concentration of about 0.001% to about 0.5% w / v. In some embodiments, the liquid formulation contains a surfactant at a concentration of about 0.001% to about 0.01% w / v. In some embodiments, the liquid formulation contains a surfactant at a concentration of about 0.001% to about 0.01% w / v. In some embodiments, the liquid formulation contains a surfactant at a concentration of approximately 0.001%, 0.002%, 0.003%, 0.004%, 0.005%, 0.006%, 0.007%, 0.008%, 0.009%, 0.01%, 0.05%, 0.1%, 0.2%, 0.3%, and approximately 0.4% to 0.5% w / v. In some embodiments, the liquid formulation contains a surfactant at a concentration of approximately 0.001% to 0.01% w / v.

[0144] The protein formulations of this disclosure optionally include a buffer. Suitable buffers include acetate buffers, glutamate buffers, citrate buffers, lactate buffers, succinate buffers, tartarate buffers, fumarate buffers, maleate buffers, histidine buffers, phosphate buffers, 2-(N-morpholino)ethanesulfonic acid buffers, or any combination thereof. In some examples, the buffer includes histidine.

[0145] Buffers are often used to control the pH of a formulation. In some embodiments, buffers are added at concentrations that maintain the pH of the liquid formulation at approximately 3–7, 4–6, 4–5, 5–6, or 6–6.5. The effect of pH on the formulation can be characterized using one or more of several methods, such as accelerated stability testing and calorimetry screening tests (Remmele RLJr., et al., Biochemistry, 38(16):5241-7(1999)).

[0146] The buffer system (if present in the protein formulation) is selected to be physiologically compatible and to maintain the desired pH. The buffer may be present in the liquid formulation (before lyophilization) at concentrations of approximately 0.1 mM to approximately 1000 mM (1 M), or approximately 5 mM to approximately 200 mM, or approximately 5 mM to approximately 100 mM, or approximately 10 mM to approximately 50 mM. Suitable buffer concentrations include concentrations of approximately 200 mM or less. In some embodiments, the buffer is present in the liquid formulation (before freeze-drying) at concentrations of approximately 190 mM, 180 mM, 170 mM, 160 mM, 150 mM, 140 mM, 130 mM, 120 mM, 110 mM, 100 mM, 80 mM, 70 mM, 60 mM, 50 mM, 40 mM, 30 mM, 20 mM, 10 mM, or 5 mM. In some embodiments, the concentration of the buffer is at least 0.1, 0.5, 0.7, 0.8, 0.9, 1.0, 1.2, 1.5, 1.7, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 500, 700, or 900 mM. In some embodiments, the concentration of the buffer is 1, 1.2, 1.5, 1.7, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, 40, 50, 60, 70, 80, or 90 mM to 100 mM. In some embodiments, the concentration of the buffer is 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, or 40 mM to 50 mM.

[0147] In a further embodiment, a kit is provided comprising the lyophilized protein composition described herein, packaged in a manner that facilitates administration to a subject. In one embodiment, the kit comprises the lyophilized protein composition described herein, packaged in a container such as a sealed bottle, vessel, single-use or multi-use vial, a pre-filled device (e.g., a syringe), or a pre-filled injection device, and optionally having a label affixed to the container or contained in packaging that describes the use of the lyophilized protein composition. In one embodiment, the pharmaceutical composition is packaged in unit dosage forms. The kit may include a device suitable for administering the re-dissolved protein composition according to a specific route of administration (but this is not required). For example, this disclosure provides a dual-chamber device for delivering the re-dissolved protein composition disclosed herein to a subject requiring it. The dual-chamber device is a combination product containing the lyophilized protein composition and diluent in two separate chambers of the device. A pre-filled dual-chamber device is a combination product containing the lyophilized drug and diluent in two separate chambers of the device. Dual-chamber devices suitable for use with this disclosure are described in the Art. For example, see Ingle R., Fang W. (2021). Int. Journal of Pharmaceutics 597, 12031.

[0148] The formulations described herein are useful as pharmaceutical formulations for the treatment or improvement of cancer in subjects requiring treatment or improvement of cancer. The terms “subjects requiring treatment” or “subjects requiring treatment” include subjects already suffering from the disorder and subjects in whom the disorder should be prevented. “Subjects requiring treatment” or “patients” include human and other mammalian subjects receiving either prophylactic or therapeutic treatment. “Treatment” does not require complete remission or eradication of the disease, but aims at any improvement of the disease and / or improvement of symptoms associated with the disease. For example, a therapeutic response refers to one or more of the following improvements in the disease: (1) a reduction in the number of neoplastic cells, (2) an increase in neoplastic cell death, (3) an inhibition of neoplastic cell survival, (4) an inhibition (i.e., slowing, preferably stopping) of tumor growth or the appearance of new lesions, (5) a delay in disease progression, (6) an increase in patient survival rate, (7) a downgrade of the stage of cancer (e.g., from stage 2 to stage 1), and / or (8) some relief from one or more symptoms associated with the disease or condition. This composition may also be administered to achieve disease prevention or delay of onset, for example, avoidance of the development or recurrence of tumors or cancer. The disease state is monitored, for example, by clinical examinations, X-rays, 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 typically also used to diagnose and stage cancer.

[0149] This disclosure provides a method for treating cancer, comprising administering a therapeutically effective amount of a redissolved composition based on a lyophilized formulation described herein to a subject in need thereof. In certain embodiments, the subject is a human. In certain embodiments, the cancer is a solid tumor. In some embodiments, the cancer is a brain tumor, bladder cancer, breast cancer (e.g., trinegative breast cancer), clear cell kidney cancer, cervical cancer, colorectal cancer, endometrial cancer, gastric cancer, head and 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 embodiments, the cancer is acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), or chronic myeloid leukemia (CML). In some embodiments, 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 disclosure also provides the use of redissolvable compositions based on lyophilized formulations in methods for treating cancer and the use of lyophilized formulations in the preparation of pharmaceuticals for treating cancer.

[0150] Preferably, the pharmaceutical formulation is administered parenterally, for example, intravenously, subcutaneously, intratumorally, or intramuscularly. Parenteral administration can be achieved by injection, such as a bolus injection, or by infusion, such as a continuous infusion. Administration can be achieved via depot for long-term release. In some embodiments, the formulation is administered intravenously by continuous infusion after an initial bolus to maintain therapeutic circulating levels of the drug product. In some embodiments, the formulation is administered as a single dose. The pharmaceutical formulation may 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.

[0151] The following examples illustrate typical features of the present disclosure. Based on the description of these embodiments, other embodiments of the present invention may be manufactured and / or carried out based on the description provided below. The method involves the use of molecular biological techniques described in specialized books such as Molecular Cloning: A Laboratory Manual, 2nd ed., vol. 1-3, Sambrook et al., ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001; and Current Protocols in Molecular Biology, Ausubel et al., ed., Greene Publishing and Wiley-Interscience, New York. The examples are for illustrative purposes only and are not intended to limit the scope of the present invention. [Examples]

[0152] This embodiment describes an alternative method including an annealing step and a process for preparing a lyophilized bispecific molecular composition according to the present invention, the latter of which yields desirable product characteristics such as a low percentage of HMWS and a desirable process time.

[0153] The use of annealing and controlled ice nucleation are two methods for addressing the challenges associated with freeze-drying, including aggregation and long processing times. Both annealing and controlled ice nucleation address this problem by reducing ice heterogeneity by generating larger ice crystals, resulting in the formation of a porous cake matrix with reduced resistance to mass transfer during sublimation.

[0154] Annealing involves holding the product between the glass transition temperature and the eutectic melting point after stochastic ice nucleation (Figure 1A). This causes the ice crystals to melt, with the smallest ones completely disappearing and the larger ones remaining. Subsequent freezing then further grows the larger crystals. Larger ice crystals result in larger pores in the freeze-dried cake, which means lower resistance for vapor passing through the dry layer and faster primary drying. However, holding the product above the glass transition temperature increases protein mobility and proximity, which can promote the formation of high molecular weight species (HMWS) in some products. This problem has been observed with bispecific molecule products (e.g., BiTE® molecules) with protein concentrations greater than 5 mg / mL. In addition, annealing prolongs the freezing phase of freeze-drying, partially offsetting the reduction in primary drying time.

[0155] Controlled ice nucleation (CIN) generates larger ice crystals using different methods. For example, one CIN technique involves injecting a seed crystal to induce ice nucleation (Figure 1B). This induces nucleation while the sample is slightly supercooled, and the crystal can be further grown using isothermal holding after nucleation. Lower degrees of supercooling result in larger ice crystals than those produced by stochastic nucleation, and shorter isothermal holding times may reduce the risk of HMWS formation compared to annealing. In this example, we evaluated bispecific molecular drug products with "high protein concentrations" (defined in this example as 15 mg / mL or higher) by comparing CIN lyophilization cycles with cycles without the CIN step, cycles with annealing, and cycles without annealing of single-stranded bispecific T cell engager molecules. BCMAxCD3 bispecific molecule (20 mg / mL) and DLL3xCD3 bispecific molecule (15 mg / mL) were used as models for high-concentration bispecific molecular products. Annealing is currently a typical option for lyophilized bispecific molecular drug products. Therefore, such lyophilization cycles were used as representative annealing lyophilization cycles in this specification. An "annealing-free cycle" was created by removing the annealing step from the lyophilization cycle. The CIN lyophilization cycle, initially developed for BCMAxCD3 bispecific molecule 1 mg 6R SKU7, was adapted for use with high-concentration bispecific molecular formulations. The lyophilization cycles were evaluated in terms of total cycle time and product quality.

[0156] Materials and methods These experiments used BCMAxCD3 bispecific molecule (20 mg / mL) and DLL3xCD3 bispecific molecule (15 mg / mL). Table 1 shows details of the active pharmaceutical ingredients (DS) used.

[0157] [Table 2]

[0158] BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules (DS) were stored at -70°C and -30°C, respectively. Before use, DS was thawed at room temperature for 72 hours. 150 mL aliquots were taken and mixed by inversion for 2 minutes. Then, before filling vials, DS was filtered through a 0.22 μm polyvinylidene fluoride (PVDF) filter.

[0159] Tables 2 and 3 list the equipment and vial components used in this study. The 6R vials were treated with a pyrotherm before use. The sterile stoppers were stored in autoclavable packaging before use, but the sterile seal was not maintained after opening.

[0160] [Table 3]

[0161] [Table 4]

[0162] Summary of the experiment The purpose of these experiments was to compare the effects of conventional lyophilization cycles with and without annealing with the CIN lyophilization cycle for high-concentration BCMAxCD3 bispecific molecule models (20 mg / mL) and DLL3xCD3 bispecific molecule models (15 mg / mL). The bispecific molecule lyophilization cycle was used as the conventional lyophilization cycle with and without annealing. The CIN lyophilization cycle used BCMAxCD3 bispecific molecule 1 mg 6R SKU. 7 This is based on past CIN development. A summary of all three freeze-drying cycles can be found in Table 4.

[0163] [Table 5]

[0164] Freeze-drying procedure Prior to filling, the Millrock lyophilization tray was completely filled using 6R vials. 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 presentations filled with 1.3 mL in the 6R vials were selected as high-concentration representations of BiTE, consisting of bispecific molecule formulations of 5 mg / mL or less filled with 1.28 mL in the 6R vials. Two sets of vials were placed side by side in one tray for lyophilization. Thermocouples were placed in front, in the middle, and behind the filled samples.

[0165] Table 5 lists the freeze-drying cycle parameters for annealed and non-annealed freeze-drying cycles, as well as the CIN cycle for the BCMAxCD3 bispecific molecule. The primary and secondary freeze-drying parameters remain unchanged. However, the nucleation conditions were adjusted to minimize the risk of HMWS formation. This involved lowering the nucleation temperature from the less favorable -7°C to the more favorable -12, -15, or -17°C, setting the post-nucleation time to 90 minutes, or further shortening the retention time from 1 hour to 30 or 20 minutes (Figure 3). These changes resulted in a CIN freeze-drying stage, reducing the time the product was in a low-temperature concentrated "slushy" state and decreasing the chance of HMWS formation.

[0166] [Table 6]

[0167] Analysis of freeze-dried data Freeze-drying process data was exported from the instrument, and the primary drying time, total cycle time, and product temperature were analyzed. Primary drying was considered to begin at the start of the primary drying temperature gradient. The end of the primary drying time was determined by observing when the instantaneous gradient of the Pirani vacuum gauge approached zero. The total freeze-drying time was the sum of the freezing, primary drying, and secondary drying stages of the freeze-drying cycle.

[0168] The maximum product temperature for each vial containing a thermocouple was determined as the highest temperature observed while the thermocouple remained within the freeze matrix during primary drying. This was performed for each thermocouple to obtain the product temperature range for the entire freeze-drying tray.

[0169] Analytical assays - BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules Lyophilized samples of BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules were analyzed for product quality over a 13-week period. Samples were analyzed at the first time point (time zero) immediately after lyophilization. The remaining samples were then randomly sorted (depending on their position in the lyophilization tray) and stored at 4°C, 25°C, and 40% relative humidity (25°C / 40RH) and 40°C, and 65% relative humidity (40°C / 65RH). Samples were retrieved from storage conditions for analytical testing at specified time points (Tables 5 and 7).

[0170] [Table 7]

[0171] Results and Discussion Freeze-drying results: cycle time and observed maximum temperature All freeze-drying cycles were completed without issue. The chamber pressure (measured with a capacitance manometer and Pirani vacuum gauge), shelf temperature, and product temperature for each freeze-drying cycle are shown in Figure 2.

[0172] The non-CIN freeze-drying cycle (Figure 2A) had a maximum BCMAxCD3 bispecific molecular product temperature of -33.3°C and a maximum DLL3xCD3 bispecific molecular product temperature of -33.1°C during primary drying. The non-CIN freeze-drying cycle without annealing (Figure 2B) had a BCMAxCD3 bispecific molecular product temperature 2°C higher (-31.3°C) and a DLL3xCD3 bispecific molecular product temperature 1.6°C higher (-31.5°C) than the non-CIN freeze-drying cycle with the same shelf temperature setting during primary drying. This is consistent with the improved resistance to sublimation due to the elimination of annealing. However, the non-CIN freeze-drying cycle without annealing had similar primary drying endpoints compared to the non-CIN freeze-drying cycle with annealing. This indicates that annealing did not provide a time-saving advantage for these high-concentration bispecific molecules using the non-CIN freeze-drying cycle.

[0173] A CIN cycle with a primary drying shelf setting of -8°C (Figure 2C) showed a 50% reduction in primary drying time compared to a freeze-drying cycle. CIN completely eliminates stochastic freezing, forming larger, more homogeneous ice crystals and a more porous cake structure. 2 The primary drying time can be shortened to promote faster sublimation. However, the observed maximum product temperatures for BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules (-27.8°C and -28.7°C, respectively) were also higher than the previously reported critical temperature of -30°C for 1 mg / mL BCMAxCD3 bispecific molecules and 5 mg / mL DLL3xCD3 bispecific molecules when using lyophilization with CIN.

[0174] Table 6 lists the minimum primary drying times and minimum total cycle times for the freeze-drying cycles tested. The primary drying time was measured when the instantaneous gradient of the Pirani vacuum gauge approached zero, and is therefore considered to be the minimum primary drying time without the additional safety buffers included in a typical freeze-drying recipe. Thus, the total cycle time calculated using the minimum primary drying time is considered the minimum total cycle time. The freeze-drying non-CIN cycle with annealing had a longer minimum total cycle time than the standard non-CIN cycle without annealing. This is because, although the primary and secondary drying times were similar for the two cycles, the annealing process added 7 hours to the freezing time of the standard cycle.

[0175] The primary drying endpoint for the CIN cycle is 13 hours, which represents a 32% reduction in the minimum total cycle time. While this can vary based on optimizing the primary drying shelf temperature to lower the maximum product temperature, CIN can also reduce the primary drying time by forming larger, more homogeneous ice crystals, resulting in a more porous cake structure. In addition, the shorter freezing stage time makes CIN a superior choice in terms of total cycle time.

[0176] [Table 8]

[0177] HMWS product quality results Following lyophilization, the samples were reconstituted and analyzed for %HMWS by SE-UHPLC, with the results compared to the pre-lyophilized samples. The difference between the %HMWS of the reconstituted samples and the %HMWS of the pre-lyophilized samples was defined as Δ%HMWS. While all results were within the specification limit of ≤10.0%HMWS, the Δ%HMWS during lyophilization was largest in samples from standard cycles with annealing for both BCMAxCD3 and DLL3xCD3 bispecific molecules. Samples from standard cycles without annealing showed the least HMWS growth during lyophilization.

[0178] CIN freeze-dried samples showed slightly higher Δ%HMWS than samples from the standard cycle without annealing, but Δ%HMWS was approximately 80% lower than samples from the standard cycle with annealing for both BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules. Because CIN samples spend less time in a partially frozen matrix, the CIN freeze-drying cycle carries a lower risk of aggregation and HMWS formation than the annealing cycle.

[0179] After storing the lyophilized samples at 4°C, 25°C / 40RH, and 40°C / 65RH, the samples were removed at specific time points for %HMWS analysis. DLL3xCD3 bispecific molecular samples were measured at time zero, 2 weeks, and 4 weeks, while BCMAxCD3 bispecific molecular samples were measured at time zero and 13 weeks. Initially, BCMAxCD3 bispecific molecular samples were also planned to be measured at 2 and 4 weeks, but due to instrumental issues, they were instead measured at 13 weeks.

[0180] None of the freeze-dried groups showed an increase in %HMWS over time (Δ%HMWS was not used when comparing stability points). All BCMAxCD3 bispecific molecular samples (Figure 4A) and DLL3xCD3 bispecific molecular samples (Figure 4B) maintained their %HMWS over time. Freeze-drying cycles with annealing consistently showed higher %HMWS than freeze-drying cycles without annealing and CIN cycles.

[0181] Water content of BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules All moisture content 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. As shown in Figure 5, no significant difference in moisture content was found between samples from the three freeze-drying conditions (p>0.05).

[0182] In this embodiment, the same primary drying temperature setting as the CIN cycle for BCMAxCD3 bispecific molecule 1 mg 6R was used. This resulted in higher cake resistance with the high-concentration formulation, leading to a higher product temperature during primary drying of the high-concentration bispecific molecule than previously observed with the 1 mg / mL product. As described above, the product temperature of the CIN sample exceeded the critical temperature of 30°C previously characterized for CIN freeze-dried bispecific molecules. Previous studies using 1 mg / mL and 5 mg / mL BiTE standard products have shown that exceeding this temperature during primary drying leads to an increase in moisture content over time, and this was also observed in this embodiment. Optimizing the primary drying temperature of the CIN freeze-drying cycle may mitigate this problem.

[0183] Results of CEX testing for BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules The results of the CEX-HPLC assays for BCMAxCD3 and DLL3xCD3 bispecific molecules showed slight differences between samples from each lyophilization cycle. BCMAxCD3 bispecific molecule samples stored at 40°C / 65RH for 13 weeks showed a slight decrease in the main peak area and a corresponding slight increase in the acidic and basic peak areas, which was observed regardless of the lyophilization cycle (Figure 6A-C). Similar behavior was observed in DLL3xCD3 bispecific molecule samples after 4 weeks (Figure 6D-F). Therefore, no differences were observed between lyophilization cycles with annealing, lyophilization cycles without annealing, and the CIN cycle.

[0184] Results of quantitative determination of BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecule HIAC particles Particle quantification of BCMAxCD3 bispecific molecules (Table 7) and DLL3xCD3 bispecific molecules (Table 8) showed similar particle counts across samples from lyophilization cycles with and without annealing, and from samples from CIN cycles. All particle counts were well below the specification limits of 6000 particles / vial for particles 10 μm or larger and 600 particles / vial for particles 25 μm or larger.

[0185] [Table 9]

[0186] [Table 10]

[0187] pH and osmotic pressure results for BCMAxCD3 bispecific molecules and DLL3xCD3 bispecific molecules The pH and osmotic pressure of the samples reconstituted from the standard cycle with and without annealing, as well as the CIN cycle, were measured at time zero. Osmotic pressure and pH were tested to confirm the product formulation and reconstitution. No difference in pH or osmotic pressure was observed between the pre-freeze-dried sample and the reconstituted sample for both the BCMAxCD3 bispecific molecule (Table 9) and the DLL3xCD3 bispecific molecule (Table 10). Therefore, pH and osmotic pressure were not monitored at other time points during the stability test.

[0188] [Table 11]

[0189] [Table 12]

[0190] conclusion Controlled ice nucleation lyophilization was evaluated in use with high-concentration products using BCMAxCD3 bispecific molecules (20 mg / mL) and DLL3xCD3 bispecific molecules (15 mg / mL) in 1.3 mL presentations packed in 6R vials as model proteins. CIN lyophilization allowed for the avoidance of the high levels of HMWS commonly observed when using annealed lyophilization non-CIN cycles for these products. CIN lyophilized samples also exhibited less HMWS than annealed lyophilized samples in a 13-week product quality test, regardless of storage conditions.

[0191] CIN freeze-drying also resulted in a 32% reduction in minimum total cycle time compared to non-CIN freeze-drying cycles for bispecific molecules. Optimization of primary drying parameters can reduce the reduction in minimum total cycle time. However, CIN freeze-drying remains an attractive option compared to non-CIN freeze-drying cycles with annealing, primarily due to its advantages in reducing HMWS formation in high-concentration bispecific molecules and its potential benefit of shorter freeze-drying cycle times.

[0192] All references cited herein, including patents, patent applications, and publications, are incorporated herein by reference in their entirety.

[0193] While the present invention has been described with an emphasis on preferred embodiments, it will be apparent to those skilled in the art that variations of preferred compounds and methods may be used, and that the invention may be carried out in ways other than those specifically described herein. Accordingly, the present invention includes all modifications that are encompassed within the spirit and scope of the invention as defined by the following claims.

[0194] [Table 13]

[0195] [Table 14]

[0196] Table 15

[0197] Table 16

[0198] Table 17

[0199] Table 18

[0200] Table 19

[0201] Table 20

[0202] Table 21

[0203] Table 22

[0204] Table 23

[0205] Table 24

[0206] Table 25

[0207] Table 26

[0208] Table 27

[0209] Table 28

[0210] Table 29

[0211] Table 30

[0212] Table 31

[0213] Table 32

[0214] Table 33

[0215] Table 34

[0216] Table 35

[0217] Table 36

[0218] Table 37

[0219] Table 38

[0220] Table 39

[0221] Table 40

[0222] Table 41

[0223] Table 42

[0224] Table 43

[0225] Table 44

[0226] Table 45

[0227] Table 46

Claims

1. A method for preparing a freeze-dried bispecific molecular composition, (a) A step of inducing ice nucleation in a liquid bispecific molecular composition having a concentration of bispecific molecules of at least about 10 mg / ml in a vial that is exposed to a first temperature of about -18°C to about -10°C for a first period of about 60 minutes to about 270 minutes, (b) A step of exposing the vial from 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) A step of drying the composition from step (b) at a third temperature of about -5°C to about -25°C for a third period of about 25 hours to about 70 hours, (d) A step of further drying the composition from 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 obtain a vial containing a lyophilized bispecific molecular composition having a percentage content of high molecular weight species (HMWS) of 1.5% (m / V) or less. The bispecific molecule includes at least three domains, The first domain binds to tumor antigens on target cells. The second domain binds to the extracellular epitope of the human and / or maca CD3ε chain, and The third domain is fused to the second domain by a peptide linker, and the third domain comprises two polypeptide monomers, each containing a hinge, a CH2, and a CH3 domain, and these two polypeptide monomers are fused to each other via a peptide linker. method.

2. The method according to claim 1, wherein step (a) comprises exposing the vial to the first temperature for about 90 minutes to about 3 hours.

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

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

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

6. The method according to any one of claims 1 to 5, further comprising step (a) maintaining the ice nucleation protein composition at the first temperature for a period of up to 90 minutes after nucleation.

7. The method according to claim 6, wherein the period after nucleation is approximately 20 minutes to approximately 90 minutes.

8. The method according to claim 7, wherein the period after nucleation is approximately 30 minutes.

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

10. The method according to any one of claims 1 to 9, wherein the second period is approximately 2 hours to approximately 4 hours.

11. The method according to claim 10, wherein the second period is approximately 3 hours.

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

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

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

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

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

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

18. The method according to any one of claims 1 to 17, wherein the transition from step (b) to step (c) includes 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 1 hour.

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

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

21. The method according to any one of claims 1 to 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 according to any one of claims 1 to 18, which yields a vial containing a freeze-dried bispecific molecular composition that exhibits aggregation of about 1.5% (m / V) or less, or preferably about 1.1% (m / V) or less, with respect to high molecular weight species (HMSW) formation.

23. The third domain described above consists of amino and carboxyl molecules, Hinge - CH2 - CH3 - Linker - Hinge - CH2 - CH3 The method according to any one of claims 1 to 22, including

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

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

26. The first binding domain of the bispecific molecule is (a) CDR-H1 shown in Sequence ID No. 4, CDR-H2 shown in Sequence ID No. 5, CDR-H3 shown in Sequence ID No. 6, CDR-L1 shown in Sequence ID No. 1, CDR-L2 shown in Sequence ID No. 2, and CDR-L3 shown in Sequence ID No. 3 (b) CDR-H1 shown in Sequence ID No. 29, CDR-H2 shown in Sequence ID No. 30, CDR-H3 shown in Sequence ID No. 31, CDR-L1 shown in Sequence ID No. 34, CDR-L2 shown in Sequence ID No. 35, and CDR-L3 shown in Sequence ID No. 36 (c) CDR-H1 shown in Sequence ID No. 42, CDR-H2 shown in Sequence ID No. 43, CDR-H3 shown in Sequence ID No. 44, CDR-L1 shown in Sequence ID No. 45, CDR-L2 shown in Sequence ID No. 46, and CDR-L3 shown in Sequence ID No. 47 (d) CDR-H1 shown in Sequence ID 53, CDR-H2 shown in Sequence ID 54, CDR-H3 shown in Sequence ID 55, CDR-L1 shown in Sequence ID 56, CDR-L2 shown in Sequence ID 57, and CDR-L3 shown in Sequence ID 58, (e) CDR-H1 shown in Sequence ID 65, CDR-H2 shown in Sequence ID 66, CDR-H3 shown in Sequence ID 67, CDR-L1 shown in Sequence ID 68, CDR-L2 shown in Sequence ID 69, and CDR-L3 shown in Sequence ID 70, (f) CDR-H1 shown in Sequence ID 83, CDR-H2 shown in Sequence ID 84, CDR-H3 shown in Sequence ID 85, CDR-L1 shown in Sequence ID 86, CDR-L2 shown in Sequence ID 87, and CDR-L3 shown in Sequence ID 88, (g) CDR-H1 shown in SEQ ID NO: 94, CDR-H2 shown in SEQ ID NO: 95, CDR-H3 shown in SEQ ID NO: 96, CDR-L1 shown in SEQ ID NO: 97, CDR-L2 shown in SEQ ID NO: 98, and CDR-L3 shown in SEQ ID NO: 99 (h) CDR-H1 shown in Sequence ID 105, CDR-H2 shown in Sequence ID 106, CDR-H3 shown in Sequence ID 107, CDR-L1 shown in Sequence ID 109, CDR-L2 shown in Sequence ID 110, and CDR-L3 shown in Sequence ID 111, (i) CDR-H1 shown in Sequence ID 115, CDR-H2 shown in Sequence ID 116, CDR-H3 shown in Sequence ID 117, CDR-L1 shown in Sequence ID 118, CDR-L2 shown in Sequence ID 119, and CDR-L3 shown in Sequence ID 120, (j) CDR-H1 shown in Sequence ID 126, CDR-H2 shown in Sequence ID 127, CDR-H3 shown in Sequence ID 128, CDR-L1 shown in Sequence ID 129, CDR-L2 shown in Sequence ID 130, and CDR-L3 shown in Sequence ID 131, (k) CDR-H1 shown in Sequence ID 137, CDR-H2 shown in Sequence ID 138, CDR-H3 shown in Sequence ID 139, CDR-L1 shown in Sequence ID 140, CDR-L2 shown in Sequence ID 141, and CDR-L3 shown in Sequence ID 142, (l) CDR-H1 shown in Sequence ID 152, CDR-H2 shown in Sequence ID 153, CDR-H3 shown in Sequence ID 154, CDR-L1 shown in Sequence ID 155, CDR-L2 shown in Sequence ID 156, and CDR-L3 shown in Sequence ID 157, (m) CDR-H1 shown in Sequence ID 167, CDR-H2 shown in Sequence ID 168, CDR-H3 shown in Sequence ID 169, CDR-L1 shown in Sequence ID 170, CDR-L2 shown in Sequence ID 171, and CDR-L3 shown in Sequence ID 172, (n) CDR-H1 shown in Sequence ID No. 203, CDR-H2 shown in Sequence ID No. 204, CDR-H3 shown in Sequence ID No. 205, CDR-L1 shown in Sequence ID No. 206, CDR-L2 shown in Sequence ID No. 207, and CDR-L3 shown in Sequence ID No. 208 (o) CDR-H1 shown in Sequence ID No. 214, CDR-H2 shown in Sequence ID No. 215, CDR-H3 shown in Sequence ID No. 216, CDR-L1 shown in Sequence ID No. 217, CDR-L2 shown in Sequence ID No. 218, and CDR-L3 shown in Sequence ID No. 219 (p) CDR-H1 shown in Sequence ID No. 226, CDR-H2 shown in Sequence ID No. 227, CDR-H3 shown in Sequence ID No. 228, CDR-L1 shown in Sequence ID No. 229, CDR-L2 shown in Sequence ID No. 230 and CDR-L3 shown in Sequence ID No. 231, and (q) CDR-H1 shown in Sequence ID No. 238, CDR-H2 shown in Sequence ID No. 239, CDR-H3 shown in Sequence ID No. 240, CDR-L1 shown in Sequence ID No. 241, CDR-L2 shown in Sequence ID No. 242, and CDR-L3 shown in Sequence ID No. 243 The method according to any one of claims 1 to 25, comprising a VH region including CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of the above, and a VL region including CDR-L1, CDR-L2, and CDR-L3.

27. The first domain mentioned above is sequence numbers 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, 131, 132, 133, 13 The method according to any one of claims 1 to 26, having an amino acid sequence selected from the group consisting of 4, 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, preferably 100 to 104.

28. The method according to any one of claims 1 to 27, wherein ice nucleation is induced via ice fog or decompression.

29. A freeze-dried bispecific molecular composition prepared by the method described in any one of claims 1 to 28.

Images