Page | 0
6-14-2021
Division of Toxicology and Drug Safety
Department of Pharmaceutical Biosciences
Faculty of Pharmacy
Uppsala University
The effect of
deactivation or silencing
of tumor stroma with
angiogenesis inhibitor
on malignancy of tumor
metastases
Nataly Tachijian
Degree Project in Toxicology, 30.0 c, Spring semester 2021
Supervisor: Faranak Azarbayjani, Docent
Page | 1
A
BSTRACT
Background: Neuroblastoma (NB) is a pediatric tumor in infants and young children. The survival rate is only around 50 percent for high-risk NB despite advanced and intense multi-modal therapy. Current research aims to find new effective treatment additional to modern therapy to improve prognosis of high-risk NB in children. As such, SU11248 may be a valuable approach for improving treatment and survival as growth factors have crucial roles in tumor growth, angiogenesis, and metastasis.
Aim: The aim of this investigation was to examine tissues from SU11248 treated and nontreated tumor-bearing animals on the abundance of tumor-associated macrophages (TAMs) in metastases found on vital organs. Our hypothesis is that if SU11248 could cause “deactivation” or “silencing” of the stroma of metastases particularly by acting on stromal immune cells such as TAMs.
Methods: Paraffin-embedded metastases developed in an orthotopic xenograft model in beige SCID mice were stained with a monoclonal rat anti-mouse antibody as a marker of TAMs. Morphological analysis of tissue slides, and macrophage quantification was performed using a microscope. Statistical analysis was achieved using an unpaired two tailed t-test.
Results: Macrophages were stained nicely, but the number of macrophages in the metastases were not statistically different between the vehicle treated controls and SU11248 treated metastases.
Conclusion: In patients with high-risk NB, SU11248 may be a useful therapeutic supplement. We believe that further research into mechanisms that target critical factors for angiogenesis and
metastasis in NB, such as TAMs, is an important step toward improving patient outcomes in high-risk NB.
Page | 2
T
ABLE OF CONTENTS
Abstract ... 1
1 Introduction ... 4
2 Angiogenesis and cancer ... 4
2.1 Angiogenic factors ... 4
3 Cancer in children ... 5
3.1 Neuroblastoma ... 5
3.1.1 Classification and survival rate ... 6
3.1.2 Angiogenic factors in high-risk NB ... 6
3.1.3 Treatment and sequelae of modern therapy ... 7
3.1.4 New treatment models ... 7
4 Angiogenesis inhibitors ... 8
4.1 Tumor stroma ... 8
4.1.1 Tumor-associated macrophages of the tumor stroma ... 9
5 Orthotopic xenograft tumor model ... 10
6 Aim ... 12
7 Materials and methods ... 12
7.1 Antibodies ... 12 7.2 Reagents ... 13 7.3 Paraffin-embedding ... 13 7.4 Sectioning ... 14 7.5 Immunohistochemistry ... 14 7.6 Method optimization ... 15 7.6.1 Refinement of sectioning ... 15 7.6.2 Improvement of IHC ... 16
7.6.3 Detection of metastatic regions ... 16
7.7 Quantification of macrophages and statistical analysis ... 17
8 Results ... 18
Page | 3 8.1.1 Background staining ... 18 8.1.2 Metastasis detection... 19 8.2 Morphological analysis ... 20 8.2.1 Spleen ... 20 8.2.2 Liver ... 21 8.3 Quantitative analysis ... 23 9 Discussion ... 24
10 Limitations of the present study ... 27
11 Conclusion ... 27
12 Relevance for future career ... 28
13 Ethical aspects ... 28
14 Acknowledgements ... 28
Page | 4
1 I
NTRODUCTION
Cancer is one of the most common cause of death. In 2018, cancer was responsible for 9.6 million deaths globally (1). It is estimated that at least every third person will develop cancer during the lifetime making the disease very common. Cancer is associated with genetic mutations of the DNA and arises from the conversion of normal cells to cancer cells in a multistage process (2). Inflammation and genome instability promote and accelerate the acquisition of the six hallmarks of cancer which are required for cancer transformation and stimulation of tumor growth. These include sustained
proliferative signaling, apoptotic and growth suppressor resistance, activation of invasion,
angiogenesis induction, and metastasis (3). Physical, chemical and biological carcinogens are three external agents that interact with the individual’s genetics and give rise to these modifications (1). However, there are yet no known lifestyle-related or environmental risk factors for most types of childhood cancers, including neuroblastoma (4).
2 A
NGIOGENESIS AND CANCER
Through the secretion of cytokines, growth factors, and proteases, tumor cells and immune cells communicate continuously during all stages in tumor development, from tumor initiation to metastatic expansion. The first phase of cancer involves the acquisition of mutations, which leads to a harmless non-lethal in situ tumor. The second phase entails transition to an angiogenic phenotype with
continuous “sprouting” of new blood vessels around the tumor, i.e. angiogenesis, which enables access of vasculature for the tumor to expand and to provide it with oxygen and nutrients (5,6). Angiogenesis is usually a necessary prerequisite and a key factor for developing malignant disease. The tumor must have blood vessels in order to be clinically detected, and without them it cannot metastasize (7). Metastasis occurs at the clinical point at which an avascular solid tumor has converted to a vascular phase after which the neoplastic mass can expand indeterminately in situ and spread to distant sites (8). It is the main cause of cancer-related deaths and has a big negative impact on the clinical outcome and therapy (7). One of the most common solid tumors in childhood cancer is neuroblastoma which is a tumor with high vascularity and of which metastases are present in 65 percent of the cases at the time of diagnosis (9). Preclinical models of neuroblastoma that can metastasize are thus suitable models for studies on drugs that inhibit neovascularization and metastasis (see section 5).
2.1 A
NGIOGENIC FACTORSPage | 5 anti-angiogenic protection. This angiogenic switch is correlated with advanced tumor stage and
metastatic disease (5,11). By the influence of angiogenic factors’ interactions, neovascularization, basement membrane breakdown, and remodeling of the extracellular matrix occur which allow tumor cells to expand and penetrate into neighboring tissues and metastasize. A traitorous expedition of tumor cells through the vasculature is thus entailed as metastasis occur, fostered by close association with macrophages which will be discussed later (see section 4.1.1) (12).
3 C
ANCER IN CHILDREN
Childhood cancer differs from cancer in adults. Tumor growth and severity is usually faster and more aggressive in children compared to cancer in adults (2). The cause of most childhood cancers is usually not known, however existing data suggest that genetics prejudices the outcome in 10 percent of all childhood cancers. Neuroblastoma is the most frequent cancer of solid tumors in children, while other common categories include leukemias, brain cancers and lymphomas (13).
3.1 N
EUROBLASTOMANeuroblastoma (NB) is one of the most prevalent pediatric extracranial solid tumors in infants and young children and may even develop in fetuses (2,9). It accounts for about 10 percent of all childhood tumors (14). In Sweden, around 20 children are diagnosed with NB each year, with the majority of children being under the age of two years, with a median age of 18 months at the time of diagnosis (2,15). It originates from special cells in the embryo called neural crest cells which are located at dorsalmost region of the neural tube. These cells migrate from neural tube to distant locations as the adrenal medulla or paraspinal ganglia to form from the developing sympathetic nervous system. This explains the most frequent location of NB at these sites. NB could however also appear from the neck, chest, abdomen, and to the pelvis as a solid mass (2,9). Neoplastic lesions can occur by changes in the NCC specification, deregulated migration, and cell differentiation, which can ultimately lead to NB. The exact causes of NB development from the neural crest are still not
completely understood (16).
One specific associated gene that is involved in NCC development include the N-myc protooncogene, also known as MYCN, which has been linked to oncogenic activity in human NB for decades (15,17). MYCN amplification is highly associated with severe stages of disease and is important in NB
Page | 6 percent of patients with NB have amplified MYCN (18). Moreover, it has been shown to be
overexpressed in the post-migratory neural crest of the precursor cells during sympathoadrenal development. Studies in zebrafish with forced MYCN expression have also confirmed NB development (16,19).
3.1.1 Classification and survival rate
Patients with NB are classified into different risk groups depending on their combination of prognostic factors, such as age, tumor histology, MYCN amplification, and the clinical stage. The International Neuroblastoma Risk Group (INRG) classification system divides the patients into the risk groups; very low, low, intermediate, and high risk. High-risk patients typically include children older than 18 months and those with four copies of MYCN (20). High risk-NB is associated with unfavorable histology, MYCN amplification, metastasis, and poor outcome, which is frequently accompanied by a high relapse rate. Low-risk NB lacks MYCN amplification and does not progress to advanced disease and will not acquire increased expression of the protooncogene either. MYCN amplification thus appears to be an early occurrence in high-risk NB tumorigenesis (18). The risk groups are then, based on the level of the disease and image-defining risk factors (IDRFs), classified into substages which describe the tumor localization and if metastasis is present (20).
NB is responsible for 15 percent of all cancer-related deaths of the pediatric population (14). The survival rate of NB varies with the outcome of disease and risk group. Despite a two-decade increase in the 5-year survival rate, defined as the percentage of patients who live for 5 years after diagnosis, approximately 2 of 3 children with NB already have metastasized disease at the time of diagnosis (21– 23). The survival rate is only around 40 to 50 percent for high-risk NB which signifies the low
survival despite advanced and intense multi-modal therapy. Contrariwise, the corresponding proportion for the low-risk population is more than 95 percent survival (9,24). This underscores the essentiality of finding new methods for treating high-risk NB (8,25).
3.1.2 Angiogenic factors in high-risk NB
NB is a tumor of high vascularity. Therefore, angiogenic factors are frequently overexpressed in a tumor of high-risk NB, in particular vascular endothelial growth factor (VEGF), platelet derived growth factor (PDGF), and their corresponding receptors (7,10). VEGF-A is considered the key factor in angiogenesis and is correlated with advanced tumor stage, maintaining the tumor vascular system, and for tumor development and differentiation (26). VEGF-A is normally produced via the
Page | 7 growth and survival (27). As a result, treating advanced stage NB by simultaneously targeting several crucial growth factors may be a valuable approach for improving treatment (see section 4).
3.1.3 Treatment and sequelae of modern therapy
There are different treatment methods for treating NB including surgery, radiotherapy, and chemotherapy, which the former composes the backbone of treatment. Treatment may differ
depending on the INRG classification and stage in that metastasized NB requires multi-modal therapy and is the most difficult to treat, while a localized tumor is sufficiently treated by surgery (21). High-risk NB treatment constitutes the three phases induction, consolidation, and maintenance therapy lasting for about 18 months in total (20). Surgical resection is attempted after chemotherapy and tumor shrinkage to reduce the risk of surgical complications. The following phase, consolidation, consists of myeloablative chemotherapy and autologous stem cell transplantation (ASCT), followed by radiation therapy, to eliminate the remains of diseased cells. Any residual disease is then treated in the
maintenance phase with differentiating agent isotretinoin (vitamin A) or with immunotherapy to reduce the risk of relapse (20). However, treatment resistance become more evident indicating that modern treatment is yet inadequate for high-risk NB (11,14).
Pertaining to curing cancer, it is important to take into account long-term side effects when choosing treatment (2). Surgical resection is a risk-taking action that could develop severe and long-lasting complications, such as bowel injury that complicates enteral feeding over a long period of time post-surgery (22). Other long-term sequelae of chemotherapy, radiotherapy and post-surgery among NB survivors include chronic health conditions, including musculoskeletal (e.g. scoliosis, osteoporosis), neurological (e.g. weakness, prolonged pain, movement problems), endocrine (e.g. hypothyroidism, growth hormone deficiency, delayed puberty), and sensory complications (e.g. hearing loss,
blindness), as well as secondary malignancies (28). A high burden of late sequelae is proven in high-risk patients that have survived (29). This highlights the need for improved therapy with decreased risks of developing late side effects.
3.1.4 New treatment models
In majority of cases, NB has already spread to areas outside of the original site at the time of diagnosis. Cancer treatment becomes complicated and surgery impractical (9). The survival rate of these patients is still very poor. Consequently, agents that inhibit tumor progression by interfering with blood vessel formation, i.e., angiogenesis inhibitors are becoming more widely recognized and appear to have a significant effect on the treatment of NB in preclinical research.
Page | 8 important role in tumor progression and expansion apart from the proliferating tumor cells. Relapse and resistance are a common problem with current therapies due to selectively affecting proliferating and genetically unstable tumor cells rather than also affecting the cells of the tumor stroma which are involved in regrowth of the surviving tumor cells. Thus, drug screening in tumoral in vitro
monocultures only explains one reason why most cancer drugs fail early since stromal cells usually are not included in these models (30). Therefore, new preclinical models of difficultly treated tumors like NB are crucial as we need accurate models that study the true relationship between tumor cells and the tumor stroma since current treatment of these tumors is yet inadequate.
4 A
NGIOGENESIS INHIBITORS
The intention of angiogenesis inhibitors is to inhibit neovascularization in order to prevent metastasis. By inhibiting angiogenesis, the oxygen level and nutrient accessibility from adjacent vasculature are reduced, leading to restricted tumor growth, alternatively tumor vasculature is normalized which improves the accessibility of other treatments. Since angiogenesis is dependent on many angiogenic factors which are overexpressed in NB, multi-targeting growth factors or their receptors may improve treatment efficacy. For instance, angiogenesis can be moderated by immune neutralizing VEGF with antibodies, or by inhibiting its receptor VEGFR with a multipotent tyrosine kinase receptor (RTK) inhibitor (11). RTKs, particularly VEGFR, FLT3, KIT, and PDGFR, are involved in the regulation of angiogenesis and MYCN stabilization in NB. Many physiological functions such as cell proliferation, survival and migration are mediated by these receptors where their deregulation contributes to tumor development, maintenance, and malignancy (31,32). Inhibition of these receptors would therefore be beneficial in NB treatment.
4.1 T
UMOR STROMAPage | 9 4.1.1 Tumor-associated macrophages of the tumor stroma
Tumor-stimulated inflammation is a recognized attribute of cancers (35). Monocytes, that differentiate into macrophages, are the most infiltrated leukocytes into the TME via recruitment by hypoxia-inducible chemokines like VEGF-A (figure 1). Macrophages play a central role in cancer-related inflammation as part of their anti-tumoral roles (36,37). However, other phenotypic macrophages known as tumor-associated macrophages (TAMs) differ from the classical, proinflammatory macrophages in that they are immunosuppressive and promote tumor growth by regulating tumor progression, metastasis, invasion, and angiogenesis (38,39). Despite the recruitment of anti-tumoral macrophages, most monocytes differentiate to TAMs because the classical M1-phenotype macrophage is reprogrammed or phenotype-switched through polarization to a reactive M2-phenotype shifting the anti-tumoral activity to pro-tumoral activity. TAMs are critical modulators of the TME that entails remodeling of the extracellular matrix which supports tumor growth in response to tumor and stromal cell-produced microenvironmental signals. Hence, the abundance of TAMs that exhibit the M2-phenotype in a tumor is clinically correlated with poor prognosis and poor patient outcome. (34,37,38,40–42). Depending on how highly developed the tumor has acquired, TAMs activation status and functional role varies alongside the TME differentiation (42).
Figure 1. Schematic illustration of VEGF activity in NB. TAMs secrete VEGF in the TME and recruit
monocytes into the growing tumor which differentiate into TAMs. The proangiogenic activity of VEGF causes stimulation of angiogenesis, while TAMs support tumor growth and survival. The abundance of TAMs and VEGF within the tumor induces the angiogenic switch causing an intensive intravasation of tumor cells to the vasculature and thus metastasis can initiate (43).
Page | 10
5 O
RTHOTOPIC XENOGRAFT TUMOR MODEL
Concerning the need for suitable preclinical models for high-risk NB of new treatment modalities targeting the tumor stroma, in a previous study “DF05,” an orthotopic model of NB in SCID beige mice was developed, on which this study will be based. In orthotopic models, human-derived cancer cell lines are injected in situ of the primary site or tumor origin in animal models which refers to as xenografting. The advantage of orthotopic models is that they mimic clinical disease in terms of tumor location, development, and spread, of which the former is valuable for evaluating the therapy effect in patients with metastatic disease (11). The drawback is that tumor volume cannot be measured of orthotopic tumors compared to subcutaneous models, if not measured with ultrasound immediately prior to autopsy (27). In comparison to subcutaneous models, they lack metastasis and because cancer cell lines are injected subcutaneously the developing tumor would not represent the primary tumor site of NB, making clinical evaluation of treatment efficacy more difficult (11).
In the DF05 study, MYCN-amplified, 1 p-deleted human NB cells (IMR-32 cell line), derived from a 47+XY karyotype abdominal tumor in child, was used for xenografting the left adrenal gland of immunodeficient mice. Selection of species was based on that human tumor cells exhibits higher rates of metastasis in SCID mice than in nude mice. The beige mutation also causes immunodeficiency which reduces the activity of NK (natural killer) cells, allowing for higher metastatic spread. All animals developed a palpable tumor nine weeks after xenografting (figure 2). A group of these tumor-bearing animals were either treated with 40 mg/kg daily of angiogenesis inhibitor SU11248 for 23 days or its vehicle (11). SU11248, sunitinib malate (Sutent®, Pfizer Inc), is a small-molecule drug
currently approved in different adult cancers which has been studied for its effect on NB owing to its multi-targeting of VEGFR, PDGFR, KIT, and FLT3 (11). In the DF05 study, SU11248 showed reduced orthotopic high-risk NB growth by approximately 80%, and a significant decrease in the viable tissue fraction, compared to controls. It has also showed increased levels of tumor cell
Page | 11 Figure 2. Orthotopic tumors at autopsy. Tumor-bearing mice were treated with vehicle (control) or
with SU11248 (40 mg/kg/day p.o.) for 23 days. Note the pale appearance of the SU11248 treated tumor. T = tumor, RK = right kidney, LK = left kidney; arrows indicate the location of the normal right adrenal gland (27).
Table 1. Metastatic spread in orthotopic NB xenografts of SCID beige mice treated with SU11248 (40
mg/kg/day) or vehicle (27). In total, 25% of treated animals and 44% of control animals had metastases.
SITE CONTROL SU11248
ANIMALS WITH METASTASES 100% (9/9) 71% (1/7)
LUNG 0% (0/9) 14% (1/7)
LIVER 44% (4/9) 57% (4/7)
SPLEEN 56% (5/9) 14% (1/7)
BONE MARROW 78% (7/9) 14% (1/7)*
*p<0.05; (compared with control); Fisher’s exact test.
The significant reduction in tumor volume by 80% cannot only be explained by insignificant (30%) reduction in angiogenesis. Therefore, we intended to investigate these orthotopic tumors further, focusing on the malignancy of metastases in terms of the effect of SU11248 on the stroma of tumor metastases, and whether the interactions between tumor cells and stromal cells can explain this outcome. We will be focusing on the presence of TAMs in the stroma as they have crucial roles in tumor development, angiogenesis, and metastasis.
SU11248
control
control
day 0
day 23
day 23
RK
LK
T
RK
LK
RK
LK
T
T
SU11248
control
control
day 0
day 23
day 23
SU11248
control
control
day 0
day 23
day 23
Page | 12
6 A
IM
Previous preclinical studies conducted in the angiogenesis group at Uppsala University has shown that SU11248 can halt the growth of neuroblastoma in an orthotopic model of this malignancy in mice (27). The aim of the current investigation was to continue and examine tissues from the same tumor-bearing animals, focusing on studying the abundance of tumor-associated macrophages in the
metastases found on vital organs. These metastasis-containing organs were retrieved from the animals at the termination of the above-mentioned study. Our hypothesis was that if SU11248 could cause “deactivation” or “silencing” of the tumor stroma particularly by acting on stromal immune cells such as tumor-associated macrophages.
7 M
ATERIALS AND METHODS
The biopsies from the tumor-bearing animals that were examined in this study have been acquired from a previous study by Dieter Fuchs (2005). Tissue blocks of the biopsies were prepared in advance with formalin fixation and paraffin-embedding (FFPE) (11). These will be referred to as “DF05” FFPE blocks. Each block was labeled as "Orto # organ met", which refers to the orthotopic tumor followed by an animal identity number and metastasized organ.
7.1 A
NTIBODIESHistological detection of TAMs was conducted by using 1:10 diluted monoclonal anti-F4/80 antibody. Undiluted secondary antibody was used according to the company's recommendations and was served as a probe to locate the primary antibody (table 2), followed by a horseradish peroxidase (HRP) polymer which binds to the probe. Furthermore, to detect the metastasis and localize NB cells of the specimen, an anti-CgA (chromogranin A) antibody was used as primary and anti-mouse HRP as secondary antibody (see table 2 and section 7.6.3).
Table 2. Primary and secondary antibody applied in F4/80 and CgA immunohistochemistry.
ANTIBODY SPECIES CONCENTRATION DILUTION PROVIDER ART. NO
Page | 13 Goat anti-mouse HRP 1 mg/ml 1:200 Abcam, Cambridge, U.K ab97240
7.2 R
EAGENTSSeveral reagents were used in immunohistochemistry (IHC) to achieve immunostaining without interfering factors. Table 3 summarizes each reagent's role of the IHC procedures.
Table 3. Procedures and materials used for immunohistochemistry.
PROCEDURE/ROLE REAGENTS PROVIDER
ANTIGEN RETRIEVAL Proteinase K Abcam, Cambridge, U.K
REMOVAL OF ENDOGENOUS PEROXIDASE
3.0% H2O2 solution diluted in TBST
Merck & Co, New Jersey, U.S.A
BLOCKING REAGENT Background sniper HistoLab Products AB,
Gothenburg, Sweden
POLYMER (LINKED WITH PEROXIDASE)
Rat-on-mouse HRP-polymer kit “polymer”
Biocare Medical, California, U.S.A
CHROMOGEN (CONTAINING HYDROGEN PEROXIDE)
Betazoid DAB kit Biocare Medical, California, U.S.A
COUNTERSTAIN Hematoxylin HistoLab Products AB,
Gothenburg, Sweden
MOUNTING Biomount BioGnost, Zagreb, Croatia
7.3 P
ARAFFIN-
EMBEDDINGBecause this thesis is a supplementary study of previous research (DF05), fixation, dehydration, clearing, and infiltration with paraffin was already prepared when the biopsies were initially removed. In addition, because the DF05 FFPE blocks had been stored for a long time before current study, the blocks should be re-embedded in new paraffin as old paraffin can be difficult to manage while sectioning. Embedding is important for the preservation of tissue morphology, and it facilitates the management of the biopsies during sectioning in order to easily cut fine sections. For re-embedding, the biopsy was transferred to a new mold of appropriate size after melting the old paraffin. The mold was filled using HISTOWAXTM paraffin (HistoLab Products AB, Gothenburg, Sweden) until the
Page | 14
7.4 S
ECTIONINGFor sectioning, we used the rotation microtome Leica RM255 (Leica Biosystems, Wetzlar, Germany). From each DF05 FFPE block, a minimum of four sets of 4μm thick sections was prepared. Section creases were stretched with two thin brushes carefully, then transferred to a glass slide with distilled water to allow creases to stretch out for a minute. The sections were after that transferred to a 43°C water bath for an additional 5–15-minutes stretching, and transferred to a labeled glass slide, dried out excess water with a paper towel, and placed on a heated plate to allow the remaining water to
evaporate. All tissue specimens were stored in a 37°C cabinet overnight after which they were stored in the refrigerator.
7.5 I
MMUNOHISTOCHEMISTRYImmunohistochemistry is a method that is used in histology for morphological examination using light microscopy and for detecting specific antigens in tissue preparations using antibody markers (44). In this study, IHC was applied on the macrophage-specific protein F4/80.
FFPE slides were deparaffinized by dissolving the paraffin in xylene 2x5 minutes (Solveco AB, Stockholm, Sweden), following a down-scaled hydration (99.5%, 95%, 70%) with ethanol (Solveco AB) 2x2 minutes, and washed with distilled water and TBST pH 7.6 (50 mM Tris, 150 mM NaCl, 0.05% Tween 20), respectively. Antigen retrieval was achieved by adding 100 μl undiluted proteinase K (Abcam, Cambridge, U.K) on the slides and incubating them for 4 minutes. Removal of endogenous peroxidase was achieved by incubating the slides in a TBST-diluted 3.0% H2O2 bath (Merck & Co,
New Jersey, U.S.A) for 20 minutes. To reduce background staining, two drops of the blocking reagent Background sniper (HistoLab Products AB) was added on the slides and was incubated for 10
Page | 15 Table 4. Overview of F4/80 immunohistochemistry. Between each incubation, the slides were placed
in a cuvette with TBST for 2x5 minutes.
IHC step Material Incubation time
Deparaffinization Xylene 2x5 min
Abs ethanol 2x2 min
95% ethanol 2x2 min
70% ethanol 1 min
Distilled water wash
TBST wash
Antigen retrieval Proteinase K 4 min
Removal of endogenous peroxidase 3.0% H2O2 solution 20 min
Blocking Background sniper 10 min
Primary antibody Rat anti-mouse F4/80 30 min
Secondary antibody Rat-on-mouse HRP “probe” 10 min
Polymer Rat-on-mouse HRP “polymer” 10 min
Chromogen Betazoid DAB 5 min
Counterstain Hematoxylin 3 min
Tap water 20 min
Dehydration 70% ethanol wash
95% ethanol 2 x wash 99.5% ethanol 2 x wash
Xylene 2 x wash
Mounting Biomount overnight
7.6 M
ETHOD OPTIMIZATIONA significant portion of this study was devoted to method optimization. Adjustments of the methods was attempted several times in order to refine sectioning, improve results for IHC, and confirm NB cells with CgA IHC.
7.6.1 Refinement of sectioning
Page | 16 microtome to stretching water and glass slides. Unfortunately, all FFPE blocks were sectioned before we discovered that incubating them in 10% glycerol and 70% ethanol solution for 5 minutes before sectioning facilitated the performance much better than the ice block. However, due to the lack of time, no new sections were prepared unless otherwise mentioned (see section 7.6.3).
7.6.2 Improvement of IHC
Antigen retrieval is an important step of IHC in order to enzymatically expose the antigen. However, other cellular components can also be exposed which can interfere with the immunoreaction. For instance, the chromogen cannot distinguish the peroxidase activity of the polymer from endogenous peroxidase activity of the specimen. Therefore, if endogenous peroxidase is present and not removed with a H2O2 solution, unspecific staining would appear. The manifestation of endogenous peroxidase
was further inhibited by increasing the H2O2 concentration by 10-fold. We also attempted to reduce the
incubation time of proteinase K to reduce retrieval of non-specificity. Table 5 summarizes our attempts for IHC optimization.
Table 5. Attempts to reduce non-specificity. Method optimization was focused on reducing nonspecific
staining with emphasis on improving the procedures of removal of endogenous peroxidase (EP) and antigen retrieval. ATTEMPT COMMENTS I:A Removal of EP w/o proteinase K w/ vs. w/o 3.0% H2O2
Proteinase K appeared to expose non-specificity, as both slides from this attempt were negative due to the lack of proteinase K. Thus, the
proteinase is required for retrieval in order for DAB to react with HRP and produce a stain. There was no way to comment on the impact of increased H2O2 concentration in this attempt.
I:B Removal of EP w/ 7 min proteinase K w/ vs. w/o 3.0% H2O2
Since attempt I:A did not result in any staining, a second attempt was done which included 7 min incubation of the specimen with proteinase K. Liver NC was negative as previously predicted. Spleen NC still had nonspecific staining, but the occurrence was not as high compared to the original method which means that spleen contains a lot of EP. This attempt showed that 3.0% H2O2 reduced non-specificity more than the original concentration of 0.3%. II Antigen retrieval w/ 4 min vs. 7 min proteinase K w/ 3.0% H2O2
Because there still appeared nonspecific staining of spleen from attempt I:B, the incubation time for proteinase K was reduced to control its exposure of EP. Considering choosing a sufficient amount of time to expose the desired antigen, we were recommended to try incubating proteinase K for 4 minutes instead of 7 minutes. Liver NC was negative, as predicted, while spleen NC was somewhat positive, however much less nonspecific compared to the original method.
7.6.3 Detection of metastatic regions
Page | 17 In attempt 1, we cut through 20 sections and used the 21st section in dark field microscopy to examine
if any deviant structures were visible. However, nothing was observed, which also was the case in attempt 2 where we tried to examine the section only by HTX staining without first deparaffinizing the slide. In attempt 3, deparaffinization failed due to the tissue coming off the slide, most likely because it was dried insufficiently on the heated plate. Instead, in attempt 4, we added another 20 minutes in 60°C cabinet before deparaffinization to allow the tissue to adhere. Subsequently, the slide was stained with HTX for 3 minutes, blued for 20 min and then stained with eosin for 20 seconds (HE). Slides were quickly rinsed in 70% alcohol up to xylene and mounted.
Except for the normal tissue cells, which we could identify by their characteristics, we were at first uncertain whether the abnormal cells we observed were tumor cells or not. Clinical tracing of NB was attempted using the positive tumor marker chromogranin A (CgA). It was used to confirm the
observed deviant regions as tumors of NB, as previously assumed with HE stained slides. CgA IHC was performed on specimens from the same biopsies using the same protocol as previous IHC except for antigen retrieval, the slides were incubated in 0.1M citrate buffer pH 6.0 for 15 minutes in a 98°C water bath. CgA was observed using mouse anti-CgA (clone LK2H10, Abcam) applied at 1:50 as primary antibody and 1:200 goat anti-mouse HRP (Abcam) as secondary antibody. Both antibodies were incubated 30 minutes. Omission of the primary antibody served as NC. Slides were incubated with DAB for 15 minutes.
7.7 Q
UANTIFICATION OF MACROPHAGES AND STATISTICAL ANALYSISEach slide was examined histologically by using the Leica DMRBE Trinocular Research Microscope (Leica Biosystems) in order to detect and localize the metastases in the tissue. Microscope images were captured using the digital microscope camera Jenoptik ProgRes® C14 Plus CCD 12.5 M.P
(Jenoptik AG, Jena, Germany).
Page | 18
8 R
ESULTS
8.1 M
ETHOD OPTIMIZATION8.1.1 Background staining
Our original IHC method (retrieval 7 min; endogenous peroxidase removal 0.3% H2O2) resulted in a
positive NC of spleen and negative NC of liver (figure 3). Microscopical observation of the slides with spleen and liver controls showed that our optimized method was successful as non-specificity was satisfactorily reduced.
Figure 3. First attempt for using the original F4/80 IHC protocol. Both images are from control
specimens and were not added primary antibody A. NC of normal mouse spleen, nonspecific background staining is abundant in the splenic red pulp surrounding the lymphatic nodules (white pulp) (×10). B. NC of normal mouse liver, negative (×10).
Regarding the effect of H2O2 for reducing background staining, endogenous peroxidase was much less
visible of the treated tissue, which indicated that the incubation with H2O2 was vital (figure 4).
Figure 4. A. NC of normal mouse spleen untreated with H2O2 (×10). B. NC of normal mouse spleen
Page | 19 Regarding the effect of reducing incubation time with proteinase K, it resulted in less nonspecific staining in the tissue with reduced duration (figure 5).
Figure 5. NC of normal mouse spleen, incubated with proteinase K (A. 7 min; B. 4 min) and treated
with 3.0% H2O2 (×5).
8.1.2 Metastasis detection
Further sectioning of the biopsy blocks and HE-staining resulted in sections with visible metastases and tumor cells. Tumor cells were stained purple with little to no cytoplasm and had small and round to oval nuclei. Normal tissue cells had both bigger cytoplasm (pink) and nuclei, respectively. An example of a liver section is provided in figure 6.
Figure 6. Orto74 liver metastasis, HE stained (×10). Dark purple cell nuclei with little to no
cytoplasm are tumor cells and surrounding cells are normal liver cells (hepatocytes). Note: Tattered section as to why the cells are not entirely distinct.
Page | 20 Figure 7. Orto71 primary tumor, positive control (×40). CgA positive granular cytoplasm (brown)
surrounding HTX stained tumor cell nuclei. Surrounding CgA negative cells are almost certainly undifferentiated NB cells too.
8.2 M
ORPHOLOGICAL ANALYSISMetastases were identified based on patterns of the studied tissue that deviated from normal tissue cells. We discovered various uniform areas of small round to oval shaped cells arranged in clusters, with large cell nuclei and very thin basophilic cytoplasm, as confirmed by NB cell cytomorphology of the undifferentiated type with no rosette formation.
8.2.1 Spleen
A section of F4/80-stained normal mouse spleen was prepared as a comparison control (figure 8) to spleen sections from the DF05 FFPE blocks shown in figure 9 and 10, respectively.
Figure 8. Positive control of normal mouse spleen, F4/80 stained (×20). Positive macrophages
Page | 21 Figure 9. Orto 77 spleen metastasis (treated), F4/80 stained. A. Several white pulp lymph nodes are
clearly distinct from the red pulp (left side of the section), while to the right, fibrous connective tissue contracts the red pulp (×10). B. Abnormal shape of the white pulp (splenic white pulp atrophy).
Figure 10 A+B. Orto 84 spleen metastasis (control), F4/80 stained (×20). Big distinct regions of
fibrosis. Megakaryocytes (cells with giant nuclei) are frequent around fibrotic tissue.
8.2.2 Liver
A section of F4/80-stained normal mouse liver was prepared as a comparison control (figure 11) to liver sections from the DF05 FFPE blocks shown in figure 12 and 13.
A
B
Page | 22 Figure 11. Positive control of normal liver from mouse, F4/80 stained (×10). Brown-stained
macrophages of the liver, Kupffer cells, appear as stretched and elongated (amoeboid-shaped) cells.
Figure 12. Orto 88A liver metastasis (control), F4/80 stained. A. Note the obvious cell border
between the metastasis (blue) and normal liver tissue (purple) (×5). There are fewer macrophages in the metastasis than in the normal tissue, generally observed. B. Higher magnification of figure A. Brown oblong cells represent macrophages which seem less frequent in the tumor stroma (×20).
Page | 23 Figure 13. Orto 74 liver metastasis, F4/80 stained (treated). Macrophages seem abundantly
decreased around normal liver tissue. The right edge of cells with dark purple nuclei, what seems to be tumor cells, have no macrophage presence at all.
8.3 Q
UANTITATIVE ANALYSISManual quantification by microscope at ×40 and ×100 magnification showed that the number of macrophages present in the metastases for both liver and spleen was, in general, reduced. Nevertheless, statistical analysis with unpaired two tailed t-test showed insignificant difference between the control group and the treated group (table 6 and figure 14).
Table 6. Quantified macrophages in selected specimens. Tumor i.d. refers to the identification number
of the tumor-bearing animal which received either SU11248 or vehicle (control). Sites refers to the (three) different placements of the grid in the metastatic regions of the tissue while counting the macrophages.
TUMOR I.D TREATMENT TISSUE SITE 1 SITE 2 SITE 3 AVERAGE SITE 1-3
ORTO 67 Control Liver 5 5 7 5,67
ORTO 71 Control Spleen 10 9 12 10,33
ORTO 73 Control Spleen 5 4 2 3,67
ORTO 76 Control Liver 2 4 3 3,00
ORTO 78 Control Liver 1 1 2 1,33
ORTO 81 Control Liver 12 10 13 11,67
ORTO 84 Control Spleen 15 16 14 15,00
ORTO 88 Control Liver 4 3 2 3,00
ORTO 70 SU11248 Liver 4 3 6 4,33
ORTO 74 SU11248 Liver 2 3 4 3,00
ORTO 82 SU11248 Liver 2 3 4 3,00
Page | 24 Figure 14. Diagram illustrating the average number of counted macrophages for the control and
SU11248 treated groups, respectively, calculated from the average sites 1-3 for each group (table 6). SU11248 treated metastases showed a reduction in the content of macrophages compared to control. Statistical analysis revealed no statistically significant difference between the groups (p-value 0.111>0.05). The error bars represent the variability of the data.
9 D
ISCUSSION
The angiogenesis group at Uppsala University has already shown that SU11248 can halt the growth of NB in an orthotopic model of malignancy in mice (27). Their results revealed a substantial decrease in tumor volume by approximately 80%, which cannot be explained by the insignificant (30%) reduction in angiogenesis. We therefore planned to further examine the impact of SU11248 on tumor stroma in order to better understand the mechanism of anti-tumor effects of the drug. We chose to start our investigation by quantifying TAMs due to their central role in tumor progression.
The results of the current study showed that the number of macrophages in the tumor stroma of metastases retrieved from tumor-bearing mice were lower in the SU11248 treated group compared with the controls, but the difference was not statistically significant. This might be due to the facts that we had a low number of metastases that were found histologically, due to difficulties experienced with sectioning and staining, and therefore not as many metastases could be included in the final results.
Page | 25 These analysis issues contributed to an uneven size distribution among the groups, with more controls than treated animals, which is normally undesirable, especially if the groups are very small. The smaller the sample, the more likely it is not normally distributed which is one of the conditions required in a t-test. The number of macrophages in the control group was more scatter with the mean values not normally distributed whereas the mean values for treated group were more evenly
distributed. The variability in the data and the small number of metastases examined may have
contributed to the low calculated significance level. Furthermore, the chosen statistical test (unpaired t-test) requires equal or close to equal variance in the groups, a condition that was not met, as evidenced by the error bars in figure 14. In the case of having an obliquely distributed quantitative variable, one can use the corresponding ranking method or other non-parametric statistical tests. However, other methods could not be tried due to time constraints. Given those circumstances, it is not surprising that the t-test was not significant.
Despite the statistically insignificant difference in the number of macrophages between the groups, visual examination of the histological sections of vital organs containing metastases showed a more massive invasion of abnormal cells in these organs in the control groups. In order to verify the type of these cells a preliminary staining with the neuroendocrine marker chromogranin-A (CgA) was conducted and the results showed positive granular cytoplasm of these cells indicating NB origin. However, adjacent cells of the same size and shape could also be negative, implying that the marker was not specific for all NB cells. Furthermore, because there are fewer tumor cells in a metastasis, finding CgA-expressing tumor cells would be even more difficult. We therefore decided not to perform CgA IHC on the metastases.
Moreover, we found that the occurrence of megakaryocytes, particularly in spleen, was more frequent in the control group. Megakaryocytes are very rare and giant cells of the bone marrow and can also be present in the liver and spleen under pathological circumstances such as cancer. They are precursors of platelets which tumor cells interact with during metastasis. This suggests that megakaryocytes
provides as drivers for metastatic cancer and why megakaryocytes were abundant in control groups (45).
Page | 26 A similar study conducted by Priyank Kumar investigated the effect of SU11248 on tumor stroma by using the presence of macrophages as a marker in the primary tumors instead of metastases in the vital organs, but the tissues examined in both studies were obtained from the same tumor-bearing animals. In both studies treatment of tumors with a multitargeted Class III/V tyrosine kinase inhibitor
(SU11248) caused a reduction in number of macrophages in both tumors and metastases in the vital organs (46).
The reduction in the number of macrophages can be due to multiple reasons with different
consequences for the growth and spread of tumors. First, the very pronounce increase in apoptosis previously shown in the DF05 study indicate that the number of viable tumor cells are decreased. Decrease in the number of tumor cells could cause a decrease in VEGF production leading to
decreased recruitment of macrophages into the TME. On the other hand, angiogenesis inhibition and the high percentage of apoptosis associated with it could lead to hypoxia, which could act as a key role in the control of macrophage recruitment and M2 phenotype-switch. VEGF expression is increased under hypoxic stress and the recruitment of macrophages into the TME would have to increase (47,48). However, this is not consistent with results from previously mentioned studies (46). Another plausible theory to explain the findings in our studies (both DF05 and current study) is that inhibiting angiogenesis with SU11248 could cause a normalization of vasculature. Normal blood vessels have pericyte coverage, which allows the drug to reach the tumor more effectively. Pericytes help to stabilize blood vessels, promote endothelial cell survival, and regulate blood flow. It has been
demonstrated that tumor blood vessels with no pericyte coverage are more reliant on VEGF signaling for endothelial cell survival, and that inhibiting VEGF leads to normalization of the tumor vasculature by increasing pericyte coverage, resulting in limited metastasis (49).
Tumor stroma normalization have received more evidence that it may be beneficial as a strategic additional anti-cancer therapy in NB. There are different targets of the angiogenic process to affect, therefore angiogenesis inhibitors can be categorized as direct inhibitors that targets vascular
endothelial cells or indirect inhibitors that either inhibit synthesis of angiogenic growth factors, block receptor signaling, or by reducing the activity of angiogenic inducers (10).Suggested TAM-based strategies include angiogenesis inhibition, immune checkpoint inhibition and phagocytosis of the macrophages including TAM depletion and TAM recruitment inhibition, as well as TAM
reprogramming. Selective inhibition of M2-TAMs is a promising anti-cancer approach (36,40,41). Regarding the anticancer activity of SU11248, it is either exerted directly on tumor cells or through anti-angiogenic mechanisms targeting endothelial and stromal cells which affects the tumor
Page | 27
10
L
IMITATIONS OF THE PRESENT STUDY
Physical examination at autopsy at the time the organs were removed suggested that metastasis should be presented in all the samples. It was however difficult to find and locate these metastases upon slicing the tumors for our histological examinations. Furthermore, the microtome which was used was not sectioning the tissues properly, which made the sections uneven, thick, or compact, and easily broken. We did not expect this to lead to a major problem for analysis and we sectioned all blocks before realizing that analysis of these badly sliced tissues was not easy and made the detailed analysis of single cells difficult in many regions of the sections. Analysis was thus difficult, but we managed to quantify macrophages in some reasonable sections. For most of the sections, we only analyzed them morphologically.
We consulted Dr. Anca Dragomir who is a clinical pathologist at Uppsala University Hospital for expert opinion in interpreting our results. Her recommendation was to redo the sectioning of the organs and complete the analysis of macrophages using multiple staining with different markers, a task that was impossible for us due to the limitations in time for the project and lack of economic resources.
Preclinical studies are often expensive and time consuming. The equipment needed should function properly in order to save time and the valuable tissues retrieved from the main study. If I were to repeat the experiments, I would check the equipment properly in advance. I would also schedule my time more efficiently which was a major obstacle due to corona restrictions in the lab. Staining for different components of the tumor stroma is a trial-and-error procedure which requires patience and hard work and might not always lead to satisfactory results. However, this investigation was not completely in vain as it led to valuable methodological development for future studies.
11
C
ONCLUSION
Page | 28 certainty. We believe that further research into mechanisms that target critical factors for angiogenesis and metastasis in NB, such as TAMs, is an important step toward improving patient outcomes in high-risk NB.
12
R
ELEVANCE FOR FUTURE CAREER
Preclinical studies are crucial in the development of new drug candidates and prior to testing of these drug candidates in humans so called first in human studies. The experiments conducted in this study are considered very relevant in the future career of a pharmacist like me who would like to pursue a career in pharmaceutical companies or regulatory agencies.
13
E
THICAL ASPECTS
Metastases used in this study was retrieved from earlier studies in tumor bearing mice, hence ethical permission was not needed for this study. All the original procedures of xenografting orthotopic tumors and the treatments of growing tumors in animals were approved by the animal welfare authority in Uppsala year 2005.
14
A
CKNOWLEDGEMENTS
I wish to show my appreciation to Faranak Azarbayjani for her dedication in this project and thank her for introducing me to this area of toxicology. I especially appreciate her decisiveness and quick decisions she made throughout her supervision during our discussions, and for her incredible help and feedback on this thesis. I also want to thank Carin Backlin for her great support at lab and motivational discussions she had with me. I appreciate her for her patience with me while I was learning and her assistance when I was not at lab. I want to thank both Azarbayjani and Backlin for being great supervisors. Despite the fact that you both had a full schedule, I appreciate the time you spent helping me on my master’s thesis. It has been a very interesting and instructive project.
Page | 29
15
R
EFERENCES
1. Cancer [Internet]. [cited 2021 Jan 28]. Available from: https://www.who.int/news-room/fact-sheets/detail/cancer
2. Bergman O, Johansson E. Cancer i siffror 2018: populärvetenskapliga fakta om cancer. Socialstyrelsen; 2018.
3. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011 Mar 4;144(5):646–74.
4. Organization ACC. Childhood Neuroblastoma Cancer: Causes, Risk Factors, and Prevention [Internet]. ACCO. 2018 [cited 2021 Jan 29]. Available from:
https://www.acco.org/blog/childhood-neuroblastoma-cancer-causes-risk-factors-and-prevention/ 5. Folkman J, Kalluri R. Cancer without disease. Nature. 2004 Feb;427(6977):787–787.
6. Rajabi M, Mousa SA. The Role of Angiogenesis in Cancer Treatment. Biomedicines. 2017 Jun 21;5(2).
7. Roy Choudhury S, Karmakar S, Banik NL, Ray SK. Targeting Angiogenesis for Controlling Neuroblastoma. J Oncol. 2011 Aug 25;2012:e782020.
8. Ribatti D, Marimpietri D, Pastorino F, Brignole C, Nico B, Vacca A, et al. Angiogenesis in Neuroblastoma. Ann N Y Acad Sci. 2004;1028(1):133–42.
9. Bhoopathi P, Pradhan AK, Bacolod MD, Emdad L, Sarkar D, Das SK, et al. Regulation of neuroblastoma migration, invasion, and in vivo metastasis by genetic and pharmacological manipulation of MDA-9/Syntenin. Oncogene. 2019 Oct;38(41):6781–93.
10. El-Kenawi AE, El-Remessy AB. Angiogenesis inhibitors in cancer therapy: mechanistic
perspective on classification and treatment rationales. Br J Pharmacol. 2013 Oct;170(4):712–29. 11. Fuchs D. Novel treatment modalities for high-risk neuroblastoma: studies in animal models
[Internet] [Doctoral dissertation]. Uppsala University; 2009. Available from: https://www.diva-portal.org/smash/get/diva2:173186/FULLTEXT01.pdf
12. Smith HA, Kang Y. The metastasis-promoting roles of tumor-associated immune cells. J Mol Med. 2013 Apr 1;91(4):411–29.
13. Cancer in Children [Internet]. [cited 2021 Feb 1]. Available from: https://www.who.int/news-room/fact-sheets/detail/cancer-in-children
14. Colon NC, Chung DH. Neuroblastoma. Adv Pediatr. 2011;58(1):297–311. 15. Shohet J, Foster J. Neuroblastoma. BMJ. 2017 May 3;357:j1863.
16. Johnsen JI, Dyberg C, Wickström M. Neuroblastoma—A Neural Crest Derived Embryonal Malignancy. Front Mol Neurosci. 2019 Jan 29;12.
Page | 30 18. Huang M, Weiss WA. Neuroblastoma and MYCN. Cold Spring Harb Perspect Med. 2013
Oct;3(10).
19. Otte J, Dyberg C, Pepich A, Johnsen JI. MYCN Function in Neuroblastoma Development. Front Oncol. 2021 Jan 27;10.
20. Smith V, Foster J. High-Risk Neuroblastoma Treatment Review. Children. 2018 Aug 28;5(9). 21. Neuroblastoma Treatment (PDQ®)–Health Professional Version - National Cancer Institute
[Internet]. 2020 [cited 2021 Jan 24]. Available from:
https://www.cancer.gov/types/neuroblastoma/hp/neuroblastoma-treatment-pdq
22. Tolbert VP, Matthay KK. Neuroblastoma: clinical and biological approach to risk stratification and treatment. Cell Tissue Res. 2018 May 1;372(2):195–209.
23. Neuroblastoma - Childhood - Statistics [Internet]. Cancer.Net. 2012 [cited 2021 May 13]. Available from: https://www.cancer.net/cancer-types/neuroblastoma-childhood/statistics
24. Definition of five-year survival rate - NCI Dictionary of Cancer Terms - National Cancer Institute [Internet]. 2011 [cited 2021 Jan 27]. Available from:
https://www.cancer.gov/publications/dictionaries/cancer-terms/def/five-year-survival-rate 25. Rössler J, Taylor M, Geoerger B, Farace F, Lagodny J, Peschka-Süss R, et al. Angiogenesis as a
target in neuroblastoma. Eur J Cancer. 2008 Aug 1;44(12):1645–56.
26. Bäckman U, Christofferson R. The Selective Class III/V Receptor Tyrosine Kinase Inhibitor SU11657 Inhibits Tumor Growth and Angiogenesis in Experimental Neuroblastomas Grown in Mice. Pediatr Res. 2005 May;57(5 Part 1):690–5.
27. Fuchs D, Christofferson R, Nilsson A, Azarbayjani F. SU11248 as single therapy restricts growth of orthotopic neuroblastoma monitored by contrast-enhanced ultrasound. Uppsala;
28. Laverdière C, Liu Q, Yasui Y, Nathan PC, Gurney JG, Stovall M, et al. Long-term Outcomes in Survivors of Neuroblastoma: A Report From the Childhood Cancer Survivor Study. JNCI J Natl Cancer Inst. 2009 Aug 19;101(16):1131–40.
29. Berthold F, Spix C, Kaatsch P, Lampert F. Incidence, Survival, and Treatment of Localized and Metastatic Neuroblastoma in Germany 1979–2015. Paediatr Drugs. 2017;19(6):577–93.
30. Kock A, Bergqvist F, Steinmetz J, Elfman LHM, Korotkova M, Johnsen JI, et al. Establishment of an in vitro 3D model for neuroblastoma enables preclinical investigation of combined tumor‐ stroma drug targeting. FASEB J. 2020 Aug 1;34(8):11101–14.
31. Calero R, Morchon E, Johnsen JI, Serrano R. Sunitinib Suppress Neuroblastoma Growth through Degradation of MYCN and Inhibition of Angiogenesis. PLoS ONE [Internet]. 2014 Apr 23 [cited 2021 Feb 10];9(4). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3997473/ 32. Maris JM, Courtright J, Houghton PJ, Morton CL, Kolb EA, Lock R, et al. Initial testing (stage 1)
of sunitinib by the pediatric preclinical testing program. Pediatr Blood Cancer. 2008 Jul;51(1):42– 8.
33. Henke E, Nandigama R, Ergün S. Extracellular Matrix in the Tumor Microenvironment and Its Impact on Cancer Therapy. Front Mol Biosci. 2020;6.
Page | 31 35. Asgharzadeh S, Salo JA, Ji L, Oberthuer A, Fischer M, Berthold F, et al. Clinical Significance of
Tumor-Associated Inflammatory Cells in Metastatic Neuroblastoma. J Clin Oncol. 2012 Oct 1;30(28):3525–32.
36. Liguori M, Solinas G, Germano G, Mantovani A, Allavena P. Tumor-Associated Macrophages as Incessant Builders and Destroyers of the Cancer Stroma. Cancers. 2011 Sep 28;3(4):3740–61. 37. Chanmee T, Ontong P, Konno K, Itano N. Tumor-Associated Macrophages as Major Players in
the Tumor Microenvironment. Cancers. 2014 Sep;6(3):1670–90.
38. Guruvayoorappan C. Tumor Versus Tumor-Associated Macrophages: How Hot is the Link? Integr Cancer Ther. 2008 Jun 1;7(2):90–5.
39. Lee H-W, Choi H-J, Ha S-J, Lee K-T, Kwon Y-G. Recruitment of monocytes/macrophages in different tumor microenvironments. Biochim Biophys Acta BBA - Rev Cancer. 2013 Apr 1;1835(2):170–9.
40. Cheng N, Bai X, Shu Y, Ahmad O, Shen P. Targeting tumor-associated macrophages as an antitumor strategy. Biochem Pharmacol. 2021 Jan 1;183:114354.
41. Lopez-Yrigoyen M, Cassetta L, Pollard JW. Macrophage targeting in cancer. Ann N Y Acad Sci. 2020 May 22;
42. Goswami KK, Ghosh T, Ghosh S, Sarkar M, Bose A, Baral R. Tumor promoting role of anti-tumor macrophages in anti-tumor microenvironment. Cell Immunol. 2017 Jun 1;316:1–10. 43. Kzhyshkowska J, Larionova I, Liu T. YKL-39 as a Potential New Target for Anti-Angiogenic
Therapy in Cancer. Front Immunol. 2020;10.
44. Ahmed N. 15 - Role of Integrins in Ovarian Cancer. In: Hayat MA, editor. Handbook of Immunohistochemistry and in Situ Hybridization of Human Carcinomas [Internet]. Academic Press; 2006 [cited 2021 Mar 3]. p. 425–39. (Handbook of Immunohistochemistry and
Hybridization of Human Carcinomas, Volume 4; vol. 4). Available from: https://www.sciencedirect.com/science/article/pii/S1874578405801011
45. Lucotti S, Muschel RJ. Platelets and Metastasis: New Implications of an Old Interplay. Front Oncol. 2020;10.
46. Kumar P. The Effect of Deactivating or Silencing Tumor Stroma for Growth of Tumors [Master’s thesis]. [Uppsala]: Uppsala University; 2020.
47. Guo Y, Tan J, Miao Y, Sun Z, Zhang Q. Effects of Microvesicles on Cell Apoptosis under Hypoxia. Oxid Med Cell Longev. 2019 Apr 17;2019:e5972152.
48. Henze A-T, Mazzone M. The impact of hypoxia on tumor-associated macrophages. J Clin Invest. 126(10):3672–9.
49. Raza A, Franklin MJ, Dudek AZ. Pericytes and vessel maturation during tumor angiogenesis and metastasis. Am J Hematol. 2010 Aug 1;85(8):593–8.
