Aim: In this study, it was aimed to investigate the prognostic importance of Tumor budding (TB) in Pancreatic ductal adenocarcinomas(PDAC) and its correlation with histopathological findings according to the International Tumor Budding Consensus Conference(ITBCC) grading. Material and Methods: A total of 75 patients diagnosed with PDAC were included in this study. The demographic features of the cases (age, sex) and the macroscopic features of the tumors (localization,size) were obtained from the electronic archive system. All Hematoxylin-Eosin-stained sections were re-evaluated in terms of differentiation, presence of lymphovascular (LVI) and perineural invasion(PNI), surgical margin positivity, primary tumor(pT), lymph node metastasis(LNM) and tumor budding. Statistically, Chi-square test, cox-regression and Kaplan-Meier test were performed. Results:Thirty four of the cases were female and 41 were male. The mean age was 64.21±9.71years. The degree of TB was TB-few in 17 cases, TB-moderate in 25cases, and TB-high in 33cases. LVI, PNI, LNM and TB-high were poor prognostic factors. Moreover, TB-high was related with poor differantiation,LVI,PNI,LNM and short survival time. Tumor budding was independent negative prognostic factor in multivariable model analyzes. Conclusion: ITBCC scoring can also be used in PDACs. In addition, high tumor budding was a poor prognostic feature and might be a target for tumor-specific treatments as it could be a predictive finding for the locally invasive character of the tumor. Evaluation and grading of TB thought to represent EMT may be a histological feature that can be used in tumor selection for advanced molecular methods to identify subtypes that may be associated with poor prognosis and drug resistance.
Keywords: Mesenchymal phenotype, pancreas carcinoma, tumor budding tumor-spesific treatment
How to cite this article:While pancreatic ductal adenocarcinoma (PDAC) is the seventh-highest cause of cancer deaths in the world, it ranks in the top four in developed countries.[1],[2] Many histopathological features that are effective in the prognosis are known (differentiation, perineural, and lymphovascular invasion, regional lymph node metastasis, surgical margin positivity, etc.).[3] Studies on the prognostic factors are ongoing to explain the pathogenesis of these aggressive tumors and to determine the targets for the therapies to be developed. Tumor budding (TB) is a type of invasive growth pattern observed on the histological examination of carcinomas. It appears as isolated tumor cells or as a small non-glandular tumor cell cluster and is generally believed to represent the epithelial-mesenchymal transition (EMT) of the tumors.[4],[5],[6] Although reported as an important prognostic factor in PDAC, it is currently not included in histopathology reports, unlike colon cancers, due to the lack of a standard scoring system.[7]
This study aimed to investigate the prognostic importance of TB in PDAC and its correlation with the histopathological findings according to the International Tumor Budding Consensus Conference (ITBCC) grading.
Material and MethodsCases' characteristics
All cases operated for pancreatic tumor in our institution between 2006 and 2019 were scanned from the electronic archives. Tumors without a histological diagnosis of PDAC, patients who received neoadjuvant therapy, patients whose paraffin blocks were not in the archive or unsuitable for use, and patients whose survival data could not be obtained, were excluded from the study. As a result, a total of 75 patients diagnosed with PDAC were included in this study. The demographic features of the patients (age, sex) and the macroscopic features of the tumors (localization, size) were obtained from the electronic archive system.
Pathological evaluation
Hematoxylin-eosin-stained sections of all patients were re-evaluated by at least two pathologists in terms of the differentiation, lymphovascular and perineural invasion, surgical margin positivity, primary tumor (pT), regional lymph nodes (pN), and TB.
The localization was classified as head, body, and tail. The World Health Organization 2018 Classification of Tumors of the Digestive System was used for the histological differentiation grade, pT, pN, and stage.[8] The presence of macroscopic and/or microscopic tumor at the surgical margin was considered surgical margin positivity. The other parameters were classified as present or absent.
TB was defined according to the ITBCC as isolated tumor cells or small non-glandular tumor cell cluster.[4] In this study, the tumor buds in at least five of the densest areas were counted at X20 magnification (area 0.785 mm2) and the average was calculated. TB was classified into three groups; TB-few: one to four buds, TB-moderate: five to nine buds, and TB-high: ≥10 buds.
Survival information and statistical analysis
The prognostic information was obtained from the archive records of the Local Cancer Monitoring and Follow-up Center. The survival time was calculated from the date of resection to the date of death or the date of the latest follow-up. The follow-up data of the patients were updated in August 2020. In a statistical analysis, the Chi-square test, Fisher's exact test, Cox-regression test, and Kaplan–Meier test were performed with the SPSS® software version 21.0. A P value of ≤0.05 was considered statistically significant.
Ethics and institutional review board approval
All procedures performed in the current study were approved by the National Research Ethics Committee (Reference Number 02, Date: April 2020) under the 1964 Helsinki Declaration and its later amendments. All authors have declared that no conflict of interest can influence the content of the papers and they participate in the research and/or article preparation.
ResultsThirty-four (% 45.33) of the cases were females and 41 (% 54.67) were males. The mean age was 64.21 ± 9.71 and the median age was 65.00 (min: 37, max: 85). While the mean of the largest sizes of the tumors was 3.19 ± 1.25 cm, the median size was 3 cm (min: 1.2, max 6.5 cm). Statistically, the age and tumor size were not correlated with survival (P = 0.106 and P = 0.750, respectively).
In the Kaplan–Meier survival analysis, the mean overall survival time was found to be 21.65 ± 3.29 months (95% confidence interval: 19.852–40.699). The cumulative proportion surviving at the 12th, 24th, 60th months was 51.7, 34.4, and 12.0%, respectively.
The distribution of the number of cases according to histopathological features and the mean overall survival times are shown in [Table 1].
Table 1: The distribution of the number of cases according to the histopathological features and their relationship with the mean overall survival timesTB was observed in all the cases included in our study. The degree of budding was TB-few in 17 (22.7%) cases [Figure 1]a, TB-moderate in 25 (%33.3) cases [Figure 1]b, and TB-high in 33 (%44.0) cases [Figure 1]c, [Figure 1]d. When the relationship between TB and other histopathological features was investigated statistically, no correlation between the TB with surgical margin positivity (P = 0.465), pT (P = 0.615), and stage (P = 0.138) was found. TB-few was observed in 50% of the tumors located in the body or tail of the pancreas and in 20.3% of the head localized. However, this finding also was not significant statistically (P = 0.239). TB-high budding was observed in 75% of the poorly differentiated tumors, while this rate was 41.8 and 12.5% in the moderately differentiated and well-differentiated tumors, respectively; 57.9% of the tumors with Lymphovascular invasion (LVI) and 29.7% of the tumors without LVI were TB-high. Similarly, 87.9% of TB-high tumors, 88.0% of TB-moderate tumors, and 41.2% of TB-few tumors had perineural invasion. Moreover, no lymph node metastasis was observed in 70.6% of the tumors with TB-few, 44.0% of the tumors with TB-moderate, and 24.2% of TB-high tumors. The statistical analysis showed that high TB is associated with poor differentiation (P = 0.038), presence of LVI (P = 0.004) [Figure 2]a, presence of Perineural Invasion (PNI) (P < 0.0001) [Figure 2]b, and lymph node metastasis (P = 0.030). In the multivariable model adjusting for LVI, PNI, Lymph Node Metastasis (LNM), and budding, TB > 5 buds/0.785 mm2 was an independent negative prognostic factor (P = 0.031; hazard ratio: 0.543 [95% confidence interval: 0.312–0.944]). Statistically, TB was not correlated with age and tumor size (P > 0.05).
Figure 1: Examples of tumor budding (hematoxylin and eosin). (a) Pancreatic ductal adenocarcinoma (PDAC) with low tumor budding (TB-few: one to four buds); (b) PDAC with intermediate tumor budding (TB-moderate: five to nine buds); (c– d) PDAC with high tumor budding (TB-high; ≥ 10 buds)Figure 2: Histological appearance in cases with prominent tumor budding with lymphovascular invasion and perineural invasion. The buds are indicated by arrows (a) lymphovascular invasion, X20, Hematoxylin-Eosin; (b) perineural invasion, X20, Hematoxylin-EosinThe relationship of TB with demographic and histopathological findings is shown in [Table 2].
Table 2: The relationship of tumor budding with the demographic and histopathological findings DiscussionThe incidence and mortality of pancreatic cancer have been increasing over the years and it is predicted that it will be in the top three in cancer-related deaths in the near future.[9],[10] In addition to the difficulties in the early diagnosis of the disease, although the treatments had serious side effects, they could prolong the survival by only a few months, are some factors responsible for the poor p rognosis of the disease.[11],[12] The overall 5-year survival rate is approximately 5–6%, especially in developed countries. 10 For PDAC, the treatment options and using neoadjuvant chemotherapy vary.[13],[14] Following the benefits of immunotherapies in various tumors, the search for the target molecules for the treatment of pancreatic cancer has gained momentum.[13],[15] As the molecular mechanisms of PDAC are elucidated, individual-targeted treatment options are being tried to be created.[16],[17] It is known that tumor cells can perform local invasion and metastasis using the EMT program.[18] EMT was defined by Thiery et al.[19] as a process that occurs as a result of the loss of cellular polarity with the disruption of connections between epithelial cells and their transformation into mesenchymal cells with highly mobile and invasive properties. The fact that cells displaying EMT-like features are frequently observed in the invasive front of tumors also supports the thesis that EMT is an important step in cancer invasion and metastasis.[16],[18],[20],[21] TB is a type of invasive growth pattern for carcinomas and is generally believed to represent the EMT of tumors.[4],[5],[6] It is recommended that TB be evaluated histopathologically on H&E-stained sections, in the region where the tumor mass is in close contact with the host microenvironment.[18],[21],[22],[23] In our study, in accordance with this idea accepted in the literature, TB was evaluated by examining the H&E-stained sections under light microscopy and no additional application was required. We concluded that a careful examination of the areas where the tumor is closely related will be sufficient to evaluate the TB.
In our study, we observed, like some studies in the literature, that TB is a morphological trait representing EMT, defined as a biological process that enables the epithelial cells to acquire the so-called “mesenchymal phenotype.”[24] This phenotype has resistance to invasion, metastasis, and apoptosis by rearrangement of the cytoskeleton and surface proteins. It is also known to contain properties that cause loss of cell polarity and downregulation of e-cadherin.[24],[25] The mesenchymal phenotype may cause pancreatic cancers to develop resistance, particularly to cytotoxic agents that induce apoptotic cell death. It is known that mesenchymal-like cells in pancreatic cancer are more resistant to gemcitabine, 5-fluorouracil (5-FU), cisplatin, and epidermal growth factor receptor (EGFR) inhibitors associated with cellular growth and viability.[26],[27],[28] TB, which can help in the evaluation of the relationship between drug resistance, which is effective in the poor prognosis of pancreatic cancers and tumor phenotype, gains great importance in this sense. Undoubtedly, evaluations with appropriate molecular techniques are essential in tumor subtyping. However, evaluation of TB with light microscopy may be useful in choosing which tumors to apply these expensive and laborious methods. The fact that high TB was associated with a short survival time in our study and a positive correlation with histological findings showing tumor aggressiveness such as LVI, PNI, and lymph node metastasis shows that this finding can be used in patient selection for aggressive phenotype determination. The phenotype relationship of TB can be revealed in more detail with studies to be conducted in large series.
Perineural invasion is defined histopathologically in the vast majority (70–98%) of pancreatic cancers, and it has been claimed that perineural invasion can be found in 100% of pancreatic cancers if sufficient sections are examined.[29],[30] The poor prognostic effect of PNI in PDACs has long been known.[31],[32] After the neoplastic cells enter the perineural space, they can spread along the nerves. This situation was thought to be an exit route from the pancreas, and it has been reported that perineural invasion in pancreatic cancers is associated with a shorter survival time and retroperitoneal invasion, which may cause recurrence after surgery.[33],[34],[35] In PDAC, TB is generally reported to be present in 100% of the samples.[36] In our study, TB was observed in all cases in line with this finding. Considering that tumors can use PNI as a tool in the local spread, the question that comes to mind is “Is there a relationship between these two histological features?” This relationship has not been the subject of much research. In 2010, Masugi et al.[37] reported that they could not find statistical significance in their study in which they investigated the relationship between TB and PNI, which they defined as solitary cell infiltration (SoCI). When the details of this study were examined, 64% of the cases with PNI had a high degree of SoCI, but it can be seen that the dual grading system, in which more than seven buds are cut-off, is used for high SoCI. In a study performed in a large series in the following years, O'Connor et al.[38] reported that high TB was associated with PNI. In our study, patients with tumors with PNI had a shorter survival time, and this finding is statistically significant. Moreover, there is a statistically positive correlation between TB and PNI. Our study is one of the current studies showing the relationship between these two histological findings, which has not been investigated much.
Other poor prognostic features reported for PDACs are poor-differentiated tumor, the presence of LVI, and LNM.[39],[40],[41],[42],[43] LVI and LNM, which can be associated with local spread, stand out as poor prognostic features in our study. However, no relationship was found between tumor differentiation and survival time. In fact, the survival time shortened as the differentiation worsened, but the imbalance in the distribution of the case numbers in the groups hindered statistically significant results. The results of the studies on the relationship between LVI and LNM of TB in PDACs are conflicting. TB in colon cancer has is positively associated with lymph node metastasis.[44] A similar relationship has been reported between TB and LNM in PDACs.[36],[45] In our study, while most of the TB-few group cases were pN0, lymph node metastasis was found in most of the TB-high group, and this finding is statistically significant. Our results support the thesis that TB is a finding that indicates tumor aggressiveness.
TB grading is now recognized as an important prognostic marker in a variety of solid tumors.[4],[46],[47],[48],[49] Since TB is an independent prognostic parameter in colorectal cancer (CRC) and affects treatment options, consensus has been provided for reporting TB grading in colorectal cancer.[4] Although many studies have shown that TB is an independent prognostic factor in PDAC, unfortunately, there is no consensus on the necessity of inclusion in the pathology reports for these tumors.[6],[7],[36],[45],[50]
In conclusion, our study showed that ITBCC scoring can be easily applied in the evaluation of TB in operated PDACs. In addition, high TB is a poor prognostic feature as well as being associated with LVI, PNI, LNM, and poor differentiation, suggesting that it can be a predictive finding for the locally invasive character of the tumor. We concluded that the evaluation and grading of TB, which is thought to represent EMT, can be used to select tumors for advanced molecular methods to identify the subtypes that may be associated with poor prognosis and drug resistance. The development of new targeted therapies that can remove PDACs from the group of “lethal tumors” is essential, and one of the issues that need to be addressed and investigated first is TB.
Author contributions
Argon A: Conceptualization, Methodology, Software; Argon A, Kebat T: Data curation; Argon A, Öz Ö: Original draft preparation; Argon A: Visualization, Investigation; Argon A: Supervision, Writing- Reviewing and Editing.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
Correspondence Address:
Asuman Argon
University of Health Sciences, Izmir Faculty of Medicine, Department of Pathology, İzmir
Turkey
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/ijpm.ijpm_905_21
留言 (0)