Narrative review: research progress of tumor budding in gastrointestinal tumor

Introduction

Tumor budding, also known as “tumor sprouting”, is a histological phenomenon encountered in all kinds of cancers, whereby single malignant cells and/or small malignant cell clusters in the front stroma of the tumor. However, significant heterogeneity in its accurate definition, evaluation methods, and patient stratification need to be addressed. The mechanism of tumor budding is not clear. Tumor budding is considered to be the transition from mirror epithelium to mesenchymal, which is related to the poor prognosis of cancer. In recent years, the research from protein level to gene level has been improved, which puts forward another possibility. Here we review the history of tumor budding, the controversy over the evaluation of tumor budding, the different explanations of the causes of tumor budding, and the progress of its application in clinical practice. We present the following article in accordance with the Narrative Review reporting checklist (available at http://dx.doi.org/10.21037/dmr-20-113).

The history of tumor budding

The phenomenon of tumor budding was first studied by Fukuoka Igaku in 1954 when he was observing gastric cancer cells (1). He found that there was “sprouting” at the edge of the invasion of gastric cancer. There were undifferentiated clusters of malignant cells in the invasive stroma of gastric cancer, which were mainly (but not completely) located in front of the tumor invasion. Then, in the 1950–1960s, other Japanese scholars found this phenomenon also existed in pathological sections of tongue, throat, breast, stomach, colon, rectum, and cervical cancer. In the 1980s, when Gabbert and his colleagues experimentally-induced colon cancers in mice treated with dimethylhydrazine-dihydrochloride, they found an obvious invasive front of the lesions displayed striking disorganization at the tumor architecture level which helps to mobilize cancer cells from major tumor masses (2). In the 1990s, Hase analyzed the pathology of surgical samples of colorectal cancer and found that poorly differentiated malignant cells moved outward from the invasive front alone or in groups, and the aggregated cells formed a large cell mass, similar to the anterior edge (3). Immunohistochemistry suggested that E-cadherin was down-regulated, accompanied by nuclear translocation of β-catenin and loss of cell polarity. They defined this phenomenon as "tumor budding". This definition has triggered some criticism related to terminology. Some people think that “budding” means that the tumor cell population is still connected to the main tumor, but this is not the case observed in 2D sections. Bronsert and his colleagues show that budding is not a static snapshot, but a dynamic process through the 3D reconstruction of 2D serial sections of various tumor types, through which the tumor extends many finger-like processes (4). Each process contains many cells, which are separated from the main tumor mass in the form of small cell clusters at a later point in time. If you see such extension buds on a 2D section, it may give the wrong impression that the cell mass has been isolated from the main tumor mass. In short, 2D sections show only small cell masses that seem to have lost contact with major masses, while serial sections and 3D reconstruction provide a real picture that the tumor is not separated from the bud.

Early studies have suggested a significant positive correlation between “tumor budding” and poor prognosis of colorectal cancer (5-year survival rates: 22% and 71%, respectively), but this feature has progressed slowly in routine reports (5). Part of the reason is that the methods reported in the literature to evaluate budding are different. But the 2017 AJCC American Society of Pathologists’ guidelines for colorectal cancer reporting explicitly identified tumor budding as an optional reporting field and recommended reporting in all stage I and II cases. The meeting reached a consensus on the assessment of tumor budding and provided us with a standardized reporting tool for colorectal cancer sprouting. Tumor budding is also increasingly reported as a useful pathological prognostic feature of other gastrointestinal cancers, including esophageal squamous carcinoma and adenocarcinoma, gastrointestinal adenocarcinoma, pancreatic ductal adenocarcinoma and ampullary adenocarcinoma (6-10).

The present situation of the research on tumor budding

The evaluation criteria of “tumor budding”

Because “tumor budding” is affected by the subjective observation of pathologists, many scholars have put forward different evaluation methods. These differences of evaluation are mainly focused on the staining, counting, and "intra-tumor budding" of tumor budding. It is now widely used to evaluate intra-tumoral and peritumoral tumors using H&E or pan-cytokeratin staining. Both stainings have advantages and disadvantages. Pan-cytokeratin contributes to visual counting, causing more real tumor buds to be detected, and the nucleus needs to be observed to classify cells as buds, but the nucleus can be masked by strong cytokeratin staining, which can lead to overestimation of non-budding objects, such as cell fragments. This leads to variability between observers, especially at the single object level. H&E is a standard staining method that can be performed in all pathological laboratories, but it is difficult to identify inflammatory areas around the tumor, and it is difficult to distinguish between tumor budding and activated fibroblasts. Two staining methods were evaluated in a group of 50 patients with colorectal cancer, the overall repeatability was good, although cytokeratin staining increased the number of tumor sprouts (partly single cells in the detected glands destroyed by inflammatory cells), it did not increase the consistency between observers (11).

The counting of tumor sprouting is also controversial. Some scholars have proposed that the evaluation of “tumor sprouting” should be counted in an area that shows the maximum sprouting. Ueno in a study on the pathological prognostic characteristics of PT1 colorectal cancer, defined tumor budding as a single cell or <5 cell clusters, counted the maximum number of buds in the budding area with a visual field of ×250 (0.385 mm²) and then divided the patients into negative buds (<10 buds) and positive buds (≥10 buds), which was called “hot spot method” (12). In a subsequent study, the team further improved the counting method by counting buds in an area of ×200 (0.785 mm²), with five buds as the dividing line (13). Karamitopoulou outlined another popular method of budding assessment, which evaluates the entire invasion frontier, using an average count of 5–10 areas of the budding count, also known as the “average method” (14). In their study of pancreatic ductal adenocarcinoma, cytokeratin staining was used to highlight buds and average bud numbers on 10 high-power fields of (HPF). In a multicenter study on the repeatability of tumor germination, the hot spot method and the average method had similar inter-observer variability (kappa 0.35 vs. 0.25) (15).

Whether “peritumoral” and “intra-tumoral” tumor budding should be evaluated, or both is also controversial. The concept of intra-tumoral budding was introduced in 2011 (16). Intra-tumoral budding was defined as a cluster of single cells or less than 5 tumor cells surrounded by stroma. The budding at the frontier of tumor invasion in malignant polyps or PT1 colorectal cancer is usually clear, but it will be difficult and poorly reproducible to pathologically describe the invasive frontiers of pancreatic ductal adenocarcinomas in biopsies or ductal adenocarcinoma of the pancreas. Giger found that intra-tumoral budding during the biopsy and peritumoral germination during resection had similar repeatability (k=0.65 vs. k=0.68) and significant correlation. This study provides help for clinicians to evaluate “tumor budding” in biopsies because biopsies rarely sample the invasive edge, so it is possible to increase tumor-related information and change treatment options for patients before operation. but a larger sample control analysis is needed. Because of the diversity of evaluation criteria for “tumor budding”, the standardized method selected at the 2017 AJCC consensus meeting on Colorectal Cancer sprouting report: according to hematoxylin-eosin staining, the consensus definition of tumor budding is as follows: a single cell or a cluster of <5 cells, counting a ×200 field of vision in the maximum germination area (“hot spot method”). Then the patients were divided into low grade (0–4 buds), medium (5–9 buds), and high grade (≥10 buds) (17). To minimize variability in the selection of maximum budding areas, they recommend that at least 10 HPF be screened at the invasive frontier. Although it is believed that intra-tumoral sprouting may also be a useful prognostic factor, it is suggested that more research is needed before it is incorporated into routine practice. At present, these are recommendations for colorectal cancer. however, the best way to evaluate tumor germination, especially the best cut-off value, may vary in other parts of the gastrointestinal tract.

The mechanism of tumor budding

The cause of tumor budding is not clear. The study of tumor budding is mostly based on immunohistochemical observation in early research. The membrane localization of the key cell-cell adhesion molecule E-cadherin in epithelial cells of tumor budding is relatively low, and the loss of E-cadherin membrane localization is a feature of epithelial-mesenchymal transition (EMT). High-grade tumor budding shows strong and uniform nuclear β-catenin staining and associated loss of membrane E-cadherin expression, which is consistent with the EMT study. Therefore, the tumor budding is considered to be “EMT-like” (18). Many studies have found that from the tumor center to the tumor frontier, several markers mainly related to cell adhesion are heterogeneously expressed in tumor tissue and tumor budding. E-cadherin, CD44, EpCAM, and CD166 are expressed in the medial membrane of tumor tissue, while is missing in tumor budding (19-22). High-grade tumor budding is associated with increased expression of protein markers (such as u-PA and u-PAR), matrix lysin, or matrix metalloproteinases (MMPs) that are closely related to ECM degradation. In lung cancer, the level of β-catenin decreased in tumor budding (23). In invasive ductal breast cancer, the E-cadherin membrane localization of budding cell population was lower than that of central tumor cell, while the cytoplasmic level of vimentin was higher than that of central tumor cell (24). In tongue squamous cancer, the membrane localization of E-cadherin in the tumor budding was less than that in the main mass, while vimentin was positive in the tumor budding, but not in the main mass. High sprouting was related to the decrease of membrane localization of E-cadherin and the increase of vimentin (25). Although there is an obvious relationship between tumor budding and migration and invasion, paradoxically, tumor budding seems to experience a low proliferation rate, with a decrease in the expression of proliferation marker Ki67, accompanied by an increase in the expression of cell cycle arrest mediators cyclinD1 and p16, which validates the “go-or-grow” dichotomy hypothesis.

However, another view is that although tumor budding show down-regulation of E-cadherin, they do not share other co-regulatory changes of EMT, and there is insufficient evidence to associate tumor sprouting with complete EMT (pure mesenchymal phenotype). The cytoplasmic projection similar to flake liposome extends from a single cell in the tumor budding observed in the early stage. However, slice artifacts cannot be ruled out, which makes the discovery of individual cells ambiguous. In 2014, Bronsert and colleagues found no evidence of single-cell migration in their studies of different tumor types (colorectal cancer, PDAC, lung adenocarcinoma, and invasive breast ductal carcinoma), so the invasion of cancer cells mainly (if not all) depends on collective cell migration rather than individual cell migration, which indicates that some epithelial cell-to-cell adhesions persist. These findings suggest that not all cells become pure mesenchymal, and some tumor cells still express epithelial cells. Interestingly, this expression has also occurred in the process of embryonic cell migration, which is called partial EMT. Partial EMT reflects the epithelial plasticity of epithelial cells, that is, EMT and MET are not all or no response, but a switch between pure epithelial and pure mesenchymal phenotypes (26). What we have observed is a multistate process of one or more intermediate phenotypes from pure epithelial to pure mesenchymal.

Part of the phenomenon of EMT has attracted widespread attention in recent years, and researchers have investigated the expression of epithelial and interstitial markers in different cell lines, patient-derived xenografts, and primary cancer. In the breast, pancreas, kidney, lung, colorectal, and ovarian cancer cell lines, these two markers were co-expressed in the same cell, suggesting the existence of EMT hybridization. In vitro, hybrid phenotypes are associated with increased invasion and migration. Similarly, epithelial and interstitial markers are co-expressed in human primary cancers, such as breast cancer, colorectal, esophageal, lung, and pancreatic cancer, and are significantly associated with poor prognosis. Carcinosarcoma is a rare tumor that contains epithelial and mesenchymal portions of clonal origin in the same tumor, representing an example of spontaneous EMT observed in primary human cancers from different organs. Sidharth V Puram identified the partial/mixed EMT regimen in 18 patients with head and neck squamous cell carcinoma (HNSCC) by analyzing 6,000 single-cell transcriptional groups (including 5 pairs of primary tumors and lymph node metastasis), which was defined as the incomplete activation of EMT transcription factor (TF) (27). Interestingly, they found that some of the EMT cells were spatially located at the leading edge of the tumor. This seems to confirm the possibility of EMT in the budding part of the tumor. In the study of PDAC, compared with the tumor center, the membrane localization of E-cadherin and β-catenin in tumor buds was lower (at the cell population level). However, few tumors showed vimentin-positive in tumor budding, which indicated that most of the EMT in patients with PDAC was partial EMT (28). The sprouting of skin squamous carcinoma shows an increasing level of laminin-5γ2, which is an extracellular matrix glycoprotein expressed at the front of collective migration cells during wound healing, suggesting the possibility of partial EMT (29). Although the tumor budding of oral squamous carcinoma significantly up-regulates fibronectin, only a small number of tumor samples of oral squamous carcinoma cells undergo a complete “cadherin conversion” from E-cadherin to N-cadherin (30). Compared with the major tumor, the sprouting of lung cancer did not express ZEB141, the main regulator of EMT. Meyer found that pan-cytokeratin (epithelial marker) and vimentin (interstitial marker) were both positive in colorectal cancer tumor budding (31). In summary, these conclusions suggest that tumor budding does not have a completely stable mesenchymal phenotype; on the contrary, they are partial manifestations of EMT.

In recent years, with the maturity of sequencing technology, the gene sequencing of tumor budding has been realized, which provides an important tool to unravel the mechanism of tumor budding. In 2015, Jensen used laser capture microdissection, RNA sequencing, and miRNA-qPCR array to detect oral squamous carcinoma tumor budding. Compared with the cells in the central part of the tumor, budding cells showed unique gene expression characteristics, including factors involved in epithelial-mesenchymal transformation and activated transforming growth factor signal TGF-β, up-regulated expression of TFs ZEB1 and Prrx1, while decreased expression of mesenchymal-epithelial TFs (such as OVOL1) (32). Subsequently, Linde used the same method to sequence the tumor sprouts and tumor masses of colon cancer and found that 296 genes were differentially expressed (33). Compared with the tumor blocks, the tumor budding is characterized by phenotypic transformation, which is related to the acquisition of migration characteristics and the reduction of cell proliferation. In addition, other studies have shown that tumor budding is associated with increased expression of hypothetical stem cell markers. Tumor budding expresses high levels of stem cell markers, such as CD133 and aldehyde dehydrogenase 1 (ALDH1), indicating that tumor budding has cancer stem cell-like characteristics. Budding tumor cells expressing ALDH1 are positively correlated with increased invasiveness and poor prognosis (34). A number of studies on the interaction between EMT and stem cells have shown that some EMT phenotypic cells may be more like stem cells than pure epithelial cells or pure mesenchymal cells, supporting the tumor budding is partial EMT (35-37). Therefore, the tumor budding at the edge of invasion can be considered as the realization of the proposed “transitional tumor stem cells”.

The tumor budding has attracted much attention in recent years for its correlation with poor prognosis. Few studies aim to its molecular mechanism because of lacking vitro model. it is generally believed that tumor budding is tumor clusters detached from tumor bulk. When and how this happen is still unknown, the phenomenon of EMT or partly EMT is a picture of this process. Because tumor budding take placed in three dimensions and Influenced by the surrounding microenvironment, the molecular mechanism is completely different from what we find in two dimensions. What we know of potential mechanisms of tumor budding is still too little. But we believe with the application of new technology, mechanisms of tumor budding will be clearly expounded in the near future.

Clinical application

For colorectal cancer, tumor budding has become an independent prognostic factor. Moreover, tumor budding has other guiding values for the choice of treatment for patients with colorectal cancer.

Tumor budding and colorectal cancer stage

First of all, for stage I colon cancer, tumor budding guides patients to benefit from oncological resection after diagnosis of primary tumors growing into the submucosa (PT1). In 2004, Ueno investigated a set of clinicopathological parameters, including tumor location, tumor diameter, macroscopic tumor morphology (stemless and pedicled), tumor grade, vascular invasion, tumor germination, and width and depth of submucosal infiltration (13). The study concluded that the lack of some features, including high tumor grade, vascular infiltration, budding, and extensive submucosal infiltration, may tend to be followed up. Similar results were obtained from a meta-analysis of 17 studies and 3,782 PT1CRC in 2013 by Bosch et al. (38). The guidelines, therefore, indicate that tumor budding is an independent predictor of lymph node metastasis in patients with PT1 colorectal cancer, and therefore strongly recommend that tumor budding be used in conjunction with other histopathological predictors of lymph node metastasis (such as poor differentiation, lymphatic vascular infiltration, and submucosal invasion depth/level) in patients with PT1 colorectal cancer to evaluate participation in treatment decision-making.

Second, the budding of stage II colorectal cancer may indicate that adjuvant therapy should be considered to improve survival. The latest management and treatment guidelines of the European Medical Oncology Society (ESMO) recommend that patients with low-risk stage II colorectal cancer be followed up, while patients with high-risk factors should consider fluorouracil adjuvant therapy, such as T4 (the tumor has grown to all layers of the colon and attaches or invades other structures and organs), the number of lymph nodes examined is less than 12, perforated or obstructed, grade 3, no microsatellite instability (MSI) and tumor budding (39). This recommendation is supported by the 2019 classification of digestive system tumors of the World Health Organization (WHO), which reports tumor budding with perineural infiltration, intramural and extramural vascular infiltration, lymphatic infiltration, and tumor deposition as risk factors, with an OR of 4.51 (95% CI, 2.55–7.99). Multivariate analysis showed that tumor budding had an independent effect on DFS. The study concluded that stage II colon cancer should routinely report tumor budding.

Third, the effectiveness of tumor budding in patients with stage III cancer has not been deeply evaluated. However, since adjuvant therapy is standardized, whether tumor budding can predict the response of this subgroup to chemotherapy remains to be studied. Rogers found that among the 89 patients with locally advanced rectal cancer, 18 (20%) showed budding in the tumor biopsy before treatment, and the patients with budding in the tumor did not show grade 1 tumor regression or complete pathological remission after neoadjuvant therapy (40). Intra-tumoral budding indicates the adverse pathological reaction to neoadjuvant radiotherapy and chemotherapy. Patients who may not have a complete pathological response to neoadjuvant therapy must be upgraded accordingly.

Fourth, in patients with stage IV colorectal cancer, the presence of tumor budding within or around metastasis (IMB and PMB, respectively) in liver metastasis may be a supporting marker for stratification of patients with different treatment options. The ESMO consensus guidelines for the treatment of patients with metastatic colorectal cancer emphasize the importance of multidisciplinary treatment, including oncology, surgery, radiology, and pathology. The most commonly used molecular markers in clinical practice are RAS, BRAF, and MSI status. The histopathological features of clinical treatment are the size of metastasis, the percentage of fibrosis and necrosis, the status of resection, and the degree of tumor degeneration (41). Several studies have shown that, tissue growth patterns such as connective tissue proliferation, push and metastasis have the potential to predict prognosis (42,43). Similar to primary tumors, tumor budding can be detected in the presence of intra-metastatic or premetastatic tumor budding (IMB and PMB, respectively) in colorectal cancer liver metastases (CRLM). Therefore, tumor budding may also be an important factor in the disease progression of patients with stage IV colorectal cancer, but compared with primary tumors, the detection of tumor budding in liver metastasis is still a major challenge.

Tumor budding and molecular typing of colon cancer

In addition, there is also a certain tendency for tumor budding in the molecular typing of colon cancer. CMS1 subtype is characterized by the defect of the DNA mismatch repair system, which is consistent with MSI. In addition, they are characterized by strong immune penetration and strong activation of immune escape pathways. Tumor budding is rarely found in patients with MSI tumors. One possible explanation is that inherent immune cell infiltration leads to the destruction of tumor budding. CMS4 subtype showed obvious activation of transforming growth factor-β pathway and Wnt- signal pathway, as well as markers of EMT and angiogenesis. This subtype showed the worst overall survival rate, the worst 5-year survival rate, and the relapse-free survival rate of all subtypes. A study also found that the frequency of budding of this type of tumor was significantly higher than that of other types. Interestingly, in this study, the researchers also found that the overall expression of the tumor was CMS2, while the tumor budding tissue was CMS4. Then, in 2018, Casasent et al. assured the genome copy number of a single tumor cell by topographic single-cell sequencing (TSCS) in the study of breast ductal carcinoma in situ, indicating that most mutations and copy number distortions evolved in the ducts before the invasion (44). These results support the polyclonal invasion model, in which one or more clones are transferred from ducts to adjacent tissues, resulting in the establishment of invasive cancer.

Tumor budding and gastric cancer

With tumor budding as an optional reporting field in the 2017 AJCC American Society of Pathologists’ Colorectal Cancer reporting guidelines, a growing body of evidence also points to tumor sprouting in gastric and pancreatic cancer, emphasizing the effectiveness of this standardized scoring method (45,46). The study of tumor budding in gastric cancer is less than that in colon cancer. A meta-analysis in 1961 analyzed seven cohorts containing data from 2,178 patients; high-level budding was positively correlated with high tumor stage and low differentiation (47). In the comprehensive analysis of 1,833 patients, high-grade tumor budding was significantly correlated with lymphatic invasion and lymph node metastasis, and with a poor 5-year overall survival rate. These results were confirmed in the subgroup analysis of intestinal type, but not in diffuse type. In the evaluation of tumor budding during radical gastrectomy in 621 patients with early submucosal gastric cancer, it was also found that high-grade tumor germination was a predictor of lymph node metastasis (48).

Discussion

Although a consensus has been reached in summarizing tumor budding in colorectal cancer, there is still great variability in evaluating budding for inexperienced pathologists. An artificial intelligence-assisted scoring system for judging tumor budding can save time and help to improve repeatability. Although tumor budding is also a very important prognostic marker in many kinds of tumors, it should be noted that the definition and scoring system of tumor budding may be different according to different tumor types, which requires further research to reach a consensus.

At present, the study on the mechanism of tumor budding still can not fully explain the formation of tumor budding. High-throughput multi-sample genomics research is helpful to further clarify the causes of tumor budding. The study of potential targets expressed in tumor budding may provide a promising method for anti-budding therapy to specifically target tumor cells that seem to be responsible for local and distant metastasis, thus improving the survival rate of tumors.

Acknowledgments

Funding: This work is supported by the Sanming Project of Medicine in Shenzhen (SZSM201911010).

Footnote

Reporting Checklist: The authors have completed the Narrative Review reporting checklist. Available at http://dx.doi.org/10.21037/dmr-20-113

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at http://dx.doi.org/10.21037/dmr-20-113). YH serves as the Editor-in-Chief of Digestive Medicine Research. CZ serves as an unpaid Associate Editor-in-Chief of Digestive Medicine Research. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Imai T. The growth of human carcinoma: a morphological analysis. Fukuoka Igaku Zasshi 1954;45:72-102.
2. Gabbert H, Wagner R, Moll R, et al. Tumor dedifferentiation: An important step in tumor invasion. Clin Exp Metastasis 1985;3:257-79. [Crossref] [PubMed]
3. Hase K, Shatney C, Johnson D, et al. Prognostic value of tumor “budding” in patients with colorectal cancer. Dis Colon Rectum 1993;36:627-35. [Crossref] [PubMed]
4. Bronsert P, Enderle-Ammour K, Bader M, et al. Cancer cell invasion and EMT marker expression—A three-dimensional study of the human cancer-host interface. J Pathol 2014;234:410-22. [Crossref] [PubMed]
5. Amato A, Scucchi L, Pescatori M. Tumor budding and recurrence of colorectal cancer. Dis Colon Rectum 1994;37:514-5. [Crossref] [PubMed]
6. Zhang J, Tian XJ, Zhang H, et al. TGF-β-induced epithelial-to-mesenchymal transition proceeds through stepwise activation of multiple feedback loops. Sci Signal 2014;7:ra91. [Crossref] [PubMed]
7. Sampson VB, David JM, Puig I, et al. Wilms’ tumor protein induces an epithelial-mesenchymal hybrid differentiation state in clear cell renal cell carcinoma. PLoS One 2014;9:e102041. [Crossref] [PubMed]
8. Schliekelman MJ, Taguchi A, Zhu J, et al. Molecular portraits of epithelial, mesenchymal, and hybrid states in lung adenocarcinoma and their relevance to survival. Cancer Res 2015;75:1789-800. [Crossref] [PubMed]
9. Hiew MSY, Cheng HP, Huang CJ, et al. Incomplete cellular reprogramming of colorectal cancer cells elicits an epithelial/mesenchymal hybrid phenotype. J Biomed Sci 2018;25:57-71. [Crossref] [PubMed]
10. Huang RY, Wong MK, Tan TZ, et al. An EMT spectrum defines an anoikis-resistant and spheroidogenic intermediate mesenchymal state that is sensitive to e-cadherin restoration by a src-kinase inhibitor, saracatinib (AZD0530). Cell Death Dis 2013;4:e915. [Crossref] [PubMed]
11. Puppa G, Senore C, Sheahan K, et al. Diagnostic reproducibility of tumor budding in colorectal cancer: a multicentre, multinational study using virtual microscopy. Histopathology 2012;61:562-75. [Crossref] [PubMed]
12. Ueno H, Murphy J, Jass J, et al. Tumour ‘budding’ as an index to estimate the potential of aggressiveness in rectal cancer. Histopathology 2002;40:127-32. [Crossref] [PubMed]
13. Ueno H, Mochizuki H, Hashiguchi Y, et al. Risk factors for an adverse outcome in early invasive colorectal carcinoma. Gastroenterology 2004;127:385-94. [Crossref] [PubMed]
14. Karamitopoulou E, Zlobec I, Born D, et al. Tumour budding is a strong and independent prognostic factor in pancreatic cancer. Eur J Cancer 2013;49:1032-9. [Crossref] [PubMed]
15. Nakamura T, Mitomi H, Kikuchi S, et al. Evaluation of the usefulness of tumor budding on the prediction of metastasis to the lung and liver after curative excision of colorectal cancer. Hepatogastroenterology 2005;52:1432-5. [PubMed]
16. Giger OT, Comtesse S, Lugli A, et al. Intra-tumoral budding in preoperative biopsy specimens predicts lymph node and distant metastasis in patients with colorectal cancer. Mod Pathol 2012;25:1048-53. [Crossref] [PubMed]
17. Lugli A, Kirsch R, Ajioka Y, et al. Recommendations for reporting tumor budding in colorectal cancer based on the International Tumor Budding Consensus Conference (ITBCC) 2016. Mod Pathol 2017;30:1299-311. [Crossref] [PubMed]
18. Brabletz T, Hlubek F, Spaderna S, et al. Invasion and metastasis in colorectal cancer: epithelial-mesenchymal transition, mesenchymal-epithelial transition, stem cells, and beta-catenin. Cells Tissues Organs 2005;179:56-65. [Crossref] [PubMed]
19. Coppola D, Hyacinthe M, Fu L, et al. CD44V6 expression in human colorectal carcinoma. Hum Pathol 1998;29:627-35. [Crossref] [PubMed]
20. Gosens MJ, van Kempen LC, van de Velde CJ, et al. Loss of membranous Ep-CAM in budding colorectal carcinoma cells. Mod Pathol 2007;20:221-32. [Crossref] [PubMed]
21. Nanashima A, Yamaguchi H, Sawai T, et al. Expression of adhesion molecules in hepatic metastases of colorectal carcinoma: relationship to primary tumors and prognosis after hepatic resection. J Gastroenterol Hepatol 1999;14:1004-9. [Crossref] [PubMed]
22. Zlobec I, Gunthert U, Tornillo L, et al. Systematic assessment of the prognostic impact of membranous CD44v6 protein expression in colorectal cancer. Histopathology 2009;55:564-75. [Crossref] [PubMed]
23. Taira T, Ishii G, Nagai K, et al. Characterization of the immunophenotype of the tumor budding and its prognostic implications in squamous cell carcinoma of the lung. Lung Cancer 2012;76:423-30. [Crossref] [PubMed]
24. Liang F, Cao W, Wang Y, et al. The prognostic value of tumor budding in invasive breast cancer. Pathol Res Pract 2013;209:269-75. [Crossref] [PubMed]
25. Wang C, Huang H, Huang Z, et al. Tumor budding correlates with poor prognosis and epithelial-mesenchymal transition in tongue squamous cell carcinoma. J Oral Pathol Med 2011;40:545-51. [Crossref] [PubMed]
26. Grigore AD, Jolly MK, Jia D, et al. Tumor Budding: The Name is EMT. Partial EMT. J Clin Med 2016;5:51. [Crossref] [PubMed]
27. Puram SV, Tirosh I, Parikh AS, et al. Single-Cell Transcriptomic Analysis of Primary and Metastatic Tumor Ecosystems in Head and Neck Cancer. Cell 2017;171:1611-1624.e24. [Crossref] [PubMed]
28. Kohler I, Bronsert P, Timme S, et al. Detailed analysis of epithelial-mesenchymal transition and tumor budding identifies predictors of long-term survival in pancreatic ductal adenocarcinoma. J Gastroenterol Hepatol 2015;30:78-84. [Crossref] [PubMed]
29. Okado Y, Aoki M, Hamasaki M, et al. Tumor budding and laminin5-γ2 in squamous cell carcinoma of the external auditory canal are associated with shorter survival. Springerplus 2015;4:814-45. [Crossref] [PubMed]
30. Attramadal CG, Kumar S, Boysen ME, et al. Tumor budding, EMT and cancer stem cells in T1-2/N0 oral squamous cell carcinomas. Anticancer Res 2015;35:6111-20. [PubMed]
31. Meyer S, Zahnd S, Galván JA, et al. The search for an epithelial-mesenchymal phenotype in colorectal cancer reveals a subset of cytokeratin- positive/vimentin-positive cells derived from the tumor. Virchows Arch 2016;469:342-6.
32. Jensen DH, Dabelsteen E, Specht L, et al. Molecular profiling of tumour budding implicates TGFβ-mediated epithelial-mesenchymal transition as a therapeutic target in oral squamous cell carcinoma. J Pathol 2015;236:505-16. [Crossref] [PubMed]
33. De Smedt L, Palmans S, Andel D, et al. Expression profiling of budding cells in colorectal cancer reveals an EMT-like phenotype and molecular subtype switching. Br J Cancer 2017;116:58-65. [Crossref] [PubMed]
34. Luo WR, Gao F, Li SY, et al. Tumour budding and the expression of cancer stem cell marker aldehyde dehydrogenase 1 in nasopharyngeal carcinoma. Histopathology 2012;61:1072-81. [Crossref] [PubMed]
35. Jolly MK, Huang B, Lu M, et al. Towards elucidating the connection between epithelial-mesenchymal transitions and stemness. J R Soc Interface 2014;11:20140962. [Crossref] [PubMed]
36. Jolly MK, Jia D, Boareto M, et al. Coupling the modules of EMT and stemness: A tunable “stemness window” model. Oncotarget 2015;6:25161-74. [Crossref] [PubMed]
37. Goldman A, Majumder B, Dhawan A, et al. Temporally sequenced anticancer drugs overcome adaptive resistance by targeting a vulnerable chemotherapy-induced phenotypic transition. Nat Commun 2015;6:6139. [Crossref] [PubMed]
38. Bosch SL, Teerenstra S, de Wilt JH, et al. Predicting lymph node metastasis in pT1 colorectal cancer: a systematic review of risk factors providing a rationale for therapy decisions. Endoscopy 2013;45:827-34. [Crossref] [PubMed]
39. Van Cutsem E, Cervantes A, Adam R, et al. ESMO consensus guidelines for the management of patients with metastatic colorectal cancer. Ann Oncol 2016;27:1386-422. [Crossref] [PubMed]
40. Rogers AC, Gibbons D, Hanly AM, et al. Prognostic significance of tumor budding in rectal cancer biopsies before neoadjuvant therapy. Mod Pathol 2014;27:156-62. [Crossref] [PubMed]
41. Van den Eynden GG, Bird NC, Majeed AW, et al. The histological growth pattern of colorectal cancer liver metastases has prognostic value. Clin Exp Metastasis 2012.29:541-9.
42. Lazaris A, Amri A, Petrillo SK, et al. Vascularization of colorectal carcinoma liver metastasis: insight into the stratification of patients for anti-angiogenic therapies. J Pathol Clin Res 2018;4:184-92. [Crossref] [PubMed]
43. Lugli A, Karamitopoulou E, Panayiotides I, et al. CD8+ lymphocytes/tumor-budding index: an independent prognostic Impact of the Microenvironment on Tumour Budding in Colorectal Cancer factor representing a ‘pro-anti-tumor approach to tumor host interaction in colorectal cancer. Br J Cancer 2009;101:1382-92. [Crossref] [PubMed]
44. Casasent AK, Schalck A, Gao R, et al. Multiclonal Invasion in Breast Tumors Identified by Topographic Single-Cell Sequencing. Cell 2018;172:205-217.e12. [Crossref] [PubMed]
45. Ulase D, Heckl S, Behrens HM, et al. Prognostic significance of tumour budding assessed in gastric carcinoma according to the criteria of the International Tumour Budding Consensus Conference. Histopathology 2020;76:433-46. [Crossref] [PubMed]
46. Petrova E, Zielinski V, Bolm L, et al. Tumor budding as a prognostic factor in pancreatic ductal adenocarcinoma. Virchows Arch 2020;476:561-8. [Crossref] [PubMed]
47. Guo YX, Zhang ZZ, Zhao G, et al. Prognostic and pathological impact of tumor budding in gastric cancer: a systematic review and meta-analysis. World J Gastrointest Oncol 2019;11:898-908. [Crossref] [PubMed]
48. Kemi N, Eskuri M, Ikalainen J, et al. Tumor budding and prognosis in gastric adenocarcinoma. Am J Surg Pathol 2019;43:229-34. [Crossref] [PubMed]