Summary
The aim of the present study was to investigate the correlation between the density of infiltrating T cells and macrophages in the parental colorectal cancer (CRC) and the growth rate of tumoroids (i.e. a patient‐derived in vitro 3D model). Tumoroids were established from fresh specimens of primary and metastatic CRC from 29 patients. The in vitro growth rate of tumoroids was monitored by automated imaging. The density of infiltrating T cells and macrophages was determined in the centre of the tumour (CT) and at the invasive margin (IM) of the parental tumours. This was performed by digital image analysis on the whole‐slide scanned images using Visiopharm® software. Tumoroids with higher density of infiltrating CD3+ lymphocytes in the IM of their parental tumour showed a higher growth rate (P < .0005). The average relative growth rate (log10) during the period from day 1 to day 11 was 0.364 ± 0.006 (mean ± SD) for the CD3+ (IM)‐high group and 0.273 ± 0.008 (mean ± SD) for the CD3+ (IM)‐low group. In contrast, the density of CD68+ infiltrating macrophages in the parental tumours showed significant inverse effect on the growth rate of the tumoroids (P < .0005). The present study showed that the density of immune cells in the parental CRC correlates with the growth rate of the tumoroids. The future perspective for such a 3D model could be in vitro investigations of the tumour‐associated inflammatory microenvironment as well as personalized cancer immunotherapy.
Keywords: colorectal adenocarcinoma, digital pathology, in vitro 3D culture, tumour‐associated macrophages, tumour‐associated T lymphocytes
1. INTRODUCTION
In vitro 3D cultures of tumour cells encapsulate the functionality of the parental tumours in vivo.1 These models are often used in preclinical screening of anti‐cancer agents and are important tools for anti‐cancer drug development. Furthermore, these 3D models have shown to be useful for investigation of functional features of the solid tumours, such as invasion and angiogenesis.2
Research over the past decade has shown that malignant tumours do not consist of a collection of homogenous cancer cells that autonomously proliferate. Tumours are rather complex tissues consisting of cancer cells in interaction with the ‘tumour microenvironment’.3 The immune cells have important roles in this tumour microenvironment. Tumour‐infiltrating immune cells have both tumour‐antagonizing and tumour‐promoting effects. The main tumour‐antagonizing immune cells include cytotoxic T cells and natural killer cells. Macrophages may have some tumour‐promoting role, but their effect can be paradoxical.4 Hence, the inflammatory microenvironment in the parental tumour could affect the growth rate of the tumoroids. One might anticipate that the presence of immune cells may have some correlation with growth patterns, but this has not been documented in detail before. Therefore in this study we investigated whether the density of the T cells and macrophages in the parental tumour correlates with the growth rate of their tumoroids. These cell types were selected based on the evidence of correlation with prognosis of colorectal cancer. Eriksen et al recently showed that the low density of CD3+ and CD8+ tumour‐infiltrating immune cells in the invasive margin of the stage II colorectal adenocarcinoma is related to worse prognosis of the disease.5 Macrophages have been shown to be an important factor for seeding of metastatic cancer cells.6 Furthermore, macrophages play a role in migration and invasion of the cancer cells.6
Thus, in summary, the aim of this study was to investigate the correlation between infiltrating T cells and macrophages in "parental" colorectal adenocarcinoma, and the growth rate of tumoroids derived from them.
2. MATERIALS AND METHODS
2.1. Tumour specimens
Nineteen fresh primary colorectal tumours and 10 liver metastases from 29 patients with colorectal adenocarcinoma were included. The patients were operated at Bispebjerg Hospital and Rigshospitalet, Copenhagen, Denmark. Non‐necrotic areas of the tumour were selected by gross inspection (performed by a pathologist at the Department of Pathology, Rigshospitalet). Approximately 0.5 cm3 of each tumour was transferred to the culturing laboratory in transfer medium. The transfer medium consisted of Dulbecco's phosphate‐buffered saline (Sigma‐Aldrich, St. Louis, Missouri, USA) containing 500 U/mL penicillin, 500 U/mL streptomycin (Sigma‐Aldrich), 500 µg/mL gentamycin (Thermo Fisher Scientific, Waltham, Massachusetts, USA) and 12.5 µg/mL amphotericin B (Sigma‐Aldrich). For each patient, the following features were registered from the pathology report: Gender, age, TNM stage [local invasion depth (T stage), lymph node involvement (N stage), presence of distant metastasis (M stage)]7 and the staging based on Union for International Cancer Control (UICC) classification of colorectal cancer. The relative growth rates of tumoroids in a 11‐day period of culture were recorded by automated surveillance as described in our previous work.8
2.2. Ethical approval
The project was approved by the local committee for scientific ethics (No. H.1.2011.125)
2.3. Immunohistochemistry
After removal of the tissue for tumoroid culture, the rest of the tumour was fixed in formalin, dehydrated in PLORIS tissue processor (Leica Biosystems, Wetzlar, Germany) and embedded in liquid paraffin. Immunohistochemistry (IHC) stains were performed on the formalin‐fixed and paraffin‐embedded (FFPE) samples. Five micrometre thick sections were used for IHC. The sections were immunohistochemically stained using the following protocol: CD3 (clone 2GV6; Roche, Tucson, AZ, USA) mouse monoclonal primary antibody for 24 minutes at 36ºC, CD8 (clone SP57; Roche) mouse monoclonal primary antibody for 8 minutes at 36ºC and CD68 (clone PG‐M1; dilutions 1:100; Agilent, Santa Clara, CA, USA) mouse monoclonal primary antibody for 28 minutes at 36ºC. IHC was performed on a BenchMark ULTRA platform (Roche/Ventana) with the OptiView DAB IHC Detection Kit (Roche). In order to see whether the immune cells can be detected in the in vitro cultures, the tumoroids established from three different randomly selected patients were IHC stained for CD3, CD8 and CD68 markers. The expression of mismatch‐repair proteins was investigated by IHC on the FFPE tissue of the parental tumours as described by Boennelycke et al.9
2.4. Digital image analysis (DIA)
The IHC‐stained slides were digitally scanned at 40X magnification by Hamamatsu NanoZoomer‐XR (Hamamatsu Photonics, Hamamatsu City, Japan). The Visiopharm® Quantitative Digital Pathology software (Visiopharm A/S, Hoersholm, Denmark) was used for digitally quantification of positive cells based on the developed algorithms. Prior to DIA, the centre of tumour (CT) and invasive margin (IM) of the tumour were delineated by an experienced pathologist on the CD3+‐stained sections guided by the haematoxylin‐ and eosin‐stained sections. Similar areas were detected on CD8+ and CD68+‐stained sections using the Tissuealign™ add‐on, and positive cells were assessed in these areas. IM was selected with approximately 500 µm width. The CT area was selected without necrosis. The algorithms were developed using a ‘Decision Forest classifier’ trained on preprocessing steps that highlight the IHC staining. The preprocessing was done with filters of HDAB‐DAB and HDAB‐haematoxylin. Postprocessing was applied to fine‐tune the algorithm by removing the false‐positive objects based on minimum size. The number of cells was calculated by dividing the total positive area by the average size of one cell. The mean area of the cells was calculated for each marker by manual assessment of 300 cells.
2.5. Statistical analysis
The statistical analysis was performed on the collective data including both the primary tumours and the liver metastases. The number of immune cells was considered to be a continuous variable for the statistical analysis. The linear regression model was performed on log10‐transformed values of cell numbers and relative growth rates of tumoroids. Paired samples t test was used to test the correlation between the density of the immune cells in CT and IM as well as the correlation between the density of different types of the immune cells. The statistical analysis was performed in IBM® SPSS® Statistics (Version 22.0.0.0, Armonk, NY, USA). P‐values less than .05 were considered significant.
3. RESULTS
Tumoroids were successfully established from all the included tumours. All the tumours were histologically diagnosed as glandular adenocarcinoma except patients no. 7 and 12, which were mucinous and low‐differentiated adenocarcinoma respectively. All the tumours showed sufficient expression of mismatch‐repair proteins. Out of 19 primary tumours, three cases were classified as stage I, six as stage II, eight as stage III and two cases as stage IV. The clinicopathologic features of the tumours are listed in Table 1. The average relative growth rate (log10) of tumoroids increased from 0.036 ± 0.070 (mean ± SD) at day 1 to 0.457 ± 0.248 (mean ± SD) at day 11. Densities (cells/mm2) of CD3+, CD8+ and CD68+ cells were successfully assessed in CT and IM areas of the original tumour by DIA (Figure 1) on the scans of IHC‐stained sections. The density of the immune cells in CT correlated positively and significantly with their density at IM (P < .0001). Furthermore, there was a positive and significant correlation between the densities of different types of the immune cells (P < .0005).
Table 1.
Overview of tumoroids
| Patient code | Gender | Age (y) | Origin of tumoroid | T stage | N stage | M stage | UICC Stage | Expression of MMR proteins |
|---|---|---|---|---|---|---|---|---|
| 1 | F | 68 | Colon | T1 | N0 | M0 | I | Sufficient |
| 2 | M | 71 | Colon | T2 | N1 | M0 | IIIA | Sufficient |
| 3 | M | 76 | Colon | T3 | N0 | M0 | IIA | Sufficient |
| 4 | F | 85 | Colon | T2 | N0 | M0 | I | Sufficient |
| 5 | M | 78 | Colon | T3 | N2 | M0 | IIIC | Sufficient |
| 6 | M | 76 | Colon | T3 | N0 | M0 | IIA | Sufficient |
| 7 | M | 77 | Colon | T3 | N1 | M1 | IV | Sufficient |
| 8 | M | 75 | Colon | T2 | N1 | M0 | IIIA | Sufficient |
| 9 | F | 64 | Colon | T2 | N1 | M0 | IIIA | Sufficient |
| 10 | F | 61 | Colon | T3 | N1 | M0 | IIIA | Sufficient |
| 11 | F | 67 | Colon | T4 | N2 | M0 | IIIC | Sufficient |
| 12 | M | 63 | Colon | T3 | N0 | M0 | IIA | Sufficient |
| 13 | F | 62 | Colon | T4 | N0 | M0 | IIB | Sufficient |
| 14 | F | 73 | Colon | T3 | N0 | M0 | IIA | Sufficient |
| 15 | M | 53 | Colon | T4 | N1 | M0 | IIIB | Sufficient |
| 16 | M | 79 | Colon | T2 | N0 | M0 | I | Sufficient |
| 17 | F | 62 | Rectum | T4 | N1 | M0 | IIIB | Sufficient |
| 18 | M | 81 | Rectum | T3 | N2 | M1 | IV | Sufficient |
| 19 | F | 76 | Rectum | T3 | N0 | M0 | IIA | Sufficient |
| 20 | M | 64 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 21 | F | 72 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 22 | F | 52 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 23 | M | 71 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 24 | M | 74 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 25 | F | 77 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 26 | F | 64 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 27 | F | 75 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 28 | M | 70 | Liver | N/A | N/A | N/A | N/A | Sufficient |
| 29 | M | 54 | Liver | N/A | N/A | N/A | N/A | Sufficient |
Abbreviations: M, Male; F, female; T, tumour stage; N, lymph node metastasis; M, distant metastasis; N/A, not applicable; UICC, Union for International Cancer Control; MMR, mismatch repair.
Figure 1.
Automated quantification of the immune cells in colon adenocarcinoma by digital image analysis. Main figure: The centre of the tumour (blue marking) and the invasive margin (red marking). A‐C, Recognition of CD3+, CD8+ and CD68+ cells respectively. The left panels show the IHC‐stained section before analysis, and the right panels show the areas detected positive, indicated by green label, adjacent to the background area, indicated by the blue label. (bar = 500 µm)
3.1. Higher density of CD3+ infiltrating cells at the invasive margin of the parental tumour is associated with higher growth rate of tumoroids
The density (cells/mm2) of CD3+ cells in CT and IM of the primary tumours was 678 ± 730 (mean ± SD) and 1209 ± 1181 (mean ± SD) respectively. The density of CD3+ cells in CT and IM of the liver metastases was 194 ± 106 (mean ± SD) and 1542 ± 619 (mean ± SD) respectively. The average relative growth rate (log10) during the period from day 1 to day 11 was 0.364 ± 0.006 (mean ± SD) for the CD3+ (IM)‐high group and 0.273 ± 0.008 (mean ± SD) for the CD3+ (IM)‐low group. The density of the CD3+ cells at the IM of the parental tumours showed a statistically significant correlation with the growth rate (log10) of tumoroids for the whole culture period (P < .0005). This correlation was not significant for CT (Figure S1). The different growth rates of the tumoroids established from the parental tumours with the highest 50% vs the lowest 50% density of infiltrating CD3+ cells at IM are shown in Figure 2 and Figure S2. No CD3+ cells were detected in the IHC‐stained sections of the investigated tumoroids.
Figure 2.
Scatter plot with fitted regression line showing relative growth rates of tumoroids established from the tumours with the highest 50% (blue colour) vs the lowest 50% (green colour) density of CD3+ cells at the invasive margin (IM) of their parental tumours. Each circle represents the relative rate of growth area in a specific culture well for a specific patient. The correlation between increasing number of the CD3+ cells at IM of the parental tumour and the higher growth rate of tumoroids was significant during the period from day 1 to day 11 (P < .0005)
3.2. Higher density of infiltrating macrophages at the invasive margin of the parental tumour was associated with lower growth rate of tumoroids
The density of CD68+ cells in CT and IM of the primary tumours was 442 ± 205 (mean ± SD) and 996 ± 633 (mean ± SD) respectively. The density of CD68+ cells in the liver metastases was 330 ± 165 (mean ± SD) and 645 ± 236 (mean ± SD) for CT and IM respectively. The average relative growth rate (log10) during the period from day 1 to day 11 was 0.276 ± 0.006 (mean ± SD) for the CD68+ (IM)‐high group and 0.378 ± 0.008 (mean ± SD) for the CD68+ (IM)‐low group. The density of macrophages at the invasive margin of the parental tumours showed an inverse correlation with the growth rate of tumoroids (P < .0005). The correlation between the number of the CD68+ cells in the centre of the parental tumours and the growth rate of tumoroids during the period from day 1 to day 11 was not statistically significant (P = .163) (Figure S3). The growth rate of the tumoroids established from the parental tumours with the highest 50% vs the lowest 50% density of infiltrating CD68+ cells at the IM is shown in Figure 3 and Figure S4. No CD68+ cells were detected in the IHC‐stained sections of the investigated tumoroids.
Figure 3.
Scatter plot with fitted regression line showing relative growth rates of tumoroids established from the tumours with the highest 50% (blue colour) vs the lowest 50% (green colour) density of CD68+ cells at the invasive margin (IM) of their parental tumours. Each circle represents the relative rate of growth area in a specific culture well for a specific patient. The statistical analysis showed a significant correlation between increasing number of the CD68+ cells at IM of the parental tumour and the lower growth rate of tumoroids during the period from day 1 to day 11 (P < .0005)
3.3. The density of CD8+ infiltrating T cells in the parental tumour was not associated with growth rate of tumoroids
The density (cells/mm2) of CD8+ T cells in CT and IM of the primary tumours was 127 ± 104 (mean ± SD) and 619 ± 596 (mean ± SD) respectively. The density of CD8+ cells for the liver metastases was 80 ± 85 (mean ± SD) and 344 ± 157 (mean ± SD) for CT and IM respectively. The average relative growth rate (log10) during the period from day 1 to day 11 was 0.393 ± 0.008 (mean ± SD) for the CD8+ (IM)‐high group and 0.284 ± 0.006 (mean ± SD) for the CD8+ (IM)‐low group. Statistical analysis did not show any significant correlation between the density of CD8+ T cells in the CT and IM of the parental tumours and growth rate of the tumoroids (P = .084 and P = .591 respectively) (Figures S5 and S6). The immunohistochemical staining of the investigated tumoroids did not detect any CD8+ cells.
4. DISCUSSION
The tumoroid in vitro 3D model is useful for investigation of the functional features of solid tumours and for preclinical development of new drugs.10,11 Previous studies have shown that by retaining the cell‐cell and cell‐matrix contacts of the tumour cells, the developed model resembles the original tissue more closely.12, 13, 14 Our group has previously shown that tumoroids can be established from colorectal adenocarcinoma with high success rate; tumoroids recapitulate the biologic features of their parental tumours;15 andtumoroids established from different parts of the tumours reflect the intra‐tumoural heterogeneity.16
The validity of such a model is based on the knowledge about factors affecting the in vitro growth of cancer cells. The mechanism behind the growth rate of tumoroids is poorly understood. Our group has shown that some mutations in the parental tumour accelerate growth rate of tumoroids established from colorectal adenocarcinoma.8 It is expected that different factors in the tumour microenvironment of the parental tumours correlate with the growth of patient‐derived tumoroids. One group of these factors are the tumour‐associated immune cells. Since 1990s, the question has been, whether tumour‐associated immune cells have a pro‐neoplastic effect on the neighbouring cancer cells.17, 18, 19, 20 In this study, we have assessed the correlation between the density of the T cells and macrophages in the parental tumours and the growth rate of the tumoroids. The quantification of the CD3+, CD8+ and CD68+ cells was performed in both CT and IM areas based on previous studies showing a more reliable correlation with prognosis than taking only one of the measures into account.5,21,22 We showed that a higher density of the tumour‐infiltrating CD3+ cells correlates with a higher growth rate of the tumoroids, while the higher density of the CD68+ cells showed an inverse correlation with the growth rate. The statistical analysis was performed on the collective data including both the primary tumours and the liver metastases, as the sample size in each group was small.
This study showed that there may be a dynamic interaction between the immune cells and cancer cells in vivo which persists in an in vitro environment. The immune cells in the parental tumour may influence the growth of the tumoroids through a variety of chemokines, cytokines, proangiogenic and pro‐invasive factors released from the immune cells. Studies have shown several signalling molecules released by the immune cells which are effectors of tumour‐promoting or tumour‐antagonizing features. These molecules include but are not limited to vascular endothelial growth factor (VEGF), fibroblast growth factor 2 (FGF2), epidermal growth factor (EGF) and matrix‐degrading enzymes.6,23 The tumour‐associated immune response is complex. It serves both as tumour‐eradicating and tumour‐promoting systems. The tumour‐eradicating effect is mediated primarily by natural killer cells and cytotoxic T lymphocytes. The tumour‐promoting role which includes angiogenesis, proliferation, invasion and dissemination of the tumour is mediated by macrophages, neutrophils, mast cells, B‐ and T lymphocytes.4,24 These diverse roles of the immune system reflect the indigenous function of the immune system to eradicate both the infectious agents and the dead cells from the host.3
We could not detect CD3+, CD8+ or macrophages in the IHC‐stained sections of the tumoroids in a random subset of the patients. This is in accordance with several previous studies.12, 13, 14 This could be explained by the inability of the immune cells to survive in the laboratory conditions after several days. As the immune cells are not present in the tumoroids, the impact of the removal of the inflammatory microenvironment could be investigated. It is plausible that the higher growth rate of the tumoroids established from parental tumours with high density of CD3+ cells may be either due to the removal of the growth‐antagonizing effect of these cells on the cancer cells or due to their long‐lasting growth‐promoting effect. In the same way, the inverse correlation between the density of the macrophages in the parental tumour and the growth rate of tumoroids may be either due to the removal of the growth‐promoting impact of these cells on the cancer cells or due to their long‐lasting growth‐antagonizing impact.
DIA provides the pathologists with a reliable tool that rapidly and objectively quantifies the immune cells in a histologic section. There are limitations to this approach. The IHC‐stained slides must be of high quality; otherwise, the background noise of the chromogen will compromise the accuracy of the DIA algorithms. Occasionally, the distinction between the two adjacent immune cells by DIA is impossible. We have minimized this risk by dividing the total positive area with the average size of one immune cell in order to calculate the density of the cells.
The immunoscore data have shown that the type, density and location of the immune cells may have a superior prognostic value to the traditional clinical and pathological staging of the tumours. Our results support these data by showing that the density of the tumour‐infiltrating immune cells has a dynamic and long‐lasting effect on the tumour cells. Therefore, selection of adjuvant chemotherapy or immunotherapy could be guided by assessment of the density of the tumour‐infiltrating immune cells independent of the clinical stage.
In conclusion, the present study showed the correlation between the density of the tumour‐infiltrating immune cells and the growth rate of tumoroids established from these tumours. There is a need for investigation of the relationship between the presence of the immune cells in the parental tumour and the growth rate of the tumoroids by implementing more elaborate markers and combination of these markers in the future studies. As an example, CD163, CD16, CD312 and CD115 would be useful to discriminate macrophages from other cells of the phagocytosis system. Furthermore, in situ identification of different subtypes of macrophages including M1, M2 and regulatory macrophages would help to understand, how these immune cells influence cancer cells. The results of this study showed that the immune cells of the microenvironment in the original tumour seem to influence the growth of tumoroids.
CONFLICT OF INTEREST
The authors declare no conflict of interests.
Supporting information
ACKNOWLEDGEMENTS
This work was supported by the Danish Cancer Society.
Mousavi N, Jespersen AJB, Jorgensen LN, Timmermans V, Heegaard S. The density of infiltrating T cells and macrophages in the parental tumour correlates with growth rate of tumoroids established from colorectal adenocarcinoma. Int J Exp Path. 2020;100:304–310. 10.1111/iep.12341
Mousavi and Jespersen contributed equally to this work.
REFERENCES
- 1. Sachs N, Clevers H. Organoid cultures for the analysis of cancer phenotypes. Curr Opin Genet Dev. 2014;24:68‐73. [DOI] [PubMed] [Google Scholar]
- 2. Vinci M, Gowan S, Boxall F, et al. Advances in establishment and analysis of three‐dimensional tumor spheroid‐based functional assays for target validation and drug evaluation. BMC Biol. 2012;10:29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646‐674. [DOI] [PubMed] [Google Scholar]
- 4. DeNardo DG, Andreu P, Coussens LM. Interactions between lymphocytes and myeloid cells regulate pro‐ versus anti‐tumor immunity. Cancer Metastasis Rev. 2010;29:309‐316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Eriksen AC, Sorensen FB, Lindebjerg J, et al. The prognostic value of tumor‐infiltrating lymphocytes in stage ii colon cancer. A nationwide population‐based study. Transl Oncol. 2018;11:979‐987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Qian BZ, Pollard JW. Macrophage diversity enhances tumor progression and metastasis. Cell. 2010;141:39‐51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Brierly JDGM, Wittekind C. TNM Classification of Malignant Tumours, 8th edn UK: Wiley‐Blackwell; 2017. [Google Scholar]
- 8. Mousavi N, Truelsen SLB, Hagel G, et al. KRAS mutations in the parental tumour accelerate in vitro growth of tumoroids established from colorectal adenocarcinoma. Int J Exp Pathol. 2019;100:12‐18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Boennelycke M, Thomsen BM, Holck S. Sebaceous neoplasms and the immunoprofile of mismatch‐repair proteins as a screening target for syndromic cases. Pathol Res Pract. 2015;211:78‐82. [DOI] [PubMed] [Google Scholar]
- 10. Van de Wetering M, Francies HE, Francis JM, et al. Prospective derivation of a living organoid biobank of colorectal cancer patients. Cell. 2015;161:933‐945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Xu X, Farach‐Carson MC, Jia X. Three‐dimensional in vitro tumor models for cancer research and drug evaluation. Biotechnol Adv. 2014;32:1256‐1268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kondo J, Endo H, Okuyama H, et al. Retaining cell‐cell contact enables preparation and culture of spheroids composed of pure primary cancer cells from colorectal cancer. Proc Natl Acad Sci USA. 2011;108:6235‐6240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lee SH, Hong JH, Park HK, et al. Colorectal cancer‐derived tumor spheroids retain the characteristics of original tumors. Cancer Lett. 2015;367:34‐42. [DOI] [PubMed] [Google Scholar]
- 14. Ashley N, Jones M, Ouaret D, Wilding J, Bodmer WF. Rapidly derived colorectal cancer cultures recapitulate parental cancer characteristics and enable personalized therapeutic assays. J Pathol. 2014;234:34‐45. [DOI] [PubMed] [Google Scholar]
- 15. Jeppesen M, Hagel G, Glenthoj A, et al. Short‐term spheroid culture of primary colorectal cancer cells as an in vitro model for personalizing cancer medicine. PLoS ONE. 2017;12:e0183074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Arnadottir SS, Jeppesen M, Lamy P, et al. Characterization of genetic intratumor heterogeneity in colorectal cancer and matching patient‐derived spheroid cultures. Mol Oncol. 2018;12:132‐147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Dvorak HF. Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med. 1986;315:1650‐1659. [DOI] [PubMed] [Google Scholar]
- 18. Schafer M, Werner S. Cancer as an overhealing wound: an old hypothesis revisited. Nat Rev Mol Cell Biol. 2008;9:628‐638. [DOI] [PubMed] [Google Scholar]
- 19. Karin M, Lawrence T, Nizet V. Innate immunity gone awry: linking microbial infections to chronic inflammation and cancer. Cell. 2006;124:823‐835. [DOI] [PubMed] [Google Scholar]
- 20. Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell. 2010;140:883‐899. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Anitei MG, Zeitoun G, Mlecnik B, et al. Prognostic and predictive values of the immunoscore in patients with rectal cancer. Clin Cancer Res. 2014;20:1891‐1899. [DOI] [PubMed] [Google Scholar]
- 22. Galon J, Costes A, Sanchez‐Cabo F, et al. Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science. 2006;313:1960‐1964. [DOI] [PubMed] [Google Scholar]
- 23. Murdoch C, Muthana M, Coffelt SB, Lewis CE. The role of myeloid cells in the promotion of tumour angiogenesis. Nat Rev Cancer. 2008;8:618‐631. [DOI] [PubMed] [Google Scholar]
- 24. Coffelt SB, Lewis CE, Naldini L, Brown JM, Ferrara N, De Palma M. Elusive identities and overlapping phenotypes of proangiogenic myeloid cells in tumors. Am J Pathol. 2010;176:1564‐1576. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials