From R&D to Clinical Supplies - Organic Process Research


From R&D to Clinical Supplies - Organic Process Research...

6 downloads 99 Views 2MB Size

Subscriber access provided by University of Newcastle, Australia

Full Paper

From R&D to Clinical Supplies Amar S Prashad, Birte Nolting, Vimalkumar Patel, April Xu, Bo Arve, and Leo Letendre Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.7b00020 • Publication Date (Web): 28 Mar 2017 Downloaded from http://pubs.acs.org on March 28, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Organic Process Research & Development is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

From R&D to Clinical Supplies Amar S Prashad1 *, Birte Nolting1, Vimalkumar Patel1, April Xu1, Bo Arve2, Leo Letendre1 * 1

Pfizer, Inc., Biotherapeutics Pharmaceutical Sciences, Worldwide R&D, Pearl River, NY

10965, USA 2

Pfizer, Inc., Biotherapeutics Pharmaceutical Sciences, Worldwide R&D, Chesterfield, MO

63017, USA

ACS Paragon Plus Environment

1

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

ABSTRACT

In the development of new Antibody Drug Conjugates (ADCs), the activities performed by discovery groups typically focus on rapid and comprehensive screening of many conjugates to find the ones with the desired efficacy and safety profiles. These conjugates are typically prepared in a combinatorial approach whereby various monoclonal antibodies (mAbs) for a specific target, linkers and payloads are combined. These efforts usually rely on efficient screening methodologies and high-throughput tools, such as solid phase conjugation and purification arrays. Development of robust and consistent processes suitable to produce the selected candidate for clinical trials is typically not a priority for discovery. Many of the ADCs in the clinic today are based on one of two conjugation technologies and associated linkerpayloads (LP). The traditional cysteine conjugation technology utilizes an auristatin as payload with a cleavable or non-cleavable linker. The other technology is based on conjugating a maytansinoid or calicheamicin based payload to lysine residues on the mAb. Selecting a single conjugation chemistry and one linker-payload for all ADCs, where only the mAb is changing based on the target, allows the building of platform processes and methods. This leads to efficient process and analytical methods development and a reduction in the work required to develop processes suitable for the production of GMP clinical trial material. However, a growing number of ADCs are being developed to further reduce toxicity and improve efficacy utilizing novel linkers and payloads as well as new conjugation technologies. These payloads include DNA damaging cytotoxins such as DNA alkylating or crosslinking agents and cytotoxic payloads derived from natural products with novel mechanism of action. New conjugation chemistries include site-specific technologies based on non-native amino acids, inserting unpaired cysteine residues and enzymatically mediated conjugation methods. As a result,

ACS Paragon Plus Environment

2

Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

developing processes for each new conjugation technology and combination of new linker and payload may require significant resources for process and methods development as well as scaleup. Therefore, aligning discovery and development (analytical and process development) efforts for clinical trial material production becomes important. To facilitate the technology transfer from Discovery to Process Development, a systematic approach of early engagement and evaluation of the mAb, linker-payload and ADC should be established. This includes harmonization of analytical methods and processes used for production of ADCs for pre-clinical screening experiments, including exploratory toxicology studies. This paper will review methodologies for technology transfer from Discovery to Development, approaches to process and methods development for a diverse portfolio of ADC technologies, as well as process scaleup and specification setting. Case studies for different conjugation chemistries and linkerpayloads will be reviewed, including conventional cysteine, lysine, and site-specific technologies.

KEYWORDS: antibody drug conjugate; conventional cysteine chemistry; lysine chemistry; site specific conjugation; transglutaminase; quality attributes; process development

ACS Paragon Plus Environment

3

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 34

INTRODUCTION Antibody drug conjugates (ADCs) are a class of biopharmaceutical drugs composed of an antibody that is chemically linked to one or more cytotoxic small molecules.1 While other uses have been explored, these conjugates have been designed as anti-cancer agents that combine the unique selectivity of antibodies as targeted therapy with very potent small molecule compounds that inhibit the growth of or kill cancer cells.2 The discovery efforts directed towards finding new ADCs as oncology therapies involves the development of an appropriate antibody, the selection of an appropriate toxin and linker chemistry as well as selecting the number of toxin molecules on the antibody in order to maximize efficiency and minimize side effects.3 There are several classes of cytotoxic molecules that are employed or being developed for use in ADC based therapies.4 Some of these are shown in Figure 1. At this time, the tubulin inhibitors and the DNA damaging classes contain a number of compounds from which the discovery efforts can choose.

Figure 1. Different Classes of Cytotoxic Drugs. As is typical of any discovery effort, the discovery of a new ADC involves the synthesis of a large number of potential candidates to be tested in a battery of in vitro and in vivo tests. Given the need to produce a large numbers of candidates, it is not surprising that the techniques used

ACS Paragon Plus Environment

4

Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

are not necessarily those that are easily scalable, thus providing one of the challenges to be addressed in bringing the selected ADC to the clinic.

Antibody drug conjugates are inherently heterogeneous materials.5 The heterogeneity comes from several sources. The antibody is inherently heterogeneous arising mainly from variations in glycosylation. The heterogeneity may then increase further due to the manner in which the toxin is attached to the antibody. One of the major challenges of progressing an ADC from discovery through early development and eventually through approval and commercialization is qualitatively and quantitatively maintaining the same level of heterogeneity. Pfizer’s current portfolio of ADCs in clinical trials include molecules that use lysines, cysteines or glutamines as the conjugation points on the antibody. Examples of each of these chemistries will be presented. Conjugation to lysine residues in an antibody generally introduces the greatest increase in heterogeneity due to the fact that there are a large number of lysine residues available to react. In the chemistries used to date by Pfizer, the relative proportion of linker/payloads attached to a given lysine residue has been kinetically controlled with four or fewer residues binding the preponderance of the linker/payload.4b Conjugation to cysteine residues in an antibody may be utilized in two different ways to produce an ADC. One chemistry utilizes the endogenous cysteine residues by reduction of the interchain disulfide bridges.3 chemistry.

This method introduces less heterogeneity than the lysine based

However, heterogeneity is nonetheless introduced since methods to selectively

reduce a specific set of disulfides (heavy-light chain or heavy-heavy chain disulfides) have not yet been developed. Thus, a kinetically controlled mixture of free sulfhydryls is produced leading to a mixture of sites where the linker/payload is attached. More recently, unpaired

ACS Paragon Plus Environment

5

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

cysteines have been introduced into the protein sequence which allows for site-specific introduction of the linker/payload.6 One current drawback to the practice of this chemistry stems from the fact that these unpaired cysteine residues typically are in the form of a disulfide with molecules captured from the cell culture process media such as free cysteine or glutathione. In order to free the unpaired cysteine residues, all of the accessible disulfides are reduced, small sulfides removed and the antibody re-oxidized. This process introduces yet another form of heterogeneity based upon potential non-natural pairing of cysteine residues in rebuilding the disulfide bridges. In this case, the addition of the linker/payload is selective but the antibody may be more heterogeneous. The third conjugation method employed by Pfizer is an enzymatic coupling utilizing transglutaminase.7 To facilitate this coupling the antibody is engineered to include a glutamate containing sequence with a requisite recognition sequence to react with an amine present on the linker. This method of conjugation increases the heterogeneity only to the extent of potentially competing deamination of the mAb versus the transamidation with the cytotoxic payload (leading to unconjugated or partially conjugated ADC). . With all of these conjugation chemistries, the option to adjust the level of heterogeneity may be available through the use of chromatography although not all chemistries lend themselves equally to the task.

TECHNOLOGY TRANSFER FROM DISCOVERY The transfer of knowledge from Discovery when the selected ADC is transferred to Development is a critical function. The level of detailed understanding of the selected molecule required at the technology transfer event is higher than that typically required for small

ACS Paragon Plus Environment

6

Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

molecules mainly due to the level of heterogeneity of the molecule. Since the selection of a specific ADC was based upon the properties and performance of a specific heterogeneous mixture, the development effort must effectively reproduce and maintain this heterogeneity throughout development and ultimately through commercialization. Technology transfer (TT) from Discovery formally starts when a lead candidate is selected. At this point the optimum mAb, the preferred linker-payload combination and conjugation methodology are selected based on pre-clinical toxicology, efficacy and pharmacokinetic data.8 We have formalized the transfer process from Discovery to include a TT checklist which was developed jointly between Discovery and Development. Over time, Pfizer’s discovery and development scientists have established a formal checklist of items to be included in the discussions during technology transfer.

Much of the checklist details the biochemical and

chemical characterization of the ADC but also includes details of the conjugation and purification techniques used to produce material for pre-clinical studies.

As part of the

technology transfer, a series of bioanalytical assessments of both the antibody and the ADC are performed to determine potential issues with the stability or with the production of the antibody. As part of the evolution of this check list, process and analytical knowledge was shared between Discovery and Development and where appropriate, the processes and methods were harmonized. This document captures all relevant information about the ADC, its characteristics, and the process and methods used by Discovery to generate material for early screening studies. This includes information about process parameters such as reducing agent (cysteine chemistry), linker-payload stoichiometries, antibody concentration during conjugation, temperature of reduction/conjugation, pH, etc. In addition to the process parameters, discovery describes target quality attributes such as the drug antibody ratio, aggregate levels, unconjugated antibody levels

ACS Paragon Plus Environment

7

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34

and residual drug levels.9 These targets are also informed by current trends in regulatory guidance. To ensure targets are understood, communication between Development and Discovery is maintained throughout the development process.

Other work processes instituted to facilitate the TT and development of new ADCs are a “road map” of all activities and deliverables as well as a detailed molecular assessment of the mAb as part of the routine selection process. The assessment reviews mAb attributes such as aggregation, hot spots, viscosity, and general manufacturability and when possible will influence the selection of the final construct. A high level view of a typical road map is shown in Figure 2

Figure 2. Typical “Road Map” of Activities and Deliverables for ADC Development (TS – Team Supply; CTM – Clinical Trial Material; MCB – Master Cell Bank)

In cases where novel technologies (conjugation modality, novel linker-payload, etc.) are utilized by Discovery, early collaboration between Discovery and Development is established prior to a project entering Development. As an example, before a candidate utilizing a novel site specific enzyme mediated conjugation technology formally entered Development, a collaboration between the Discovery and Development groups was established, which included early exchange

ACS Paragon Plus Environment

8

Page 9 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

of materials and data on various model ADC constructs. Such close collaboration prior to candidate selection helps to drive the underlying science and adds to the understanding of the conjugate providing insight into potential process and analytical challenges. Based upon the outcome of the collaboration, Development is afforded the opportunity to start exploratory development work to minimize potential delays once the final candidate has been selected.

PROCESS AND METHODS DEVELOPMENT Discovery typically provides an initial quantity of antibody and linker-payload to Development to allow Development to reproduce the Discovery process and initiate analytical method development and pharmaceutics screening. Once technology transfer from Discovery to Development is complete, the development work is initiated according to a project specific work plan based on a template road map. The project plan includes the following major activities:

Antibody o Transfection, cell line development and cell banking o Cell culture and purification process development and scale-up o “Development Team Supplies” of antibody for ADC, analytical method and formulation development o Stability studies o Regulatory toxicology and clinical supplies – drug substance intermediate (DSI) Linker-payload o Route development o LP analytical methods development o Stability Studies

ACS Paragon Plus Environment

9

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

o Regulatory toxicology and clinical supplies – drug substance intermediate (DSI) ADC o Conjugation and purification process development o “Development Team Supplies” of ADC for analytical method and formulation development o ADC analytical methods and formulation development o Stability studies o Regulatory toxicology and clinical supplies – drug substance and drug product (DS and DP) Pack, Label and Ship IND Regulatory Submission

The process developed by Discovery usually forms the basis for the manufacturing method eventually used by Development to produce clinical material. The initial Development work is focused on understanding any potential issues which might be encountered during scale-up and which could lead to generating an ADC with an unacceptable quality attribute profile. These studies evaluate reaction concentrations, input ratios, reaction temperature, addition rates and other parameters. In addition, if chromatographic purification is necessary, resin screening and process optimization in terms of binding capacity, selectivity and impurity removal is performed. Also, for the ultrafiltration/diafiltration step, studies for optimum performance in terms of protein to membrane area, (potentially) free drug and co-solvent removal and flux are carried out. Once the basic process steps and operating parameters have been determined additional process and product understanding is gained by performing experiments in multivariate formats such as

ACS Paragon Plus Environment

10

Page 11 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Design of Experiments (DOE) and also One/Multiple Factor at a Time (O/MFAT) to better understand process parameters, determine potential parameter interactions and their impact on process performance and product quality attributes. Based on this data the process is used by Development to generate regulatory toxicological material (non-clinical) and transferred to a manufacturing organization to produce early phase clinical trial material.

As part of streamlining the development and scale-up of the ADC process appropriate equipment design and process engineering methodologies have been established to enable delivery of preclinical ADC supplies and scalable manufacturing processes for clinical supplies. Therefore, a library of conjugation reactors that have been characterized in terms of blend times and computational fluid dynamics (CFD) models has been compiled. This effort is further aided by utilization of high throughput screening for chromatography resins and implementing scalable ultrafiltration and diafiltration protocols.

Due to the large diversity of early stage ADCs, it is difficult to have one set of platform methods for all ADCs, including in-process methods to support process development. In addition, the large heterogeneity associated with conjugation chemistry presents particular challenges in analytical characterization and method development. As a result, multiple orthogonal analytical technologies are required to provide detailed characterization including cutting edge liquid chromatography and mass spectrometry. For example, to characterize the non-covalent linked different DAR species in cysteine chemistry (or lysine chemistry ADC with acid labile linker), native mass spectrometry will be required.

Peptide mapping will often be required to

identify/confirm sites of conjugation. However, not all the heightened characterization may be

ACS Paragon Plus Environment

11

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34

needed at early stage of development. To enable rapid process development, an initial set of analytical methods need to be identified to measure the key quality attributes of the ADCs, including drug to antibody ratio (DAR), drug loading distribution, residual drug and related impurities, size distribution (including high and low molecular mass species), and charge distribution, if needed. Methods from existing similar ADC constructs can be used as a starting point for the development. Often a rapid method verification, minor parameter adjustments and minimum method characterization can suffice for early process development support.

In cases where the method is not fit for purpose, more efforts are needed to develop a suitable method. Additional method characterization may be required to ensure a good understanding of method capability. One particular challenge for in-process methods is the potential matrix interference as sample matrices may vary significantly during early process development. Many methods, such as iCE (Imaged Capillary Electrophoresis), HIC (Hydrophobic Interaction Chromatography), SEC (Size Exclusion Chromatography), IEX (Ion Exchange Chromatography) can be sensitive to matrix interference. In particular, the methods for iCE and free drug assay (typically reversed phase chromatography) can be highly dependent on buffer salt concentrations and other excipients. Either an additional sample preparation step or a spike recovery study is required to confirm method suitability in conjunction with in process sample testing. It is worth mentioning that particular effort should be spent on developing a suitable assay for monitoring/determining free drug and related impurities, as it obviously is a critical quality attribute which could impact off-site toxicity for ADCs. It is typically challenging to develop a method for free drug and related species in the presence of ADC which often requires high sensitivity and specificity. While multiple methodologies have been used in the area of ADC

ACS Paragon Plus Environment

12

Page 13 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

free drug method development, our experience indicates that a suitable LC method with UV detection may be optimum in supporting process development and characterization as well as stability studies.

To facilitate rapid conjugation process development, a tailored testing strategy is important to provide rapid feedback so that requisite process improvements can be implemented. For example, to optimize the process for residual impurity clearance, only the residual impurity testing will be performed and not the entire battery of methods for quality attributes of the ADC. While bioassays (such as antigen binding and cytotoxicity assays) are typically not needed for inprocess support, they are performed for each DS team supply to ensure the quality of ADC is within expected ranges. Although the in-process methods may not be the same as those used for final drug substance release, they typically serve as an appropriate starting point for further method optimization. The method optimization is typically concurrent with the in-process testing process development support so that minimum bridging is needed between measurements of quality attributes of the ADC at different stages of development. For specification setting for early clinical stage ADCs, knowledge on key attributes is gained from across different ADC molecules. Wider ranges are typically used in early stage compared to late clinical stage since process and product knowledge is limited during early stage. Platform knowledge built from other ADCs can support the setting of specification strategies.

ACS Paragon Plus Environment

13

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34

CASE STUDIES Conventional Cysteine Conjugation This technology is based on the well-established cysteine conjugation chemistry developed by Seattle Genetics.5 Inherent to this technology, the partial reduction of IgG1 mAbs yields eight possible conjugation sites and a drug-loading distribution ranging from 0, 2, 4, 6 to potentially 8 drug molecules per mAb molecule, some of which may also be present in different isoforms. Based on current ADCs in clinical trials, an average loading of four drug molecules per mAb molecule is targeted, minimizing unconjugated antibody and higher-loaded species.10 Minimal amounts of unconjugated antibody is desired because this species doesn’t possess any cytotoxic moiety. While minimal amount of higher-loaded species; e.g., DAR 8, is desired because these species tend to exhibit rapid plasma clearance. A schematic diagram for potential drug loaded species and their isoforms are shown in Figure 3. Scheme with Different Drug Loaded Species and Isoforms from Conventional Cysteine Conjugation.

Figure 3. Scheme with Different Drug Loaded Species and Isoforms from Conventional Cysteine Conjugation.

Three HPLC assays (HIC, RP, SEC) are typically used for in-process and drug substance testing to support process development by quantifying the key quality attributes of an ADC at various

ACS Paragon Plus Environment

14

Page 15 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

steps in the process. Analytical hydrophobic interaction chromatography (HIC) is one of the most commonly used methods for ADC in-process testing, as it can provide information about several attributes in a single assay.11 Examples of HIC chromatograms for an auristatin ADC are shown in Figure 4. Analytical HIC Chromatograms.

A.

Discovery Process

B.

Optimized Development Process

Figure 4. Analytical HIC Chromatograms. A. Discovery Process B. Optimized Development and Manufacturing Process

The loading distribution (0, 2, 4, 6 and 8 loaded) as well as the two positional isoforms of 4loaded species (4A and 4B) are shown in Figure 7. Attributes such as DAR (drug antibody ratio), drug loading distribution, and unconjugated mAb content can be determined by HIC analysis alone. Residual drug levels can be quantified using reversed phase RP-HPLC, where drug related species are eluted separately from the ADC. Size exclusion chromatography (SEC) is used to measure the high molecular mass species (HMMS) and potential low molecular mass species. Additional testing includes mAb concentration, TCEP stock concentration as well as free drug clearance in various process steps.

The conjugation parameters transferred from Discovery to Development were suitable to produce an ADC with the targeted average drug loading of 4 (DAR 4). During the conjugation

ACS Paragon Plus Environment

15

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

process development, a more homogeneous drug loading distribution was obtained (Figure 4B) while maintaining the overall average target loading of 4 Figure 5).

Conditions A - Optimized Development Conditions B - Early Development Conditions C - Discovery

Figure 5. Comparing Average Drug loading from Discovery Process to optimized Development Process.

In particular, unconjugated mAb and higher-loaded ADC isomers were being reduced. This was achieved by the screening of the various process parameters in DOEs and M/OFAT experiments for better process understanding and the factors contributing to the drug-loading distribution (shown in Figure 6, 2-loaded ADC not included). This was accompanied, by a change of the ratio between the most prevalent 4-loaded isomers 4A and 4B, as was evident in the analytical HIC profiles (shown in Figure 4. Analytical HIC Chromatograms. and diagram in Figure 7).

ACS Paragon Plus Environment

16

Page 17 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Conditions A - Optimized Development Conditions B - Early Development Conditions C - Discovery

Figure 6. Comparing Drug Loading Distribution from Discovery Process to optimized Development Process.

Development Discovery

Figure 7. Loading Distribution

The challenge for further development of this ADC molecule with less heterogeneity, and what appears to be an improved drug loading distribution, was to demonstrate comparability to the Discovery ADC used in candidate-enabling exploratory toxicology and efficacy studies. This may be of importance as each of the different loading isomers may be associated with different toxicological and pharmacokinetic properties. Comparability was addressed by several approaches.

ACS Paragon Plus Environment

17

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Page 18 of 34

Binding and cytotoxicity assays showed comparable results for ADC with either loading distribution.



The different mAb batches used by Discovery and Development as well as Discovery and Development processes were bridged by generating a number of ADCs with the expected loading profiles and an average loading of 4 drugs per mAb. o These ADCs were subjected to further biophysical analyses such as thermal stability and forced aggregation demonstrating comparability between the Discovery and Development ADC material. o Additional regulatory toxicology studies, using ADCs with the different loading profiles and bridging mAb sources also demonstrated comparability between the broader and narrower loading distribution ADCs.

Therefore, the manufacture of clinical materials followed the process producing the more homogeneous drug distribution with lower content of unconjugated mAb and higher loaded species while maintaining the target average loading of 4 drugs per mAb. A comparison of the loading distribution of several pre-clinical (across different scales) and clinical batches in shown in (Figure 8).

ACS Paragon Plus Environment

18

Page 19 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Figure 8. Loading Distribution for Preclinical and Clinical Batches.

Lysine Conjugation This conjugation technology involves conventional lysine chemistry whereby the lysine residues of the mAb are conjugated to a linker-payload, in this case a calicheamicin-based linkerpayload.12 This is shown in Figure 9.

O O

NH

O Me Me NHN CH3 O I O

CH3 HO

O OCH3 OH

S

Me CH 3

O

HO

S

NH O

S O

OCH3

O

CH3

HN HO

OCH3 OH OCH3

Et N

CH3

O O

H

O

O OCH3

O

0-9

Figure 9. ADC with a calicheamicin-based linker-payload.

ACS Paragon Plus Environment

19

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

When this ADC was transferred from Discovery to Development, two major process challenges had to be addressed. The first was the aggregation level during the conjugation reaction and subsequent increase over time in the crude ADC. The second was a low process yield, approximately 25%. Since this process requires a chromatography step, reducing the aggregate level during the conjugation reaction and improving the step yield of the chromatography step will both improve the overall process yield.

Since this is an ADC produced using lysine chemistry with an acid labile linker, there are unique challenges associated with the analytical support for ADC process development. One is the potential drug loading heterogeneity in the ADC due to the presence of over 80 lysine residues in an IgG mAb. The other is the narrow pH range suitable for analytical methods due to the acid labile nature of the linker. Based on the special concerns regarding the aggregation level during conjugation, the validity of the SEC method needed to be established concurrently with early process development to ensure accurate quantification. Challenges were also encountered in the development of appropriate methods for other key quality attributes such as DAR, drug loading distribution, unconjugated mAb and residual free drug. For example, due to the difficulty in separating different drug loaded ADC species using conventional HPLC separation techniques such as HIC, drug loading distribution was monitored using an imaged capillary electrophoresis method (iCE), while a HIC method was used to quantify unconjugated mAb levels. A typical HIC chromatogram is shown in Figure 10.

ACS Paragon Plus Environment

20

Page 21 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Figure 10. Typical HIC Profile for ADC with Spiked mAb.

Due to the acid-labile nature of the linker, a non-TFA based RP-HPLC method was developed to measure residual calicheamicin and its degradants in the ADC. Average drug loading (DAR) was determined by a UV method leveraging knowledge from similar ADCs.

Additional

information on sites of conjugation and site occupancies was obtained using peptide mapping.

In an effort to better understand the underlying factors that may be responsible for the level of aggregation observed during the conjugation reaction and subsequent hold time for this ADC, reaction kinetics were determined, and a DOE evaluating several conjugation reaction parameters was performed. These included concentration of buffer, antibody, solubilizing agent, co-solvent, linker-payload and pH. Some results of the DOE study are shown in Figure 11.

ACS Paragon Plus Environment

21

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 34

Figure 11. Example of Results from DOE Studies. As evident from this data, certain process parameters have to be carefully controlled in order to minimize the level of unconjugated antibody in the ADC while also maintaining an acceptably low level of aggregation in the crude reaction mixture. Therefore, selected reaction conditions were screened in the preparation of ADC (Figure 12).

Effect of Parameter 1 on Agg & UmAb Levels

Effect of Parameter 5 on Agg & UmAb Levels 20 18 16 14 12 % 10 8 6 4 2 0 32.5

14 12 10 %8 6 4 2

Effect of Parameter 6 on Agg & UmAb Levels 16 14 12 10 % 8 6 4 2 0

37.5

42.5

Agg (%) UmAb (%)

4

6

8

10

0 225

275

325

375

Figure 12. Effect of Selected Parameters on Aggregate and Unconjugated mAb Levels.

Furthermore, several additives were explored in an effort to reduce and stabilize the level of aggregation observed in the crude ADC reaction mixture. These additives were added upon completion of the reaction. The effect of the evaluated additives on the aggregation levels is

ACS Paragon Plus Environment

22

Page 23 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

shown in Figure 13. Based on this data Agent 3 was used to stabilize the level of aggregation in the crude reaction mixture.

18

16

14

12

10 % Agg 8

6

No Additive Agent 1 Agent 2 Agent 3

4

2

0 0

1

2

3

4

5

6

7

Time (h)

Figure 13. Effect of Evaluated Additives on Aggregate Levels. To address the low yield in the purification step, several chromatography resins were evaluated for their efficiency to remove unconjugated antibody, aggregates and residual linker-payload. The results for some resins are summarized in Figure 14.

80 % Aggregates

% Unconjugated mAb

% Residual Drug

Yield

70 60 50 40 30 20 10 0 Resin 1

Resin2

Resin 3

Resin 4

Figure 14. Results from Resin Screening.

ACS Paragon Plus Environment

23

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

The results show that Resin1, a HIC resin, performed well in removing the undesired species, however, the yield was unacceptably low. Resin 2, a less hydrophobic HIC resin, had improved yield, however, did not perform well in removal of aggregates. Resin 3, a mixed mode resin, performed well in all areas except removal of residual drug. Resin 4, another HIC resin, performed well in all areas, including impurity removal and yield, and was chosen as the resin to for chromatography purification of this ADC.

The combination of optimized reaction conditions, additive addition to control aggregation levels during the hold of the crude ADC and an optimal chromatography purification step resulted in an ADC process enabling the manufacture of clinical material with the targeted quality attributes. A comparison of four pre-clinical and clinical batches produced by this process at 5 – 100 g scale is shown below. % Agg

UmAb

Upper limit

Upper limit

%

1

2

3

1

4

2

3

4

Batch of ADC DS

Batch of ADC DS DAR

Target

1

2

3

4

Batch of ADC DS

ACS Paragon Plus Environment

24

Page 25 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Figure 15. Comparison of Results for Selected Quality Attributes for Pre-Clinical and Clinical ADC Batches. Site Specific Conjugation Enzyme mediated conjugation is one of the emerging site-specific technologies. In this case bacterial transglutaminase (TG) is used to mediate an enzymatic trans-amidation leading to covalent linking of engineered glutamine residues (Gln) at specific sites on the mAb with amino groups on a drug derivative.13 Since the enzyme binding is limited to specific “glutamine tags” or engineered glutamines in the mAb, naturally occurring glutamines are not recognized by the enzyme. This allows sites on the mAb to be selected that provide the best properties for the ADC, including a homogeneous drug loading, favorable PK, and optimum therapeutic index (TI).

For the construct in this case study, two glutamine residues were genetically incorporated as part of short peptide tags (LLQGA) into the C-termini of the heavy chains of the mAb, thereby leading to a target loading of two drugs per mAb (DAR2). For determination of average loading and loading distribution an analytical HIC method was used. A typical chromatogram for the crude ADC reaction mixture is shown in Figure 16.

DAR 2

ACS Paragon Plus Environment

25

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34

Figure 16. Typical HIC Profile for Crude ADC Reaction Mixture Based on TG-Mediated Conjugation. The HIC profile indicates a significant amount of 1-loaded species (DAR 1) due to varying levels of truncation of the glutamine tag, which were encountered during the mAb production switch from a transient expression system to a stable cell line. Under mAb platform conditions, over 90% of the tag was truncated due to proteolysis. Through a series of experiments, specific conditions were developed to significantly reduce the level of tag-truncation of the mAb. This was crucial for the development of this ADC since an intact tag is required as substrate for transglutaminase recognition and thereby for conjugation (Figure 17).

Figure 17. Intact Tag is Required for TG-Mediated Conjugation.

Increasing levels of truncation of the tag lead to decreasing levels of the expected 2-loaded species and increasing levels of 1-loaded species and unconjugated mAb (see Figure 18).

ACS Paragon Plus Environment

26

Page 27 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Figure 18. Increased Tag-Truncation Leads to Lower ADC DAR2 Content.

The determination of the level of truncation in the mAb is challenging as no conventional HPLC method is capable of distinguishing intact mAb from mAb with a truncated tag. In addition to extensive screening using the TG-mediated conjugation itself to optimize mAb production conditions, analytical methods were employed to support mAb process development. The Two methods used were intact mass analysis by mass spectrometry and a focused peptide mapping approach by quantifying relevant peptides. The latter assay was used for the final mAb characterization.

Considering the site-specificity of this technology, mostly homogeneous drug-loading at the engineered mAb sites is required in successful ADC manufacture. During the development of this ADC several challenges had to be addressed and overcome. In the Discovery process for this ADC construct, a very large molar excess of drug to mAb was used to achieve 2-loaded ADC and minimize 1-loaded species. This excess was about 10-fold higher than used in conventional cysteine chemistry. Systematic screening of a variety of

ACS Paragon Plus Environment

27

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 34

reaction parameters such as concentration of mAb, buffer composition, pH and temperature in addition to drug stoichiometry and TG input in DOEs and M/OFAT experiments provided a better process understanding. Adjustment of the conjugation reaction conditions yielded consistently high levels of 2-loaded species while requiring only about one tenth of the required molar drug equivalents compared to the earlier process. Results for selected screening experiments are shown in Figure 19.

Figure 19. Results from Selected Conjugation Parameter Screening.

Furthermore, the input of TG could be significantly reduced compared to the Discovery process. This is advantageous for another challenge specific to the transglutaminase mediated conjugation technology: the necessity to remove the enzyme from the ADC after conjugation. A sensitive and selective analytical method is needed to determine residual transglutaminase levels to demonstrate removal in the ADC manufacturing process. The determination of residual transglutaminase levels is challenging due to two reasons: 1) a very low limit of quantitation (LOQ) is required to ensure ADC quality, 2) the difficulty to resolve transglutaminase from

ACS Paragon Plus Environment

28

Page 29 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

ADC using common chromatographic methods especially the quantitation of minute amounts in the presence of large amounts of ADC. Two methods were developed: an ELISA assay and a LC/MS method. Both methods were used for process development support.

In the Discovery process, an affinity chromatography approach using Protein A was used for purification of the ADC. While this allowed the removal of TG and excess drug, it is not suitable for manufacture of clinical material unless followed by an additional purification step to remove residual Protein A. Furthermore, unconjugated mAb and 1-loaded species (present due to the tagtruncation) need to be reduced to maintain a homogeneous 2-loaded ADC. During process development, several resins (mostly HIC and mixed mode) were screened with the aim of finding a resin that could achieve removal of all impurities including transglutaminase, free drug, unconjugated mAb and 1-loaded ADC species. Results from initial screening experiments are shown in Table 1.

Table 1. Results from Resin Screen. Resin

TG Removal

LP Removal

(varying Residual TG)

mAb/DAR1/DAR2 Separation

A

yes

yes

no

B

yes

no

no

C

yes

no

no

D

yes

yes

yes

E

yes

yes

yes

Resin D and E appeared to be the most promising and were further evaluated for loading capacity, impurity removal and DAR 2 recovery with or without mobile phase modifiers. Resin

ACS Paragon Plus Environment

29

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 34

E was most suitable based on removal of transglutaminase and free drug and provided good recovery for the target ADC (2-loaded species) (Figure 20) and purified ADC met the acceptance criteria for all critical quality attributes.

Figure 20. Resin Challenge and Recovery with and without Modifier.

A comparison of several pre-clinical and clinical batches shows the site-specific conjugation resulting in an ADC with a homogeneous DAR of 2 (Figure 21).

ACS Paragon Plus Environment

30

Page 31 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Figure 21. Comparison of Pre-Clinical and Clinical ADC Batches.

SUMMARY Having processes, methods and systems in place to transition projects from Discovery to Development and then Manufacturing is key to the successful and expedient delivery of clinical trial material for early phase studies. Notwithstanding, as exemplified by this paper, different conjugation chemistries present unique challenges that have to be addressed in the process and methods development of an ADC and innovative solutions to challenges will always be an important strategy for advancing a diverse portfolio of ADCs to the clinic.

AUTHOR INFORMATION

Corresponding Author *E-mail: [email protected]; [email protected]

ACS Paragon Plus Environment

31

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 34

ACKNOWLEDGEMENTS We acknowledge the following individuals who played important roles on some of these projects described herein. Process Development colleagues:

He Meng, Ryan Manetta, & Bhumit Patel

Discovery colleagues:

Frank Loganzo, Kiran Khandke, Manoj Charati & Nadira Prashad

REFERENCES

1. Gerber, H.-P.; Dijoseph, J. F.; Khandke, K. M.; Marquette, K. A.; Sapra, P.; Tchistiakova, L. G. Antibody-drug conjugates. WO2012131527A1, 2012. 2. (a) Carter, P. J.; Senter, P. D., Antibody-drug conjugates for cancer therapy. Cancer J 2008, 14 (3), 154-69; (b) Drachman, J. G.; Senter, P. D., Antibody-drug conjugates: the chemistry behind empowering antibodies to fight cancer. Hematology Am Soc Hematol Educ Program 2013, 2013, 306-10. 3. Hamblett, K. J.; Senter, P. D.; Chace, D. F.; Sun, M. M. C.; Lenox, J.; Cerveny, C. G.; Kissler, K. M.; Bernhardt, S. X.; Kopcha, A. K.; Zabinski, R. F.; Meyer, D. L.; Francisco, J. A., Effects of drug loading on the antitumor activity of a monoclonal antibody drug conjugate. Clin. Cancer Res. 2004, 10 (20), 7063-7070. 4. (a) Maderna, A.; Doroski, M.; Subramanyam, C.; Porte, A.; Leverett, C. A.; Vetelino, B. C.; Chen, Z.; Risley, H.; Parris, K.; Pandit, J.; Varghese, A. H.; Shanker, S.; Song, C.; Sukuru, S. C. K.; Farley, K. A.; Wagenaar, M. M.; Shapiro, M. J.; Musto, S.; Lam, M.-H.; Loganzo, F.; O'Donnell, C. J., Discovery of cytotoxic dolastatin 10 analogues with N-terminal modifications. J Med Chem 2014, 57 (24), 10527-43; (b) Peipp, M.; Gramatzki, M. In Calicheamicin conjugates: gemtuzumab ozogamicin (Mylotarg), inotuzumab ozogamicin, Wiley-VCH Verlag GmbH & Co. KGaA: 2014; pp 1545-1564; (c) Dirico, K. J.; Eustaquio, A. S.; Green, M. E.; He, H.; He, M.; Koehn, F. E.; O'Donnell, C. J.; Puthenveetil, S.; Ratnayake, A. S.; Subramanyam, C.; Teske, J. A.; Yang, H. Y. Preparation of spliceostatin analog antibody conjugates for cancer treatment. WO2014068443A1, 2014; (d) Lambert, J. M., Antibody-maytansinoid conjugates: a new strategy for the treatment of cancer. Drugs Future 2010, 35 (6), 471-480. 5. Lyon, R. P.; Meyer, D. L.; Setter, J. R.; Senter, P. D., Conjugation of anticancer drugs through endogenous monoclonal antibody cysteine residues. Methods Enzymol 2012, 502, 12338. 6. (a) Junutula, J. R.; Mallet, W. Cysteine-engineered humanized or chimeric anti-human MUC16 antibodies and antibody drug conjugates for diagnosis and treatment of cancer. WO2008141044A2, 2008; (b) Panowksi, S.; Bhakta, S.; Raab, H.; Polakis, P.; Junutula, J. R., Site-specific antibody drug conjugates for cancer therapy. MAbs 2014, 6 (1), 34-45. 7. Dennler, P.; Chiotellis, A.; Fischer, E.; Bregeon, D.; Belmant, C.; Gauthier, L.; Lhospice, F.; Romagne, F.; Schibli, R., Transglutaminase-Based Chemo-Enzymatic Conjugation Approach Yields Homogeneous Antibody-Drug Conjugates. Bioconjugate Chem. 2014, 25 (3), 569-578. 8. Lin, K.; Tibbitts, J.; Shen, B.-Q., Pharmacokinetics and ADME characterizations of antibody-drug conjugates. Methods Mol Biol 2013, 1045, 117-31.

ACS Paragon Plus Environment

32

Page 33 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

9. Wiggins, B.; Liu-Shin, L.; Yamaguchi, H.; Ratnaswamy, G., Characterization of Cysteine-Linked Conjugation Profiles of Immunoglobulin G1 and Immunoglobulin G2 Antibody-Drug Conjugates. J. Pharm. Sci. 2015, 104 (4), 1362-1372. 10. Janin-Bussat, M.-C.; Dillenbourg, M.; Corvaia, N.; Beck, A.; Klinguer-Hamour, C., Characterization of antibody drug conjugate positional isomers at cysteine residues by peptide mapping LC-MS analysis. J Chromatogr B Analyt Technol Biomed Life Sci 2015, 981-982, 9-13. 11. Wakankar, A.; Chen, Y.; Gokarn, Y.; Jacobson, F. S., Analytical methods for physicochemical characterization of antibody drug conjugates. MAbs 2011, 3 (2), 161-72. 12. Sapra, P.; DiJoseph, J.; Gerber, H.-P. In Calicheamicin antibody-drug conjugates and beyond, Wiley-VCH Verlag GmbH & Co. KGaA: 2012; pp 395-410. 13. Strop, P.; DELARIA, K. A.; DORYWALSKA, M.; Foletti, D. L.; Dushin, R. G.; Shelton, D. L.; Rajpal, A., Antibody-drug conjugates with high drug loading. Google Patents: 2015.

ACS Paragon Plus Environment

33

Organic Process Research & Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 34

For Table of Contents Only

ACS Paragon Plus Environment

34