Chapter 1 General introduction General introduction 9

Antibody-based therapy has become one of the most important and successful therapeutic modalities to treat cancer. The first application was just 20 years ago, 1 but the concepts are much older. Antibody-based therapy is actually one of the old- est treatments for cancer patients. Back in 1895, two French researchers described a serotherapy in which they treated patients with antisera derived from animals that were injected with cancer cells of the patient themselves1. In some of the patients improvements were observed, however, severe immunological reactions occurred in most cases. The antisera they used consisted of a large variety of antibodies directed towards different antigens. Some of these antigens are expressed not only on malignant cells, but also on normal cells, which caused the immunological reactions. If a method could be developed to produce antibodies directed against one specific antigen, a possible treatment against cancer and many other diseases could be envisioned. This was the foundation of the ‘magic bullet’ hypothesis, proposed a decade later by Paul Ehrlich2. His theory implied that if a compound could be made that selectively targets a disease, then a toxin for that organism could be delivered along with this selective agent.

Antibody structure and function Natural human antibodies are large protein molecules produced by plasma cells, a specific type of white blood cells, and are used to identify and neutralize foreign organisms or molecular structures, called antigens. The structure of antibodies was discovered in the late 1960s3,4. Antibodies consist of two identical heavy chains that are interconnected forming a Y-shape and two light chains that form extra legs to the top of the Y (Figure 1). Heavy and light chains both have a constant and a variable region. Heavy chains determine the class of the antibody, of which there are five types in mammals. The tail of the Y is indicated as the Fc part and the variable regions are the upper parts of the heavy and light chains that form the two upper legs (Fab-part) of the Y. The two light chains are identical and also consist

Figure 1 - Schematic representation of a mAb. 10 Chapter 1

of a bottom constant region and a top variable region. Antigen binding occurs at the variable part. When the antibody is bound to the antigen by the Fab (Fragment, antigen binding) region on the top, the tail, called Fc (Fragment, crystallization) region makes it possible to initiate the further immune response that will remove the antigen from the system.

Development of monoclonal antibodies From the early 1900s on, many researchers tried to find a method to develop anti- bodies directed against one specific antigen, i.e. monoclonal antibodies (mAbs). In 1951, Henri Kunkel investigated cells from multiple myeloma patients and discov- ered that the malignant plasma cells produce antibody molecules with one single specificity instead of several5. However, these cells produce antibodies at random, and researchers had no method to find out against which antigen the antibody was directed. In the late 1960s, researchers began to develop methods to isolate and propagate B-cells that only produce one specific antibody directed against one specific antigen6. The limitation of these methods was the restricted lifespan of the B-cells and the small amount of antibodies produced. In 1975, Köhler and Milstein succeeded in fusing a mouse B-cell, producing a specific and well-known antibody, with a mouse myeloma cell7. This method was called the ‘hybridoma technique’. By using this hybridoma technology they obtained an immortal cell line, which was capable of very high and pure production of one specific antibody. This method to develop specific antibodies did not immediately lead to convinc- ing clinical successes, since the antibodies were derived from mice. Application of murine mAbs met a few limitations: they are quickly removed from the blood, inflammatory reactions can occur, and human anti-mouse antibodies (HAMAs) can develop. When HAMAs are present, patients cannot receive multiple doses of the mAb, because of anaphylactic reactions that might occur8. By genetic engineering, chimeric antibodies were developed that are partially mouse and partially human. The DNA encoding the variable parts of the murine heavy and light chains are fused to the DNA encoding the constant regions of the human heavy and light chains, and these fusion constructs are brought into cultured cells for chimeric antibody expression. This engineering also allows the design of antibodies of a particular isotype such as IgG1 or IgG4. With this approach immune reactions directed against the Fc part of the mAb can be avoided, while the antigen binding part is not altered. As a next step it became possible to replace most of the murine heavy and light chains by the human heavy and light chains, with just the antigen binding parts of murine origin remaining. These mAbs are called “humanized”. Nowadays techniques are available to develop also fully human mAbs, and in some approaches the use of mice or other animals is no longer needed. These methods General introduction 11 also enable the engineering of a variety of antibody fragments, such as: diabodies, minibodies, affibodies and nanobodies9. With the increase in mAb-development, 1 rules for the nomenclature also have been set. All mAbs should have -mab as a suffix, and are preceded by the origin-determination: -o for mouse, -u for human, -xi for chimeric, -xizu for chimeric/humanized, and -zu for humanized.

First mAbs and their mechanisms of action All of the current FDA-approved anti-cancer mAbs are immunoglobulins G (IgGs), with a variety of mechanisms of action, not always predictable and not always known for a new mAb. Often these mechanisms work in parallel, depending on the targeted molecule and the isotype of the mAb. Mechanisms of action include: (i) inhibition of cell signaling, (ii) induction of apoptosis, (iii) antibody-dependent cellular cytotoxicity (ADCC), and (iv) complement-dependent cytotoxicity (CDC). While the first two mechanisms are direct effects of the antibody, the next two are immune-mediated responses. In addition a mAb can serve as a carrier for a toxic agent. The delivery of a toxic agent is central part of this thesis and will be discussed later. The different mechanisms of action are explained hereafter in more detail: (i) in the inhibition of cell signaling the antibody has an antagonistic effect by binding a membrane-bound receptor, often a growth factor receptor that shows and increased level of expression in tumor cells. This either prevents the growth factor from bind- ing to its receptor or interferes with dimerization of the receptor. The result is a decreased signaling in the downstream pathway, leading to inhibition of unwanted cell proliferation and, ideally, cell death. The antibody could also target the growth factor itself, and as a result binding to the receptor cannot occur and cell growth is halted; (ii) antibodies that induce apoptosis have an agonistic effect, and binding to a receptor directly triggers apoptotic signaling in the tumor cells. The Fc-region of IgGs mediates immune-mediated responses: after binding to a target molecule, the Fc-regions engage effector cells of the immune system that subsequently kill the tumor cells; (iii) ADCC is induced by binding of the Fc-region of the mAb to the FcγRIIIa (CD16) receptors on e.g. natural killer (NK) cells, after which NK cells are activated and release cytolytic granules that cause cell death; and (iv) CDC is induced when the Fc-region of the mAb binds to soluble complement factor C1q. This triggers the complement cascade that ultimately leads to cell death.

Clinically approved mAbs The developments of the last years have led to 17 currently FDA- and European Medicines Agency (EMA)-approved mAbs (Table 1). Some of these are exploited as naked mAbs and some as antibody-conjugates. 12 Chapter 1

Table 1 - FDA- and EMA-approved mAbs for use in oncology, adapted from Glassman & Balthasar18 Name Marketed by Class Target First approved Reported Approval indication mechanisms of year action Biogen Idec./ Chimeric CD20 Non-Hodgkin’s ADCC, CDC, 1997 (Rituxan) Genentech IgG1 Lymphoma Induction of Apoptosis Genentech Humanized HER2 Signal Inhibition, 1998 (Herceptin) IgG1 ADCC Sanofi-Aventis Humanized CD52 B cell Chronic CDC, Induction 2001 (Campath) IgG1 Lymphocytic of Apoptosis Leukemia Ibritumomab Biogen Idec Murine IgG1 CD20 Non-Hodgkin’s Radioisotope 2002 tiuxetan Lymphoma Delivery (90Y) (Zevalin) GlaxoSmith- Murine CD20 Non-Hodgkin’s Radioisotope 2003 (Bexxar) Kline IgG2a Lymphoma Delivery (131I), ADCC, CDC, Induction of Apoptosis Bristol-Myers Chimeric EGFR Squamous Cell Signal Inhibition, 2004 (Erbitux) Squibb/Eli Lilly IgG1 Carcinoma of the ADCC, CDC Head and Neck Genentech Humanized VEGF Colorectal Signal Inhibition 2004 (Avastin) IgG1 Cancer Human IgG2 EGFR Colorectal Signal Inhibition, 2006 (Vectibix) Cancer ADCC Genmab/GSK Human IgG1 CD20 Chronic ADCC, CDC 2009 (Arzerra) Lymphocytic Leukemia Denosumab Amgen Human IgG2 RANKL Bone Metastases Signal Inhibition 2010 (Xgeva) Ipilumumab Bristol-Myers Human IgG1 CTLA-4 Metastatic Signal Inhibition 2011 (Yervoy) Squibb Brentuximab Seattle Chimeric CD30 Hodgkin ADC 2011 vedotin Genetics IgG1 Lymphoma (Adcetris) Genentech Humanized HER2 Breast Cancer Signal Inhibition, 2012 (Perjeta) IgG1 ADCC Trastuzumab Genentech Humanized HER2 Breast Cancer ADC, Signal 2013 emtansine IgG1 Inhibition, (Kadcyla) ADCC Obinituzumab Genentech Humanized CD20 Chronic ADCC, CDC, 2013 (Gazyva) IgG1; Glyco- Lymphocytic Induction of engineered Leukemia Apoptosis Merck Sharp & Humanized PD-1R Metastatic Signal Inhibition 2014 (Keytruda) Dohme Corp IgG4 Melanoma Bristol-Myers Human IgG4 PD-1R Metastatic Signal Inhibition 2014 (Opdivo) Squibb Melanoma General introduction 13

Naked mAbs In 1997, rituximab was the first mAb in oncology that was approved by the US 1 Food and Drug Administration (FDA). It is directed against the CD20 antigen and used for patients with non-Hodgkin’s lymphoma (NHL). Hematologic malignan- cies are relatively easy to target, because the cells are in close connection with the blood stream and therefore directly accessible for mAb-therapy. CD20 is highly expressed on normal B cells and B-cell lymphoma cells and is not shed into the plasma. In vitro, no internalization after binding of rituximab is observed, but in vivo internalization cannot be excluded. Although also normal B cells are cleared by rituximab treatment, this seems to have no effect on the immune system, since no increase in the incidence of infections is observed in these patients10. Upon antibody binding, B-cell proliferation is directly inhibited, nuclear DNA fragmenta- tion is induced and this leads to cell death by apoptosis11. After being a decade on the market, rituximab still makes 5 billion dollar sales per year12. One year after the approval of rituximab, a mAb targeting solid tumors was ap- proved: trastuzumab. This humanized mAb is directed against human epidermal growth factor receptor 2 (HER2), which is expressed by many epithelial tissues and over-expressed in 20-30% of breast cancers. Trastuzumab binds to the extracellular domain IV of the HER2 receptor and blocks activation of the downstream signaling cascade. Other indicated effects are prevention of HER2-receptor dimerization, increased endocytotic destruction of the receptor, inhibited shedding of the extra- cellular domain, and activation of the immune system by ADCC13. Trastuzumab is used for treatment of metastatic HER2-positive breast cancer and as adjuvant therapy for other HER2-positive breast tumors. Alemtuzumab became approved in 2001, and is an anti-CD52 mAb used for patients with B-cell chronic lymphocytic leukemia. It exerts anti-tumor activity by ADCC. In 2012, alemtuzumab was withdrawn from the market and was re- launched in 2014 as a treatment for multiple sclerosis at a much higher price. Cetuximab, approved in 2004, is directed against the epidermal growth factor receptor (EGFR = HER1) and was first used in advanced colorectal cancer. It is only effective for tumors that express EGFR and have a wild type KRAS gene. Like trastu- zumab, cetuximab blocks the downstream signaling cascade after binding with high affinity to the receptor. KRAS is downstream in the signaling cascade, when mutated it causes uncontrolled cellular proliferation and as it acts downstream of EGFR it makes the tumor cell insensitive for cetuximab treatment. It also triggers ADCC and CDC. Currently, cetuximab is registered for the treatment of metastatic colorectal cancer (mCRC), metastatic non-small cell lung cancer (mNSCLC) and head-and-neck squamous cell carcinoma (HNSCC). 14 Chapter 1

Bevacizumab, also approved in 2004, is a mAb directed against vascular en- dothelial growth factor (VEGF), a growth factor associated with angiogenesis and overexpressed in many solid tumors. Bevacizumab binds to VEGF, thereby causing inhibition of angiogenesis. When tumors grow larger, the cells become hypoxic and start releasing VEGF and other angiogenic factors. This makes bevacizumab suitable for use in many different solid tumor types. Currently it is approved for treatment of patients with mCRC, NSCLC, metastatic breast cancer, metastatic renal cell cancer, prostate cancer, and glioblastoma14. Panitumumab, approved in 2006, is an anti-EGFR mAb, like cetuximab, used in the treatment of mCRC and also requires wild type KRAS to be effective. In contrast to cetuximab, which is a chimeric IgG1 and therefore ideally suited for ADCC activity, panitumumab is a human IgG2 mAb, less active in ADCC. However, clini- cal data do not show differences in the efficacy of these two anti-EGFR mAbs in colorectal cancer 15, while there is a difference in HNSCC. Ofatumumab, approved in 2009, targets CD20 like rituximab and is approved for use in treatment of chronic lymphocytic leukemia (CLL). The binding and im- munological properties of ofatumumab differ from those of rituximab16, therefore it might be a suitable mAb to treat rituximab-refractory patients. In 2012, a clinical study showed a modest efficacy of ofatumumab treatment in rituximab-refractory patients17. Denosumab, approved in 2010, targets less frequent occurring tumors. It is directed against the receptor activator of nuclear factor kappa-B ligand (RANKL), a surface molecule on osteoclasts, and is used for treatment of bone metastases from solid tumors and for unresectable giant cell bone tumors18. , the first immune checkpoint mAb (elaborated in the part ‘Other next generation mAbs’) approved in 2011, is used for treatment of metastatic melanoma and is directed against cytotoxic T lymphocyte–associated antigen 4 (CTLA-4), a protein receptor that down-regulates the immune system19. Pertuzumab, approved in 2012, is also directed against HER2 and used to treat patients with metastatic breast cancer. Pertuzumab targets a different motif in HER2 than trastuzumab: it binds domain II of the extracellular domain of HER2, and is more effective in preventing receptor heterodimerization with EGFR, HER3, and HER4, and subsequent cell signaling. Preclinical studies showed synergy between trastuzumab and pertuzumab, providing possibilities for combination therapy20. Therefore, a phase II clinical study was performed, which showed that the addition of pertuzumab to the standard trastuzumab and docetaxel combination treatment resulted in a significant clinical benefit21. General introduction 15

Conjugated mAbs Using mAbs as delivery vehicles of toxic agents is one of the most promising ap- 1 plications in antibody therapy. During the earliest mAb research this approach was tested because of the initial lack of successes with naked therapeutic murine mAbs. Delivery of toxic payloads can be divided in three categories: radioimmunocon- jugates, immunotoxins and antibody-drug conjugates (ADCs). The development of mAbs coupled to radioactive or toxic compounds makes the chemistry more complicated and demands a higher standard of safety measurements: (i) conju- gates preferably have exactly the same pharmacokinetics as their non-conjugated counterparts; (ii) biological half-life and binding properties to the target should not be altered after conjugating an additional group to the mAb; (iii) in vivo stability of the chemical bond between the payload and the mAb is also of major importance, since it is undesirable that some of the toxic molecules are released from the con- jugate before they reach the tumor site. Radioimmunoconjugates are mAbs with a radionuclide or radionuclide-contain- ing compound attached to it. When the mAb binds to the tumor cell, the radiation damage causes cell death of the targeted cell and possibly the surrounding cells, depending on the radiation energy of the radionuclide. For use in this therapeutic setting, the radionuclide needs to have a physical half-life long enough to allow the mAb to reach the tumor and have an α- or β– -decay. The two radionuclides used in FDA-approved radioimmunoconjugates are 131I and 90Y, with half-lifes of 8.02 days and 2.66 days, respectively. 90Y emits high-energy β’s that have a mean range in tissue of 2.7 mm, and is therefore most suitable for large tumors; 131I emits β’s with lower energy, the mean tissue range is 0.8 mm, and is better suited for treatment of smaller tumors22. The radionuclide is coupled directly to the tyrosines of the mAb (131I) or via a chelate (90Y). The two radioimmunoconjugates that made it to approval are (90Y) in 2002 and tositumomab (131I) in 2003 (see Table 1). Both antibodies are of murine origin with CD20 as the target. They are used for treatment of patients with NHL, which is attractive since these cells are sensitive for radiation. Nevertheless, unlabeled mAbs like rituximab and its successors are still favored in clinical use, since working with radiopharmaceu- ticals always requires extra precautions for the patient and physician. In recent years, new radioimmunoconjugates have not been approved for clinical use. Main reasons are the inadequate targeting of many tumor types for radioimmunotherapy and the previously mentioned extra precautions that need to be implemented for transportation and clinical use when working with radiopharmaceuticals. Immunotoxins have a catalytic toxin as therapeutic entity. These toxins are enzymes of bacteriological origin or other types of pathogens. After endocytosis, the cleaved toxins cause termination of protein synthesis, which results in cell 16 Chapter 1

death by apoptosis. A complication in the development of immunotoxins is that the toxins are of non-mammalian origin, which causes a reaction of the human immune system in patients. Neutralizing antibodies are produced against the toxin, limiting the number of treatments23. Also, most toxins have binding sites for surface molecules expressed by normal tissues, and these binding sites have to be modified before the toxins can be used therapeutically. In addition, therapy studies revealed that efficacy could be impeded by cellular resistance mechanisms24. While im- munotoxins have a lot of potential, a large effort in optimization is required before they can be successfully brought to the clinic.

ADCs are currently the most thoroughly explored immunoconjugates. For the last 20 years researchers have tried to develop the perfect combination of a highly toxic payload, linked through a stable linker to a specifically and selectively targeting mAb. Early studies tested classical chemotherapeutic compounds, such as doxoru- bicin, coupled to a mAb, but these conjugates failed to show therapeutic efficacy25. These failures can be explained by the small amount and relatively low potency of the toxic compound that is released in the tumor. Compared to a single injec- tion of the free toxic compound, the amount of toxic compound that is delivered to the tumor is much lower when coupled to a mAb26. Therefore, the ideal ADC allows the application of highly toxic compounds, much more toxic than current chemotherapeutics, that otherwise have a therapeutic window which is too narrow to allow their use as free drug27. This also requires stable conjugation to avoid systemic toxicity by release of the toxic compound in the circulation, but once internalized the drug should be released to be able to kill the tumor cell. Only then, the desired therapeutic effect will be observed. The toxic compounds used for ADCs can be divided in two main categories: DNA-damaging agents and microtubule inhibitors. The most studied toxins are tubulin inhibitors derived from maytansine, maytansoids, and the auristatins, synthetic analogues derived from the natural occurring dolastatin 10, and the DNA-damaging agent calicheamicin28. The development of synthetic versions of these natural occurring toxins provided the possibility of adding specific chemi- cal groups, which enables easy coupling to linkers while keeping toxicity in the nanomolar range. Gemtuzumab-ozogamicin was in 2000 the first FDA approved ADC and used for the treatment of acute myeloid leukemia (AML) (Table 1). This IgG4 mAb is directed against CD33, a surface glycoprotein, and the cytotoxic antibiotic cali- cheamicin is linked to this mAb by a hydrazone (pH labile) linker. The IgG4-isotype was used because it does not induce ADCC activity and related cytokine release. Gemtuzumab-ozogamicin was used as single agent for patients over 60 with General introduction 17

CD33-positive AML in first relapse or those that are not considered candidates for cytotoxic chemotherapy29. However, 10 years after its approval, gemtuzumab- 1 ozogamicin was voluntarily withdrawn from the market due to the lack of clinical benefit and unacceptable systemic cytotoxicity30. The poor results with this first approved ADC demonstrate the importance of getting more knowledge on the in vivo behavior of ADCs, e.g. in animal models. A decade after the approval of gemtuzumab-ozogamicin, the second ADC was approved for clinical use: (Table 1), directed against CD30 and used to treat patients with Hodgkin’s lymphoma. The drug used in this ADC is the auristatin derivative (MMAE), coupled to the mAb via a cleavable peptide (valine-citrulline) linker. Brentuximab vedotin was developed after early-phase patient data revealed only modest clinical activity of the naked mAb, originally named cAC10. In 2003, the ADC was developed under the name cAC10-vcMMAE/SGN-35, but the ADC was renamed brentuximab vedotin in 2008 when clinical studies were started. The next approved ADC was ado-trastuzumab-emtansine in 2013 (Table 1). The ADC consists of trastuzumab and the maytansine derivate DM1, coupled via a thioether linker proclaimed to cause the release of the drug upon proteolytic degradation only. The thioether linked ADC gave the better therapeutic effect, since the release of the toxic compound in the circulation was minimized compared to other, less stable, hydrazone linkers31. Ado-trastuzumab emtansine is approved for treatment of patients with metastatic breast cancer who previously received trastuzumab and a taxane, separately or in combination32.

At the moment more than 30 ADC’s are in early- or late-stage clinical develop- ment28. The ADCs in late stage clinical development all contain the previously mentioned toxic compounds (Table 2). Researchers tried to learn from the withdrawal of and developed a closely related ADC: , the same drug and linker, coupled to CD22, which is another surface glycoprotein that can be used to treat patients with AML/NHL. The naked CD22 antibody showed no clinical efficacy, but is readily internalized, which makes it ideal for use in an ADC, as was confirmed by preclinical and early clinical studies33. Clinical phase III studies with izotuzumab ozogamicin in combination with rituximab are currently ongoing. is also directed against CD22 and contains an auristatin toxic compound, like in the approved brentuximab vedotin. It is also a potential competitor of rituximab, the current standard of care for treatment of NHL in com- bination with chemotherapy, as was shown in preclinical in vivo efficacy studies34. 18 Chapter 1

Therefore, a phase II study was started that combines pinatuzumab vedotin with rituximab35, to find a possible treatment for rituximab-refractory patients. For treatment of solid tumors, and are currently in phase II studies. Glembatumumab vedotin contains a mAb directed against glycoprotein non-metastatic melanoma protein B (GPNMB), which is mainly expressed in breast cancers and . The toxic MMAE is conjugated to the mAb via a cleavable peptide linker, like in brentuximab vedotin. Clinical studies with glembatumumab vedotin for treatment of both advanced melanoma36 and breast cancer37 show promising results. Lorvotuzumab mertansine is composed of an anti-CD56 antibody, linked via a cleavable disulfide linker to the maytansinoid DM1. CD56 is the neuronal cell adhesion molecule and is expressed by various tumor types. Lorvotuzumab mertansine has shown efficacy in combina- tion therapy in multiple myeloma patients38 and shows potential in small-cell lung cancer (SCLC) xenograft models39. These combined results were promising enough to immediately start with the phase II clinical trial for treatment of SCLC35. Despite increasing popularity of ADCs, it is fair to state that the development of ADCs is still at its infancy. At this moment, most ADCs under clinical development contain maytansines or auristatins as the toxic payload (Table 2). It is not totally clear whether releasable, cleavable or non-cleavable linkers are best suited for appropriate drug delivery to the tumor. To get better understanding of optimal ADC design, in vivo tracking of ADCs would be very informative.

Table 2 - Clinical-stage ADC pipeline, adapted from Mullard35 ADC Lead Lead indications Target Payload Phase Inotuzumab Pfizer Aggresive non- CD22 Calicheamicin III ozogamicin (CMC- Hodgkin’s lymphoma; 544) ALL RG-7596 Genentech DLBCL and follicular CD79b MMAE II non-Hodgkin’s lymphoma Pinatuzumab Genentech DLBCL and follicular CD22 MMAE II vedotin (RG-7593) non-Hodgkin’s lymphoma Glembatumumab Celldex Breast cancer GPNMB MMAE II vedotin SAR-3419 Sanofi DLBCL; ALL CD19 DM4 II Lorvotuzumab ImmunoGen Small-cell lung cancer CD56 DM1 II mertansine (IMGN- 901) BT-062 BioTest Multiple myeloma CD138 DM4 II PSMA-ADC Progenics Prostate cancer PSMA MMAE II General introduction 19

Table 2 (continued) ADC Lead Lead indications Target Payload Phase 1 ABT-414 AbbVie Glioblastoma; NSCLC; EGFR Not disclosed I/II solid tumor Melatuzumab Immunomedics CLL; multiple CD74 Doxorubicin I/II doxorubicin myeloma; non- Hodgkin’s lymphoma IMMU-132 Immunomedics Solid tumor TACSTD2 Irinotecan I (also known as metabolite TROP2 or EGP1) - Immunomedics Cancer; colorectal CEA (also known Irinotecan I SN-38 cancer as CD66e) metabolite IMGN-853 ImmunoGen Ovarian tumor; solid Folate receptor 1 DM4 I tumor IMGN-529 ImmunoGen B cell lymphoma; CD37 DM1 I CLL; non-Hodgkin’s lymphoma RG-7458 Genentech Ovarian tumor Mucin 16 MMAE I RG-7636 Genentech Melanoma Endothelin MMAE I receptor ETB RG-7450 Genentech Prostate cancer STEAP1 MMAE I RG-7600 Genentech Ovarian tumor; Not disclosed Not disclosed I pancreatic tumor RG-7598 Genentech Multiple myeloma Not disclosed Not disclosed I RG-7599 Genentech NSCLC; ovarian tumor Not disclosed Not disclosed I SGN-CD19A Seattle Genetics ALL; aggressive non- CD19 MMAE I Hodgkin’s lymphoma Vorsetuzumab Seattle Genetics Non-Hodgkin’s CD70 MMAF I mafodotin lymphoma; renal cell carcinoma ASG-5ME Agensys Cancer; pancreatic SLC44A4 MMAE I tumor; stomach tumor (AGS-5) ASG-22ME Agensys Solid tumor Nectin 4 MMAE I AGS-16M8F Agensys Cancer; renal cell AGS-16 MMAF I carcinoma MLN-0264 Millennium Gastrointestinal tumor; Guanylyl MMAE I solid tumor cyclase C SAR-566658 Sanofi Solid tumor Mucin 1 DM4 I AMG-172 Amgen Cancer; renal cell CD70 Not disclosed I carcinoma AMG-595 Amgen Glioma EGFRvIII DM1 I BAY-94-9343 Bayer Cancer; mesothelioma Mesothelin DM4 I

ALL, acute lymphoblastic leukemia; CEA, carcinoembryonic antigen; CLL, chonic lymphoblastic leuke- mia; DLBCL, diffuse large B cell lymphoma; GPNMB, glycoprotein NMB; MMAF, monomethyl auristatin F; NSCLC, non-small-cell lung cancer; PSMA, prostate-specific membrane antigen; STEAP1, six-trans- membrane epithelial antigen of prostate 1; TACSTD2, tumour-associated calcium signal transducer 2. 20 Chapter 1

Other next generation mAbs While in the ADC-development the focus is on improving the linker/toxic com- pound chemistry, the mAb itself is also undergoing development, contributing to next generation mAbs. Glycoengineered mAbs are modified in their glycosylation pattern in the Fc-region to enhance ADCC and CDC, via improved binding to Fcγ receptors on immune effector cells. This creates a more potent naked mAb, compared to a normal IgG1, with ofatumumab40 as the first successful example of this strategy and obinituzumab the second that was approved in 2013 (Table 1). A new and exciting development is the clinical application of mAbs that in- hibit immune-checkpoints. When an immune response is elicited, also a negative feedback mechanism is activated to dampen the response, and these mechanisms are indicated as immune-checkpoints. These mechanisms are protective, but hamper effective anti-tumor immune responses. In fact tumor cells can exploit these mechanisms to counteract immune responses41. Cytotoxic T-lymphocyte- associated antigen 4 (CTLA-4) and programmed cell death protein 1 (PD-1) have proven to be important immune-checkpoint proteins and promising therapeutic targets for mAbs. The previously mentioned ipilimumab is directed against CTLA-4 and nivolumab and pembrolizumab are directed against PD-1. Nivolumab and pembrolizumab, approved in 2014 (Table 1), reactivate T cells by blocking APC (antigen presenting cell) inhibition of T cells or by blocking tumor cells, which of- ten inactivate T cells in the microenvironment via their overexpressed PD-1 ligands (e.g. PD-L1)42. An advantage of these immune-checkpoint inhibiting mAbs is that they can be beneficial in many types of tumors, hematological and solid. Multi-specific antibodies are able to recognize more than one target molecule, by exploiting single heavy chain-light chain pairs from antibodies with distinct speci- ficities. The different pairs can be linked to form two arms with different specificity within a single IgG molecule. MEHD7945A, currently in phase II clinical trials, is directed against EGFR as well as HER3, which are closely related receptors, and shows promising results in the treatment of multidrug resistant tumors43. Besides intact mAbs and antibody fragments, nanobodies included, advances in genetic engineering have led to the development of several non-immunoglobin- like binding molecules. They have antibody-like binding properties, but completely lack the antibody scaffold; and the novel structures are designed as derivatives of naturally occurring substrates or proteins. Anticalins, affibodies, adnectins, designed ankyrin repeat proteins (DARPins) and avimers are among the most developed non-Ig-based protein scaffolds44. Claimed advantages of these ligands are their small size, max. 200 amino acids, and the lack of immunogenicity. Ad- nectin CT-322 and DARPin AGN150998/MP0112 are currently in clinical phase II development. General introduction 21

Nuclear imaging in mAb development and applications The continuous progress in the development of mAbs and mAb conjugates demands 1 for accurate, preferably quantitative, methods to monitor the in vivo behavior of these potentially toxic molecules. Therefore, not only mAbs have been in develop- ment for therapeutic exploitation the last decades, but many researchers have put as much effort in the development of diagnostic mAbs as well. Most commonly used diagnostic mAbs for imaging purposes consist of a mAb carrying a radioactive isotope for use in single photon emission computed tomog- raphy (SPECT) or positron emission tomography (PET)45. The quality of gamma cameras as well as PET-scanners in terms of sensitivity and resolution has improved over the last years, causing an increase in mAb-imaging. However, PET imaging has a higher sensitivity and more accurate quantification than SPECT. Immuno-PET, imaging of mAbs labeled with positron emitters, can provide whole body images, thereby revealing important information both about the tumor and possible metastases. Also the level of antigen expression in tumor and other tissues can be determined, indicating the tumor-specificity of the mAb: crucial knowledge for therapeutic potential of the mAb. Immuno-PET can also be used to optimize mAb dose scheduling and to track changes in tumor biology over time, including assessments of the antigen expression before and after therapy. Further- more, immuno-PET provides the possibility to select patients for therapy, thereby increasing therapeutic efficacy and decreasing adverse events. The radionuclides that can be used for immuno-PET, need to have a physical half- life that matches the biological half-life of the mAb. Since almost all approved mAbs and mAbs in clinical development are intact mAbs with a half-life of several days, 124I 89 (t1/2=100.3 hours) and Zr (t1/2=78.4 hours) are attractive candidate radionuclides. For both isotopes labeling procedures have been developed that ensure safety-controlled production for human use all over the world9. Conjugation with 124I is relatively easy, only an oxidizing agent is necessary to enable the direct labeling reaction of 124I to the tyrosine moiety of the mAb. Labeling mAbs with 89Zr requires premodification of the mAb, which implies the introduction of an extra group - a chelator - to be able to attach the metal radionuclide in a stable way. The first method used for clinical immuno-PET was published in 200346 and required a two steps conjugation. The same group published an improved version in 2010 with a single step conjugation47. An important reason to choose between 124I or 89Zr is the nature of the mAb, whether it is internalizing or not. Iodine is released from the cells after internalization, while the zirconium is residualizing and remains in the cell after internalization. The latter makes 89Zr the radionuclide of choice for many immuno-PET applications. Many successful preclinical and clinical studies with 124I- and 89Zr labeled mAbs have been performed and are described in detail in comprehensive reviews9, 48-51. 22 Chapter 1

Near-infrared imaging in mAb applications Although immuno-PET has many advantages for whole body imaging, resolution is limited and real-time imaging is impossible. Photoimmunodetection, in which mAbs are labeled with fluorescent dyes, might have a complementary clinical potential to immuno-PET. It allows high-resolution, real-time, dynamic imaging of superficial tissue layers at the cellular level, without radiation burden to the patient. Therefore, it might be ideal for confirmation of mAb binding to subsets of cells. In addition, it might be a powerful clinical tool for the delineation and characterization of early-stage or residual disease, for example of cancer during surgery or in a screening setting. A limitation of optical imaging in comparison to PET imaging is the difficulty of deep tissue imaging and of quantification. Fluorescent dyes in the near-infrared (NIR) region are the most suitable for use in imaging. The light emitted after excitation is in the region of 650-900 nm. At this wavelength tissue absorption and scattering is low, tissue penetration is in the range of 1 to 2 centimeters and tissue auto-fluorescence in not interfering with dye signal52. The challenge is to develop a fluorescent probe that maintains its fluorescent properties after conjugation to mAbs. Studies with ICG, the only FDA-approved NIR dye, showed promising results when used as a single agent53, however, when the molecule was modified for coupling to a mAb a serious loss of fluorescence was observed54. Different fluorescent dyes have been tested for coupling abilities and the dye with the highest potential to be used in the clinical setting is IRDye800CW (Licor Bioscience). IRDye800CW can be functionalized with either an NHS or a maleimide reactive group, ideal for conjugation to mAbs, as illustrated in several preclinical studies with IRDye800CW-mAb conjugates55-59. The first clinical trials with IRDye800CW labeled mAbs – bevacizumab and cetux- imab – are currently ongoing (Clinicaltrials.gov, NCT01987375, NCT02113202, NCT01508572, NCT01972373, NCT02129933). Even more attractive seems the use of dual-labeled mAbs, since combining PET and optical imaging techniques would provide the ideal probe: the location and size of the tumor and metastases can be determined with immuno-PET, followed by a complete tumor resection aided by imaging of the NIR-labeled mAb dur- ing surgery. For these dual-labeled conjugates with IRDye800CW, FDA-approved mAbs trastuzumab60 and bevacizumab61 as well as new mAbs were used. Also dif- 89 62,63 64 ferent PET radionuclides were used. Besides the expected Zr , also Cu (t1/2= 12.7 hours) was investigated64, 65. All studies show tumor specific uptake of the conjugate with both imaging modalities, confirming the potential of IRDye800CW. However, clinical translation of dual-labeled conjugates still has some hurdles to take. Clinical use of 64Cu with intact mAbs seems less desirable, since the half-life of 64Cu does not match the biological half-life of intact mAbs in humans. Moreover, General introduction 23 stable coupling of 64Cu to mAbs is challenging, and sequestration of 64Cu in the liver might occur. In addition, the number of conjugated dye and chelator groups 1 per mAb was different for every study performed, and this might affect antigen binding and pharmacokinetics. When clear protocols would be available for the reproducible production of stable dual-labeled mAbs, clinical use of PET/NIR molecules will most probably be the next translational step.

OUTLINE OF THIS THESIS

As outlined above in Chapter 1, mAbs have shown to be very promising candidates in cancer therapy as naked single agents, but mostly when used as carrier for tumor- specific delivery of potent toxic compounds. The latter requires a reproducible synthesis protocol that is relatively straightforward and GMP compliant. Moreover, labeling of the mAbs and mAb-conjugates facilitates product development and their clinical applications. The more toxic the naked mAb or the toxic compound to be delivered by the mAb are, the more important it becomes that tumor targeting is “precise”; precise at the level of the individual patient and precise at the level of a high uptake in the tumor and a low uptake in the normal tissues. In this thesis the development of tools to allow the quantitative tracking of mAbs in vivo by PET or optical imaging is described, which might be of help in steering efforts in the development and (pre)clinical application of precision medicine with mAbs and other biologicals. In Chapter 2, we focus on the use of radioactive labels in the development and in vivo validation of novel antibody drug conjugates (ADCs). ADCs comprise a class of drugs receiving a lot of attention from pharmaceutical companies involved in the development of anti-cancer drugs (see General introduction). We introduce a new family of highly potent drugs, tubulysins, to be coupled to mAbs as novel ADCs. For efficient ADC development, tubulysins are labeled with 131I, while the mAb under investigation in this study (trastuzumab) is labeled with 89Zr. By using this approach of “dual-radiolabeling” we monitor the direct conjugation of tubulysin derivatives to trastuzumab, to get full control over drug-antibody ratio (DAR), and we characterize the tubulysin-trastuzumab conjugates for their pharmacokinetics and tumor targeting properties. Finally, we describe the in vivo efficacy of these ADCs, which have been developed by guidance of radiolabeling technology. As outlined in the introduction section, 89Zr labeling of mAbs allows the quan- titative in vivo characterization of mAbs and other biologicals using PET imaging. PET imaging is particularly well suited for whole body imaging, but is not qualified to show tumor targeting at the cellular level. For the latter, labeling of mAbs with 24 Chapter 1

fluorescent dyes in combination with optical imaging, is better suited as it allows high resolution and real time imaging. Irrespective the use of PET radionuclides or fluorescent dyes for mAb tracking, to be acceptable for pharmaceutical companies and regulatory authorities, the procedures for coupling to mAbs should we well controlled and inert, not affecting the biological properties of the mAb. In Chapter 3, we use 89Zr labeling as an internal reference standard to prove that labeling of the mAbs cetuximab and bevacizumab with the NIR dye IRDye800CW can be performed in a controlled an inert way, without affecting binding and biodistribution characteristics of the mAbs. To this end, different equivalents of IRDye800CW are coupled to the mAbs. To allow broad scale preclinical and clinical application of single- as well as of dual-modal fluorescence and PET imaging, Chapter 4 presents a detailed protocol for the inert coupling of IRDye800CW and/or 89Zr to mAbs in a current good manufacturing practice-compliant (cGMP) way. The previous study (Chapter 3) provided the necessary information to design a user and product friendly protocol: to avoid unnecessary radiation exposure to the product and personnel, first the chelator desferal (Df) is conjugated to the mAb, second the NIR-dye IRDye800CW is coupled, and finally the conjugate is labeled with 89Zr. While imaging of mAbs labeled with NIR dyes showed promising results, quan- tification of the fluorescent signal is more challenging. In Chapter 5, we evaluate a new method to quantify the uptake of IRDye800CW-cetuximab in organs and tumors. Also here we use dual-labeling with 89Zr and standard gamma ray quanti- fication method as reference and for validation purposes. In Chapter 6 a summary of the results is given as well as a general discussion and future perspectives on the topics. In Chapter 7 a summary in Dutch is provided. General introduction 25

REFERENCES 1 1. Hericourt J, Richet CH. De la serotherapie dans le traitement du cancer. CR Hebd Acad Sci 1895; 120: 567-9. 2. Ehrlich P. Experimental researches on specific therapy on immunity with special reference to the relationship between distribution and action of antigens. Collected papers 1908; 3: 1956-60. 3. Porter RR. The structure of antibodies. The basic pattern of the principal class of molecules that neutralize antigens (foreign substances in the body) is four cross-linked chains. This pattern is modified so that antibodies can fit different antigens. Sci Am 1967; 217: 81-7 passim. 4. Edelman GM, Gall WE, Waxdal MJ, Konigsberg WH. The covalent structure of a human gamma G-immunoglobulin I. Isolation and characterization of the whole molecule, the polypeptide chains, and the tryptic fragments. Biochemistry 1968; 7: 1950-8. 5. Kunkel HG, Slater RJ, Good RA. Relation between certain myeloma proteins and normal gamma globulin. Proc Soc Exp Biol Med 1951; 76: 190-3. 6. Potter M, Boyce CR. Induction of plasma-cell neoplasms in strain BALB/c mice with mineral oil and mineral oil adjuvants. Nature 1962; 193: 1086-7. 7. Köhler G, Milstein C. Continuous cultures of fused cells secreting antibody of predefined speci- ficity. Nature 1975; 256: 495-7. 8. Green C, Murray L, Hortobagyi N. therapy for solid tumors. Cancer Treat- ment Reviews 2000; 26: 269-86. 9. van Dongen GA, Vosjan MJ. Immuno-Positron Emission Tomography: shedding light on clinical antibody therapy. Cancer Biother Radiopharm 2010; 25: 375-85. 10. Stern M, Herrmann R. Overview of monoclonal antibodies in cancer therapy: present and promise. Critical Reviews in Oncology/Hematology 2005; 54: 11-29. 11. Shan D, Ledbetter JA, Press OW. Apoptosis of malignant human B cells by ligation of CD20 with monoclonal antibodies. Blood 1998; 91: 1644-52. 12. Scolnik PA. mAbs: A business perspective. MAbs 2009; 1: 179-84. 13. Hudis CA. Trastuzumab--mechanism of action and use in clinical practice. N Engl J Med 2007; 357: 39-51. 14. Mukherji SK. Bevacizumab (Avastin). AJNR Am J Neuroradiol 2010; 31: 235-6. 15. Price TJ, Peeters M, Kim TW, Li J, Cascinu S, Ruff P, et al. Panitumumab versus cetuximab in patients with chemotherapy-refractory wild-type KRAS exon 2 metastatic colorectal cancer (ASPECCT): a randomised, multicentre, open-label, non-inferiority phase 3 study. Lancet Oncol 2014; 15: 569-79. 16. Barth MJ, Mavis C, Czuczman MS, Hernandez-Ilizaliturri FJ. Ofatumumab exhibits enhanced in vitro and in vivo activity compared to rituximab in pre-clinical models of mantle cell lymphoma. Clin Cancer Res 2015; 21: . 17. Czuczman MS, Fayad L, Delwail V, Cartron G, Jacobsen E, Kuliczkowski K, et al. Ofatumumab monotherapy in rituximab-refractory follicular lymphoma: results from a multicenter study. Blood 2012; 119: 3698-704. 18. Glassman PM, Balthasar JP. Mechanistic considerations for the use of monoclonal antibodies for cancer therapy. Cancer Biol Med 2014; 11: 20-33. 19. Lipson EJ, Drake CG. Ipilimumab: an anti-CTLA-4 antibody for metastatic melanoma. Clin Cancer Res 2011; 17: 6958-62. 26 Chapter 1

20. Scheuer W, Friess T, Burtscher H, Bossenmaier B, Endl J, Hasmann M. Strongly enhanced anti- tumor activity of trastuzumab and pertuzumab combination treatment on HER2-positive human xenograft tumor models. Cancer Res 2009; 69: 9330-6. 21. Baselga J, Cortés J, Kim SB, Im SA, Hegg R, Im YH, et al. Pertuzumab plus trastuzumab plus docetaxel for metastatic breast cancer. N Engl J Med 2012; 366: 109-19. 22. Karagiannis TC. Comparison of different classes of radionuclides for potential use in radioim- munotherapy. Hell J Nucl Med 2007; 10: 82-8. 23. Schrama D, Reisfeld RA, Becker JC. Antibody targeted drugs as cancer therapeutics. Nat Rev Drug Discov 2006; 5: 147-59. 24. Redman JM, Hill EM, AlDeghaither D, Weiner LM. Mechanisms of action of therapeutic anti- bodies for cancer. Molecular Immunology 2015; 67: 28-45. 25. Jaracz S, Chen J, Kuznetsova LV, Ojima I. Recent advances in tumor-targeting anticancer drug conjugates. Bioorg Med Chem 2005; 13: 5043-54. 26. McCombs JR, Owen SC. Antibody drug conjugates: design and selection of linker, payload and conjugation chemistry. AAPS J 2015; 17: 339-51. 27. Lambert JM. Drug-conjugated monoclonal antibodies for the treatment of cancer. Curr Opin Pharmacol 2005; 5: 543-9. 28. Gerber HP, Koehn FE, Abraham RT. The antibody-drug conjugate: an enabling modality for natural product-based cancer therapeutics. Nat Prod Rep 2013; 30: 625-39. 29. Ricart AD. Antibody-drug conjugates of calicheamicin derivative: gemtuzumab ozogamicin and inotuzumab ozogamicin. Clin Cancer Res 2011; 17: 6417-27. 30. Bornstein GG. Antibody drug conjugates: preclinical considerations. AAPS J 2015; 17: 525-534. 31. Lewis Phillips GD, Li G, Dugger DL, Crocker LM, Parsons KL, Mai E, et al. Targeting HER2- positive breast cancer with trastuzumab-DM1, an antibody-cytotoxic drug conjugate. Cancer Res 2008; 68: 9280-90. 32. Amiri-Kordestani L, Blumenthal GM, Xu QC, Zhang L, Tang SW, Ha L, et al. FDA approval: ado-trastuzumab emtansine for the treatment of patients with HER2-positive metastatic breast cancer. Clin Cancer Res 2014; 20: 4436-41. 33. Shor B, Gerber HP, Sapra P. Preclinical and clinical development of inotuzumab-ozogamicin in hematological malignancies. Mol Immunol 2015; 67: 107-16. 34. Li D, Poon KA, Yu SF, Dere R, Go M, Lau J, et al. DCDT2980S, an anti-CD22-monomethyl auristatin E antibody-drug conjugate, is a potential treatment for non-Hodgkin lymphoma. Mol Cancer Ther 2013; 12: 1255-65. 35. Mullard A. Maturing antibody-drug conjugate pipeline hits 30. Nat Rev Drug Discov 2013;12: 329-32. 36. Ott PA, Hamid O, Pavlick AC, Kluger H, Kim KB, Boasberg PD, et al. Phase I/II study of the antibody-drug conjugate glembatumumab vedotin in patients with advanced melanoma. J Clin Oncol 2014; 32: 3659-66. 37. Bendell J, Saleh M, Rose AA, Siegel PM, Hart L, Sirpal S, et al. Phase I/II study of the antibody- drug conjugate glembatumumab vedotin in patients with locally advanced or metastatic breast cancer. J Clin Oncol 2014; 32: 3619-25. 38. Berdeja JG. Lorvotuzumab mertansine: antibody-drug-conjugate for CD56+ multiple myeloma. Front Biosci (Landmark Ed) 2014; 19: 163-70. 39. Whiteman KR, Johnson HA, Mayo MF, Audette CA, Carrigan CN, LaBelle A, et al. Lorvotuzumab mertansine, a CD56-targeting antibody-drug conjugate with potent antitumor activity against small cell lung cancer in human xenograft models. MAbs 2014; 6: 556-66. General introduction 27

40. Evans JB, Syed BA. From the analyst’s couch: Next-generation antibodies. Nat Rev Drug Discov 2014; 13: 413-4. 1 41. Shuptrine CW, Surana R, Weiner LM. Monoclonal antibodies for the treatment of cancer. Semin Canc Biol 2012; 22: 3-13. 42. Postow MA, Callahan MK, Wolchok JD. Immune checkpoint blockade in cancer therapy. J Clin Oncol 2015; 33: 1974-82. 43. Sliwkowski MX, Mellman I. Antibody therapeutics in cancer. Science 2013; 341: 1192-8. 44. AlDeghaither D, Smaglo BG, Weiner LM. Beyond peptides and mAbs--current status and future perspectives for biotherapeutics with novel constructs. J Clin Pharmacol 2015; 55 Suppl 3: S4-20. 45. Wu AM, Olafsen T. Antibodies for molecular imaging of cancer. Cancer J 2008; 14: 191-7. 46. Verel I, Visser GW, Boerman OC, van Eerd JE, Finn R, Boellaard R, et al. Long-lived positron emitters zirconium-89 and iodine-124 for scouting of therapeutic radioimmunoconjugates with PET. Cancer Biother Radiopharm 2003; 18: 655-61. 47. Vosjan MJ, Perk LR, Visser GW, Budde M, Jurek P, Kiefer GE, et al. Conjugation and radiolabel- ing of monoclonal antibodies with zirconium-89 for PET imaging using the bifunctional chelate p-isothiocyanatobenzyl-desferrioxamine. Nat Protoc 2010; 5: 739-43. 48. Nayak TK, Brechbiel MW. Radioimmunoimaging with longer-lived positron-emitting radionu- clides: potentials and challenges. Bioconjug Chem 2009; 20: 825-41. 49. Boerman OC, Oyen WJ. Immuno-PET of cancer: a revival of antibody imaging. J Nucl Med 2011; 52: 1171-2. 50. Knowles SM, Wu AM. Advances in immuno-positron emission tomography: antibodies for molecular imaging in oncology. J Clin Oncol 2012; 30: 3884-92. 51. Wright BD, Lapi SE. Designing the magic bullet? The advancement of immuno-PET into clinical use. J Nucl Med 2013; 54: 1171-4. 52. Bai M, Bornhop DJ. Recent advances in receptor-targeted fluorescent probes for in vivo cancer imaging. Curr Med Chem 2012; 19: 4742-58. 53. Polom K, Murawa D, Rho YS, Nowaczyk P, Hünerbein M, Murawa P. Current trends and emerg- ing future of indocyanine green usage in surgery and oncology: a literature review. Cancer 2011; 117: 4812-22. 54. Ito S, Muguruma N, Hayashi S, Taoka S, Bando T, Inayama K, et al. Development of agents for reinforcement of fluorescence on near-infrared ray excitation for immunohistological staining. Bioorg Med Chem 1998; 6: 613-8. 55. Yang Y, Zhang Y, Hong H, Liu G, Leigh BR, Cai W. In vivo near-infrared fluorescence imaging of CD105 expression during tumor angiogenesis. Eur J Nucl Med Mol Imaging 2011; 38: 2066-76. 56. Vermeulen JF, van Brussel AS, Adams A, Mali WP, van der Wall E, van Diest PJ, et al. Near- infrared fluorescence molecular imaging of ductal carcinoma in situ with CD44v6-specific antibodies in mice: a preclinical study. Mol Imaging Biol 2013; 15: 290-8. 57. Day KE, Sweeny L, Kulbersh B, Zinn KR, Rosenthal EL. Preclinical comparison of near-infrared- labeled cetuximab and panitumumab for optical imaging of head and neck squamous cell carcinoma. Mol Imaging Biol 2013; 15: 722-9. 58. Bhattacharyya S, Patel N, Wei L, Riffle LA, Kalen JD, Hill GC, et al. Synthesis and biological evaluation of panitumumab-IRDye800 conjugate as a fluorescence imaging probe for EGFR- expressing cancers. Med Chem Commun 2014; 5: 1337-46. 59. Korb ML, Hartman YE, Kovar J, Zinn KR, Bland KI, Rosenthal EL. Use of monoclonal antibody- IRDye800CW bioconjugates in the resection of breast cancer. J Surg Res 2014; 188: 119-28. 28 Chapter 1

60. Sampath L, Kwon S, Hall MA, Price RE, Sevick-Muraca EM. Detection of cancer metastases with a dual-labeled near-infrared/positron emission tomography imaging agent. Transl Oncol 2010; 3: 307-217. 61. Zhang Y, Hong H, Engle JW, Yang Y, Barnhart TE, Cai W. Positron emission tomography and near-infrared fluorescence imaging of vascular endothelial growth factor with dual-labeled bevacizumab. Am J Nucl Med Mol Imaging 2012; 2: 1-13. 62. Zhang Y, Hong H, Severin GW, Engle JW, Yang Y, Goel S, et al. ImmunoPET and near-infrared fluorescence imaging of CD105 expression using a monoclonal antibody dual-labeled with (89) Zr and IRDye 800CW. Am J Transl Res 2012; 4: 333-46. 63. Hong H, Zhang Y, Severin GW, Yang Y, Engle JW, Niu G, et al. Multimodality imaging of breast cancer experimental lung metastasis with bioluminescence and a monoclonal antibody dual- labeled with 89Zr and IRDye 800CW. Mol Pharm 2012; 9: 2339-49. 64. Zhang Y, Hong H, Orbay H, Valdovinos HF, Nayak TR, Theuer CP, et al. PET imaging of CD105/ endoglin expression with a 61/64Cu-labeled Fab antibody fragment. Eur J Nucl Med Mol Imaging 2013; 40: 759-67. 65. Hall MA, Pinkston KL, Wilganowski N, Robinson H, Ghosh P, Azhdarinia A, et al. Comparison of mAbs targeting epithelial cell adhesion molecule for the detection of prostate cancer lymph node metastases with multimodal contrast agents: quantitative small-animal PET/CT and NIRF. J Nucl Med 2012; 53: 1427-37.