Advertisement

Tumor cell budding in preoperative biopsies of esophageal and gastroesophageal junction carcinoma independently predicts survival in a grade-dependent manner

Open AccessPublished:April 11, 2022DOI:https://doi.org/10.1016/j.surg.2022.02.020

      Abstract

      Background

      Tumor budding is a prognostic factor in biopsies of different tumor entities. Recent evidence suggests that this also applies to esophageal squamous cell carcinomas. Since esophageal cancer is diagnosed by biopsy, the aim of this study was to investigate whether tumor budding in pretherapeutic biopsies of a mixed tumor population of the esophagus and gastroesophageal junction might predict survival.

      Methods

      In this retrospective analysis, samples of 78 patients were analyzed (55 adenocarcinomas, 17 squamous cell carcinomas, 5 adenosquamous carcinomas, 1 carcinosarcoma). In addition to preoperative biopsies, budding foci in corresponding resection specimens were assessed and related to overall and relapse-free survival.

      Results

      The main finding was that the number of budding foci in preoperative biopsies predicted overall survival independent of the patient’s age and disease stage in a grade-specific (P = .009) manner. In patients with grade 2 tumors, each additional budding focus was associated with an increased chance of death by a factor of 1.28 (hazard ratio 95% confidence interval 1.06–1.55, P = .011). There was no significant association between survival and the number of budding foci in patients with grade 3 tumors, and no budding was observed in grade 1 tumors. Budding foci in resection specimens also showed a certain association with survival, but to a lesser degree.

      Conclusion

      Budding foci in preoperative biopsies might serve to improve prognostic accuracy in esophageal carcinomas.

      Graphical abstract

      Introduction

      Esophageal cancer is the seventh leading cause of cancer-related morbidity and the sixth leading cause of death from cancer.
      • Arnold M.
      • Abnet C.C.
      • Neale R.E.
      • et al.
      Global burden of 5 major types of gastrointestinal cancer.
      The 2 most common histological subtypes—adenocarcinoma (AC) and squamous cell carcinoma (SCC)—account for about 90% to 95% of all esophageal cancers.
      • Arnold M.
      • Abnet C.C.
      • Neale R.E.
      • et al.
      Global burden of 5 major types of gastrointestinal cancer.
      ,
      • Malhotra G.K.
      • Yanala U.
      • Ravipati A.
      • Follet M.
      • Vijayakumar M.
      • Are C.
      Global trends in esophageal cancer.
      Since early esophageal cancer has no specific symptoms, most tumors are diagnosed at a late stage, when therapeutic options are limited and curative treatment is usually unfeasible.
      • Arnold M.
      • Abnet C.C.
      • Neale R.E.
      • et al.
      Global burden of 5 major types of gastrointestinal cancer.
      However, prognostic accuracy is still poor. In this regard, tumor budding is a promising biomarker candidate that might serve to optimize the therapeutic strategy.
      It is defined as a single tumor cell or a cluster of tumor cells consisting of up to 4 cells within the tumor or at its invasion front. Based on the recommendations of the International Tumor Budding Consensus Conference 2016, which refer to colorectal carcinomas, tumor budding should be assessed on the hematoxylin and eosin (H & E)-stained slide(s) in a hotspot field at the invasion front. To ease risk stratification, a 3-tiered classification is recommended, although a continuous scale allows for a more precise risk assessment.
      • 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.
      Despite the use of different scoring methods in many studies, tumor budding is considered prognostic in carcinomas of various histological subtypes and origins.
      • Rogers A.C.
      • Winter D.C.
      • Heeney A.
      • et al.
      Systematic review and meta-analysis of the impact of tumour budding in colorectal cancer.
      • Ueno H.
      • Ishiguro M.
      • Nakatani E.
      • et al.
      Prospective multicenter study on the prognostic and predictive impact of tumor budding in stage II colon cancer: results from the SACURA trial.
      • van Wyk H.C.
      • Roseweir A.
      • Alexander P.
      • et al.
      The relationship between tumor budding, tumor microenvironment, and survival in patients with primary operable colorectal cancer.
      • Lino-Silva L.S.
      • Salcedo-Hernández R.A.
      • Gamboa-Domínguez A.
      Tumour budding in rectal cancer: a comprehensive review.
      • Demir A.
      • Alan O.
      • Oruc E.
      Tumor budding for predicting prognosis of resected rectum cancer after neoadjuvant treatment.
      • Guo Y.X.
      • Zhang Z.Z.
      • Zhao G.
      • Zhao E.H.
      Prognostic and pathological impact of tumor budding in gastric cancer: a systematic review and meta-analysis.
      • Kemi N.
      • Eskuri M.
      • Ikäläinen J.
      • Karttunen T.J.
      • Kauppila J.H.
      Tumor budding and prognosis in gastric adenocarcinoma.
      • Ohike N.
      • Coban I.
      • Kim G.E.
      • et al.
      Tumor budding as a strong prognostic indicator in invasive ampullary adenocarcinomas.
      • Lawlor R.T.
      • Veronese N.
      • Nottegar A.
      • et al.
      Prognostic role of high-grade tumor budding in pancreatic ductal adenocarcinoma: a systematic review and meta-analysis with a focus on epithelial to mesenchymal transition.
      • Berg K.B.
      • Schaeffer D.F.
      Tumor budding as a standardized parameter in gastrointestinal carcinomas: more than just the colon.
      • Tanaka M.
      • Yamauchi N.
      • Ushiku T.
      • et al.
      Tumor budding in intrahepatic cholangiocarcinoma: a predictor of postsurgery outcomes.
      • Boxberg M.
      • Kuhn P.H.
      • Reiser M.
      • et al.
      Tumor budding and cell nest size are highly prognostic in laryngeal and hypopharyngeal squamous cell carcinoma.
      • Kale A.D.
      • Angadi P.V.
      Tumor budding is a potential histopathological marker in the prognosis of oral squamous cell carcinoma: current status and future prospects.
      • Mäkitie A.A.
      • Almangush A.
      • Rodrigo J.P.
      • Ferlito A.
      • Leivo I.
      Hallmarks of cancer: tumor budding as a sign of invasion and metastasis in head and neck cancer.
      • Zhu Y.
      • Liu H.
      • Xie N.
      • et al.
      Impact of tumor budding in head and neck squamous cell carcinoma: a meta-analysis.
      • Zare S.Y.
      • Aisagbonhi O.
      • Hasteh F.
      • Fadare O.
      Independent validation of tumor budding activity and cell nest size as determinants of patient outcome in squamous cell carcinoma of the uterine cervix.
      In esophageal carcinomas, numerous studies on resection specimens have shown that increased tumor budding predicts both shorter survival and time to recurrence. Moreover, the number of budding foci was correlated with lymph node metastases.
      • Berg K.B.
      • Schaeffer D.F.
      Tumor budding as a standardized parameter in gastrointestinal carcinomas: more than just the colon.
      ,
      • Koelzer V.H.
      • Langer R.
      • Zlobec I.
      • Lugli A.
      Tumor budding in upper gastrointestinal carcinomas.
      • Almangush A.
      • Karhunen M.
      • Hautaniemi S.
      • Salo T.
      • Leivo I.
      Prognostic value of tumour budding in oesophageal cancer: a meta-analysis.
      • Fuchinoue K.
      • Nemoto T.
      • Shimada H.
      • et al.
      Immunohistochemical analysis of tumor budding as predictor of lymph node metastasis from superficial esophageal squamous cell carcinoma.
      • Lohneis P.
      • Hieggelke L.
      • Gebauer F.
      • et al.
      Tumor budding assessed according to the criteria of the International Tumor Budding Consensus Conference determines prognosis in resected esophageal adenocarcinoma.
      Besides the resection specimens, where tumor budding can usually be assessed in plenty of tumor tissue, budding has already been examined in biopsies with far less available tumor tissue. It has been found that it has informative value—for example, for prognosis.
      • Almangush A.
      • Youssef O.
      • Pirinen M.
      • Sundström J.
      • Leivo I.
      • Mäkitie A.A.
      Does evaluation of tumour budding in diagnostic biopsies have a clinical relevance? A systematic review.
      • Dawson H.
      • Blank A.
      • Zlobec I.
      • Lugli A.
      Potential clinical scenarios of tumour budding in colorectal cancer.
      • Zengın M.
      • Çıfcı A.
      Tumour budding in preoperative biopsy specimens is a useful prognostic index for identifying high-risk patients in early-stage (pN0) colon cancer.
      However, to our knowledge, there are only 2 studies to date that have examined budding in preoperative biopsies of esophageal carcinomas, and they are limited to SCC.
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      ,
      • Jesinghaus M.
      • Boxberg M.
      • Wilhelm D.
      • et al.
      Post-neoadjuvant cellular dissociation grading based on tumour budding and cell nest size is associated with therapy response and survival in oesophageal squamous cell scarcinoma.
      The primary aim of this study was to test the hypothesis that tumor budding in pretherapeutic biopsies of esophageal and gastroesophageal junction carcinomas independently predicts survival in a grade-dependent manner. Second, we aimed to assess whether this also applies to tumor budding in resection specimens. Finally, we aimed to quantify the relationship of budding foci with tumor grade, stage, and lymph node metastases.

      Materials and methods

      Study design

      Seventy-eight patients with malignant tumors of the esophagus and the gastroesophageal junction in care of the Medical University of Vienna were included in this retrospective study. Prerequisite for inclusion was the availability of both a preoperative biopsy with a clearly invasive tumor and a resection specimen of the same tumor archived at the Department of Pathology. Exclusion criteria were age below 18 years and lack of required clinical data.
      H & E-stained tumor sections of the samples available from 2009 to 2018, which were routinely fixed in 4% neutral formaldehyde and embedded in paraffin, were assessed, and the section with the highest budding content was selected. Tumor entity and tumor grading were re-evaluated, and pathological staging was adjusted to the current version of the TNM classification of malignant tumors according to UICC (eighth edition, 2017); separate stage evaluations were done for AC and SCC. Clinical data were extracted from a database of the Clinical Division of General Surgery; where possible, patients were followed up until June 2019. No formal sample size calculation was performed. The Ethics Committee of the Medical University of Vienna approved this study (study number 2194/2018). Methods were carried out in accordance with approved ethical guidelines and regulations. Individual informed consent was waived by the Ethics Committee of the Medical University of Vienna, Austria.

      Clinical endpoint

      Overall survival

      Overall survival (OS) was defined as the days elapsing between the biopsy and death, that is, an event, or the last patient contact, that is, censoring.

      Analysis of tumor budding

      On H & E-stained slides, the total number of tumor buds was analyzed in resection specimen in one high-power-field (HPF) centrally in the tumor (intratumoral budding) and in one HPF at the invasion front (peritumoral budding), both in a hotspot area. In biopsies, where discrimination into intra or peritumoral budding is not possible due to lack of topographic orientation within the tumor, only one hotspot HPF was analyzed. For this purpose, primarily all available tumor slides were screened at 200× magnification for the area with the most buds. The evaluation was done in pairs of one pathologist and one medical student (ABe/AR and AIS/AR) on a multiheaded microscope. One HPF corresponded to an area of 0.24 mm.
      • Malhotra G.K.
      • Yanala U.
      • Ravipati A.
      • Follet M.
      • Vijayakumar M.
      • Are C.
      Global trends in esophageal cancer.
      The evaluation was done blinded for clinicopathological data and outcome.

      Statistical analysis

      To analyze whether the number of budding foci was associated with survival duration, log rank tests and Cox regression models were applied. To assess associations between different types of budding foci, and between budding foci and positive lymph nodes, regression analyses were used. To investigate associations between different types of budding foci, and between budding foci and positive lymph nodes, linear regression analyses were applied. Statistical analyses were carried out with IBM SPSS Statistics 26. Graphs were generated using GraphPad Prism 8.4. Adjustment for multiple testing was not performed; therefore, the results have to be interpreted accordingly. The reported P values are the result of two-sided tests. A detailed description of the statistical methods can be found in the Supplementary Materials.

      Results

      Patient and tumor characteristics

      Patient and tumor characteristics are presented in Table I.
      Table IBasic demographic data, treatment modalities, outcome, and tumor-related details
      CategoriesTotal n (%)MeanSD
      Number of patients78 (100)
      Sex
       Female18 (23.1)
       Male60 (76.9)
      Age61.9011.25
      Neoadjuvant therapy
       No27 (34.6)
       Yes51 (65.4)
      Relapse
       No45 (57.7)
       Yes33 (42.3)
      OS event
       No38 (48.7)
       Yes40 (51.3)
      Tumor type
       Adenocarcinoma55 (70.5)
       Squamous cell carcinoma17 (21.8)
       Adenosquamous carcinoma5 (6.4)
       Carcinosarcoma1 (1.3)
      Lymph node metastases
       No35 (44.9)
       Yes43 (55.1)
      Grade–Biopsy
       No tumor0 (0)
       12 (2.6)
       246 (59.0)
       330 (38.5)
      Grade–Resection specimen
       No tumor6 (7.7)
       10 (0)
       226 (33.3)
       346 (59.0)
      UICC
      Eighth edition, 2017.
      –Pathological staging
       No tumor6 (7.7)
       Ia3 (3.8)
       Ib16 (20.5)
       IIa5 (6.4)
       IIb2 (2.6)
       IIIa9 (11.5)
       IIIb19 (24.4)
       IVa15 (19.2)
       IVb3 (3.8)
      Budding Activity
      According to Jesinghaus et al.27
      –Biopsy
       No budding23 (29.5)
       Low budding35 (44.9)
       High budding20 (25.6)
      Budding activity—intratumoral
       No budding18 (23.1)
       Low budding34 (43.6)
       High budding26 (33.3)
      Budding activity—peritumoral
       No budding18 (23.1)
       Low budding29 (37.2)
       High budding31 (39.7)
      OS, overall survival; SD, standard deviation.
      According to Jesinghaus et al.
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      Eighth edition, 2017.

      Incidence of tumor buds in biopsies and resection specimens

      To allow comparison with previous research,
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      we categorized budding foci in biopsies and intra or peritumoral foci in resections into (1) no budding, (2) 1 to 4 budding foci, and (3) ≥5 budding foci. With this categorization, 23 patients (29.49%) had 0, 35 (44.87%) had 1 to 4 (low budding), and 20 (25.64%) had ≥5 buds (high budding) in their biopsies (Table I).
      In the resected specimens, 18 patients (23.08%) had no intra and peritumoral budding, low intratumoral budding was observed in 34 (43.59%) and peritumoral budding in 29 (37.18%), 26 patients (33.33%) had high budding intratumorally, and 31 (39.74%) peritumorally (Table I). H & E-stained slides showing low and high tumor budding are depicted in Figure 1.
      Figure thumbnail gr1
      Figure 1Histopathological characteristics of tumor budding in biopsies and corresponding resection specimens of esophageal squamous cell and gastroesophageal adenocarcinoma. Tumor budding in biopsies of a squamous cell carcinoma (A, B) and a gastroesophageal adenocarcinoma (E, F) each with the corresponding resection specimen (squamous cell carcinoma: C, D, adenocarcinoma: G,H) at intermediate (200×, A, C, E, G) and high magnification (600×, B, D, F, H). Tumor buds are highlighted by a circle.

      Association of tumor budding with overall survival

      Without adjustment for confounders, OS did not differ between groups with different budding numbers, either within biopsies or in resection specimens (Figure 2). In line with this, age-adjusted analyses using budding foci as continuous predictor showed no evidence for an association with OS.
      Figure thumbnail gr2
      Figure 2Association of the number of budding foci within the biopsy or within the resection of esophageal and gastroesophageal junction carcinomas with overall survival. (A) Kaplan-Meier estimates for patients grouped according to the number of budding foci within the biopsy. Plots are primarily shown to facilitate direct comparison with previous research.
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      Accordingly, cut-off values were chosen for consistency with this research. (B) Patient age-adjusted estimated hazard ratio change associated with increased budding foci number within biopsy. The error band represents 95% CI limits. (C) Estimated survival curves corresponding to the points on the hazard ratio curve (B) indicated by lowercase letters. (D) Kaplan-Meier estimates for patients grouped according to the number of intratumoral budding foci in the resection. (E) Patient age-adjusted estimated hazard ratio change associated with increased budding foci number within the resected tumor. (F) Estimated survival curves corresponding to the points on the hazard ratio curve (E) indicated by lowercase letters. (G) Kaplan-Meier estimates for patients grouped according to the number of peritumoral budding foci in the resection. (H) Patient age-adjusted estimated hazard ratio change associated with increased budding foci number at the invasion front of the resected tumor. (I) Estimated survival curves corresponding to the points on the hazard ratio curve (H) indicated by lowercase letters. CI, Confidence interval; Exp(B), Hazard Ratio.

      Prognostic value of the number of budding foci in biopsies adjusted for established prognostic markers

      The number of budding foci within preoperative biopsies significantly predicted OS in an analysis adjusted for age, stage, and grade (Figure 3).
      Figure thumbnail gr3
      Figure 3Prognostic value of the number of budding foci within biopsies of esophageal and gastroesophageal junction carcinomas for overall survival, adjusted for the patient’s age, stage, and grade, estimated by multivariable Cox regression models. (A) Overview of the primary multivariable regression model and its 2 blocks. Block 1 contains the known prognostic variables patient age, disease stage, and tumor grade. Block 2 additionally contains the number of budding foci within biopsies, and a “Budding foci × grade” interaction term, which tests whether the prognostic value of budding foci is grade-specific. Note that the number of budding foci was used in its metric form to avoid arbitrary categorization. The P value “Change from block 1 to block 2” refers to the hypothesis that taking into account budding foci in a grade-specific manner improves the overall prognostic accuracy of the model. (B) and (C) aim to visualize this improvement and show how accurately survival can be predicted using the previously established prognostic variables alone (B) or together with budding (C). To unravel the grade-specific prognostic value of budding foci, it was analyzed separately for grades 2 and 3. Grade 1 was omitted, since no budding foci were observed in grade 1 biopsies. (D) Multivariable Cox regression model estimating the change of survival associated with more budding foci within biopsies with grade 2. (E) Estimated age- and stage-adjusted hazard ratio change associated with more budding foci in grade 2 biopsies with 95% confidence interval. (F) Estimated survival functions for the points a-d shown in (E). Note that the survival function changes continuously with the number of budding foci, and there are no groups to be compared, as no categorization was performed. (G, H, I) Analogous plots to D, E, and F for grade 3. CI, Confidence interval; Exp(B), Hazard Ratio.
      In the first part of the analysis, age, stage, and grade were used to predict survival (Figure 3, A, block 1, Figure 3, B). In the second part (Figure 3, A, block 2), the number of budding foci was added in a grade-dependent manner, statistically reflected by the budding foci × grade interaction, which tests whether the prognostic value of budding foci is grade-specific, that is, whether budding foci are predictive of survival in one grade, but not in another grade. This addition resulted in a significant (P = .032) improvement of discrimination of patients with good and bad prognosis (Figure 3, C), confirming the primary hypothesis. The budding foci × grade interaction (P = .009) indicates that the relationship between the number of budding foci and the probability of death depends on grading. In patients with grade 2 tumors (Figure 3, D–F), each additional budding focus in their preoperative biopsy was associated with an increased chance of death during the observational period by a factor of 1.28 (Exp(B) = hazard ratio). In contrast, in patients with grade 3 tumors, the number of budding foci was not significantly associated with altered survival (Figure 3, G–I).
      Since grade 1 was only diagnosed in 2 biopsy specimens, which showed no budding, a corresponding estimate was not possible. A grade-dependent prediction of budding foci on OS was also observed when only AC were analyzed (P = .012, Supplementary Figure S3, A–C). In SCC the statistical power was considered to be too low due to the smaller patient number and the lack of variability regarding the number of buds (Supplementary Figure S3, D). However, there is no evidence that our hypothesis does not apply for SCC.

      Prognostic value of the number of budding foci in resection specimens for overall survival

      Adding the number of peri or intratumoral budding foci in a grade-specific manner into survival analyses including the factors age, stage, grade, and neoadjuvant therapy significantly improved the prognostic accuracy of these statistical models regarding OS (each with P = .011, Supplementary Figure S1 and S2, A–C). Also, the analyses suggested grade-dependent effects (peritumoral budding foci × grade interaction: P = .018; intratumoral budding foci × grade interaction: P = .02). However, in separate analyses within each grade, there was no evidence for a budding foci-dependent change in OS (Supplementary Figure S1 and S2, D–I).

      Correlation of tumor budding in biopsies and resection specimens: Correlation of intra and peritumoral buds

      When we correlated the number of budding foci found in biopsies with those in the resection specimens, we observed moderate, nonlinear associations (P < .0001 and r = 0.632 intratumorally, Figure 4, A; P < .0001 and r = 0.631 peritumorally, Figure 4, B). There was a high, linear association between intratumoral and peritumoral buds in the resected tumors (P < .0001 and r = 0.84, Figure 4, C).
      Figure thumbnail gr4
      Figure 4Relationship of different budding foci with each other and with established prognostic variables. (A) Moderate, nonlinear association between budding foci within biopsies and budding foci within subsequently resected tumors. The area of each circle is proportional to the number of patients with the respective x-y-value combination. The error band represents 95% confidence interval limits of the regression line. (B) Moderate, nonlinear association between budding foci within biopsies and budding foci at the invasion front. (C) Strong, linear association between intra and peritumoral budding foci. (D, E, F) Dependency of the number of budding foci in biopsies and resections on stage. (G, H, I) Dependency of the number of budding foci on grade. (J) No relevant association between the number of intratumoral budding foci and lymph nodes positive for malignant cells. (K) Weak, nonlinear association between peritumoral budding foci and positive lymph nodes. (L) Significant yet irrelevant association between the number of budding foci in biopsies and positive lymph nodes.

      Correlation of tumor budding with stage, grade and lymph node metastases

      We observed a weak but significant correlation of the amount of tumor buds with tumor stage in resected specimens (both intra und peritumorally; Figure 4, D and 4, E) and in the biopsies (Figure 4, F). Also, the bud number was highly dependent on tumor grading (Figure 4, G–I). When correlating the number of buds with the number of lymph node metastases, a weak positive but significant correlation was seen only in peritumoral buds in the resected specimens (P < .0001 and r = 0.44; Figure 4, K).

      Discussion

      Outcome for patients with esophageal and gastroesophageal junction carcinoma is poor, and the pretherapeutic treatment decision is based mainly on clinical staging and patient performance status. In this study we found evidence that a higher number of tumor buds in preoperative biopsies is associated with a worse prognosis in patients with grade 2 but not with grade 3 tumors, even after adjustment for patient age and disease stage. Thus, tumor buds in preoperative biopsies might serve to improve the prognostic accuracy in the future, allowing for patient stratification and optimized treatment regimens.
      With our focus on biopsy specimens we decided to evaluate budding in the hotspot HPF on H & E-stained slides according to the recommendation of WHO for evaluation of tumor buds in AC of the colon.
      WHO Classification of Tumours Editorial Board
      Digestive System Tumours.
      Variable cut-off levels and budding subclassifications based on subjective valuation have been used by various authors.
      • 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.
      Since we aimed to compare our results with data from a recently published study
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      of tumor buds in esophageal biopsies, we categorized tumor budding accordingly into no, low (1–4 buds), and high budding (≥5 buds). For subsequent approaches, we calculated tumor buds as a continuous, noncategorized variable to prevent loss of information and hazard of false-positive results.
      Unlike the data presented by Jesinghaus et al,
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      ,
      • Jesinghaus M.
      • Boxberg M.
      • Wilhelm D.
      • et al.
      Post-neoadjuvant cellular dissociation grading based on tumour budding and cell nest size is associated with therapy response and survival in oesophageal squamous cell scarcinoma.
      we were not able to show a significant association of tumor buds with survival when analyzed in a univariable categorized model. Patients without budding showed a nonsignificantly higher probability compared to those with any budding. Intriguingly, the extent of tumor budding with cut-off levels as described above revealed no discriminatory effect, either in the biopsy or in the resection specimen. One possible explanation for these divergent results may be the heterogeneity of tumor samples: We included different kinds of esophageal and gastroesophageal junction carcinomas, whereas Jesinghaus et al only analyzed SCC. The role of tumor budding as a predictive biomarker has been established in both AC and SCC.
      • Lino-Silva L.S.
      • Salcedo-Hernández R.A.
      • Gamboa-Domínguez A.
      Tumour budding in rectal cancer: a comprehensive review.
      ,
      • Demir A.
      • Alan O.
      • Oruc E.
      Tumor budding for predicting prognosis of resected rectum cancer after neoadjuvant treatment.
      ,
      • Kemi N.
      • Eskuri M.
      • Ikäläinen J.
      • Karttunen T.J.
      • Kauppila J.H.
      Tumor budding and prognosis in gastric adenocarcinoma.
      ,
      • Ohike N.
      • Coban I.
      • Kim G.E.
      • et al.
      Tumor budding as a strong prognostic indicator in invasive ampullary adenocarcinomas.
      ,
      • Kale A.D.
      • Angadi P.V.
      Tumor budding is a potential histopathological marker in the prognosis of oral squamous cell carcinoma: current status and future prospects.
      • Mäkitie A.A.
      • Almangush A.
      • Rodrigo J.P.
      • Ferlito A.
      • Leivo I.
      Hallmarks of cancer: tumor budding as a sign of invasion and metastasis in head and neck cancer.
      • Zhu Y.
      • Liu H.
      • Xie N.
      • et al.
      Impact of tumor budding in head and neck squamous cell carcinoma: a meta-analysis.
      Furthermore, both carcinoma types are diagnosed by primarily biopsy and can be treated with neoadjuvant chemotherapy or radiochemotherapy when staged cT2 or higher.
      Oncology Guideline Program (German Cancer Society, German Cancer Aid, AWMF)
      S3 Guideline diagnosis and therapy of squamous cell carcinoma and adenocarcinoma of the esophagus, long version 2.0, 2018, AWMF registration number: 021/023OL.
      Thus, the feasibility of evaluation of tumor budding in esophageal biopsies is relevant and subject of our study and applies to both AC and SCC.
      In the biopsy samples, a bivariable model adjusted for age revealed a nonsignificant trend toward shortened OS with an increase of tumor bud foci.
      In a multivariable model, the number of tumor bud foci predicted survival independent of established prognostic factors, that is, patient’s age, disease stage, and tumor grade. Specifically, a significant influence of buds (number of budding foci and grade-dependent budding foci) in biopsies on OS was seen. Prognostic accuracy improved when tumor buds were incorporated in the Cox regression model. Therefore, a systematic inclusion of tumor buds, especially at the time of diagnosis, which is usually based on a biopsy, would improve the survival assessment for patients with esophageal and gastroesophageal junction carcinomas. For other tumor types, tumor budding has already been included in the grading system (within the framework of studies) with good success.
      • Jesinghaus M.
      • Strehl J.
      • Boxberg M.
      • et al.
      Introducing a novel highly prognostic grading scheme based on tumour budding and cell nest size for squamous cell carcinoma of the uterine cervix.
      • Elseragy A.
      • Salo T.
      • Coletta R.D.
      • et al.
      A proposal to revise the histopathologic grading system of early oral tongue cancer incorporating tumor budding.
      • Shi H.
      • Ye L.
      • Lu W.
      • Lu B.
      Grading of endocervical adenocarcinoma: a novel prognostic system based on tumor budding and cell cluster size.
      The fact that the independent prognostic impact of tumor buds is grade-specific and only existent in grade 2 carcinomas in our study needs to be confirmed in larger cohorts, before considering a general change in the grading system for gastroesophageal tumors. One explanation may be that the minimal differentiation is the crucial factor for survival prediction in grade 3 tumors and other variables cannot further stratify in contrast to grade 2 tumors where tumor buds can very well exhibit its influence on survival. However, grade per se was no independent survival factor in multivariable analyses. Notably, the amount of budding foci was higher in grade 3 tumors, comparable to data from a previous study of carcinomas of the esophagus and the gastroesophageal junction.
      • Brown M.
      • Sillah K.
      • Griffiths E.A.
      • et al.
      Tumour budding and a low host inflammatory response are associated with a poor prognosis in oesophageal and gastro-oesophageal junction cancers.
      In contrast to Jesinghaus et al,
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      ,
      • Jesinghaus M.
      • Boxberg M.
      • Wilhelm D.
      • et al.
      Post-neoadjuvant cellular dissociation grading based on tumour budding and cell nest size is associated with therapy response and survival in oesophageal squamous cell scarcinoma.
      the correlation of buds in biopsies with buds in corresponding resection specimens only revealed a moderate, nonlinear association. The lack of strong concordance might be caused by inclusion of different tumor types or the partial use of neoadjuvant therapy, which was administered to the majority of our patients (65.4%). The application of neoadjuvant therapy in the major fraction of our cohort might also be the reason why only the peritumoral buds of the resected specimens showed a weak but significant correlation with tumor infiltrated lymph nodes, whereas budding foci in biopsies had no relevant association.
      Although our results seem promising, they have to be interpreted in the light of some limitations. First, the inclusion of patients receiving neoadjuvant therapy in our cohort might be viewed as a limitation; however, this reflects the real-life situation. Furthermore, due to the limited sample size, a considerable amount of uncertainty concerning effect sizes remain. Related to this, our study does not include a validation cohort. Therefore, the grade-specific prognostic value of budding foci needs to be validated in different, ideally larger cohorts by independent researchers. Finally, only patients with complete data were selected for this retrospective study. Bias might have been introduced by omitting patients with incomplete data.
      In conclusion, our data nevertheless provide further evidence for the applicability of tumor budding as a prognostic variable. Despite an improvement in early detection with increased application of gastroesophagoscopy, gastroesophageal tumors are still often diagnosed at an advanced stage in which patients require neoadjuvant therapy, which often leads to pronounced regression. Therefore, the assessment of chemotherapy-naive tumors, and thus especially tumor biopsies, is of great importance. The feasibility of evaluation of tumor budding in biopsies and its consideration as prognostic biomarker is given, as shown by us, by Jesinghaus,
      • Jesinghaus M.
      • Brühl F.
      • Steiger K.
      • et al.
      Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
      and by other authors
      • Almangush A.
      • Youssef O.
      • Pirinen M.
      • Sundström J.
      • Leivo I.
      • Mäkitie A.A.
      Does evaluation of tumour budding in diagnostic biopsies have a clinical relevance? A systematic review.
      • Dawson H.
      • Blank A.
      • Zlobec I.
      • Lugli A.
      Potential clinical scenarios of tumour budding in colorectal cancer.
      • Zengın M.
      • Çıfcı A.
      Tumour budding in preoperative biopsy specimens is a useful prognostic index for identifying high-risk patients in early-stage (pN0) colon cancer.
      who investigated the utility of tumor buds in biopsies of other locations; therefore, it seems to be relevant to include tumor budding in biopsies for the planning of preoperative management.

      Funding/Support

      None of the authors received funding for this study.

      Conflicts of interest/Disclosure

      None of the authors has related conflicts of interest to declare.

      Acknowledgments

      We thank Renate Kain, MD, PhD, for critical revision of the article and Gertrude Krainz for proofreading of the article.

      Supplementary materials

      References

        • Arnold M.
        • Abnet C.C.
        • Neale R.E.
        • et al.
        Global burden of 5 major types of gastrointestinal cancer.
        Gastroenterology. 2020; 159: 335-349.e15
        • Malhotra G.K.
        • Yanala U.
        • Ravipati A.
        • Follet M.
        • Vijayakumar M.
        • Are C.
        Global trends in esophageal cancer.
        J Surg Oncol. 2017; 115: 564-579
        • 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: 1299-1311
        • Rogers A.C.
        • Winter D.C.
        • Heeney A.
        • et al.
        Systematic review and meta-analysis of the impact of tumour budding in colorectal cancer.
        Br J Cancer. 2016; 115: 831-840
        • Ueno H.
        • Ishiguro M.
        • Nakatani E.
        • et al.
        Prospective multicenter study on the prognostic and predictive impact of tumor budding in stage II colon cancer: results from the SACURA trial.
        J Clin Oncol. 2019; 37: 1886-1894
        • van Wyk H.C.
        • Roseweir A.
        • Alexander P.
        • et al.
        The relationship between tumor budding, tumor microenvironment, and survival in patients with primary operable colorectal cancer.
        Ann Surg Oncol. 2019; 26: 4397-4404
        • Lino-Silva L.S.
        • Salcedo-Hernández R.A.
        • Gamboa-Domínguez A.
        Tumour budding in rectal cancer: a comprehensive review.
        Wspolczesna Onkol. 2018; 22: 61-74
        • Demir A.
        • Alan O.
        • Oruc E.
        Tumor budding for predicting prognosis of resected rectum cancer after neoadjuvant treatment.
        World J Surg Oncol. 2019; 17: 1-8
        • Guo Y.X.
        • Zhang Z.Z.
        • Zhao G.
        • Zhao E.H.
        Prognostic and pathological impact of tumor budding in gastric cancer: a systematic review and meta-analysis.
        World J Gastrointest Oncol. 2019; 11: 898-908
        • Kemi N.
        • Eskuri M.
        • Ikäläinen J.
        • Karttunen T.J.
        • Kauppila J.H.
        Tumor budding and prognosis in gastric adenocarcinoma.
        Am J Surg Pathol. 2019; 43: 229-234
        • Ohike N.
        • Coban I.
        • Kim G.E.
        • et al.
        Tumor budding as a strong prognostic indicator in invasive ampullary adenocarcinomas.
        Am J Surg Pathol. 2010; 34: 1417-1424
        • Lawlor R.T.
        • Veronese N.
        • Nottegar A.
        • et al.
        Prognostic role of high-grade tumor budding in pancreatic ductal adenocarcinoma: a systematic review and meta-analysis with a focus on epithelial to mesenchymal transition.
        Cancers (Basel). 2019; 11
        • Berg K.B.
        • Schaeffer D.F.
        Tumor budding as a standardized parameter in gastrointestinal carcinomas: more than just the colon.
        Mod Pathol. 2018; 31: 862-872
        • Tanaka M.
        • Yamauchi N.
        • Ushiku T.
        • et al.
        Tumor budding in intrahepatic cholangiocarcinoma: a predictor of postsurgery outcomes.
        Am J Surg Pathol. 2019; 43: 1180-1190
        • Boxberg M.
        • Kuhn P.H.
        • Reiser M.
        • et al.
        Tumor budding and cell nest size are highly prognostic in laryngeal and hypopharyngeal squamous cell carcinoma.
        Am J Surg Pathol. 2019; 43: 303-313
        • Kale A.D.
        • Angadi P.V.
        Tumor budding is a potential histopathological marker in the prognosis of oral squamous cell carcinoma: current status and future prospects.
        J Oral Maxillofac Pathol. 2019; 23: 318-323
        • Mäkitie A.A.
        • Almangush A.
        • Rodrigo J.P.
        • Ferlito A.
        • Leivo I.
        Hallmarks of cancer: tumor budding as a sign of invasion and metastasis in head and neck cancer.
        Head Neck. 2019; 41: 3712-3718
        • Zhu Y.
        • Liu H.
        • Xie N.
        • et al.
        Impact of tumor budding in head and neck squamous cell carcinoma: a meta-analysis.
        Head Neck. 2019; 41: 542-550
        • Zare S.Y.
        • Aisagbonhi O.
        • Hasteh F.
        • Fadare O.
        Independent validation of tumor budding activity and cell nest size as determinants of patient outcome in squamous cell carcinoma of the uterine cervix.
        Am J Surg Pathol. 2020; 44: 1151-1160
        • Koelzer V.H.
        • Langer R.
        • Zlobec I.
        • Lugli A.
        Tumor budding in upper gastrointestinal carcinomas.
        Front Oncol. 2014; 4: 216
        • Almangush A.
        • Karhunen M.
        • Hautaniemi S.
        • Salo T.
        • Leivo I.
        Prognostic value of tumour budding in oesophageal cancer: a meta-analysis.
        Histopathology. 2016; 68: 173-182
        • Fuchinoue K.
        • Nemoto T.
        • Shimada H.
        • et al.
        Immunohistochemical analysis of tumor budding as predictor of lymph node metastasis from superficial esophageal squamous cell carcinoma.
        Esophagus. 2020; 17: 168-174
        • Lohneis P.
        • Hieggelke L.
        • Gebauer F.
        • et al.
        Tumor budding assessed according to the criteria of the International Tumor Budding Consensus Conference determines prognosis in resected esophageal adenocarcinoma.
        Virchows Arch. 2021; 478: 393-400
        • Almangush A.
        • Youssef O.
        • Pirinen M.
        • Sundström J.
        • Leivo I.
        • Mäkitie A.A.
        Does evaluation of tumour budding in diagnostic biopsies have a clinical relevance? A systematic review.
        Histopathology. 2019; 74: 536-544
        • Dawson H.
        • Blank A.
        • Zlobec I.
        • Lugli A.
        Potential clinical scenarios of tumour budding in colorectal cancer.
        Acta Gastroenterol Belg. 2019; 82: 515-518
        • Zengın M.
        • Çıfcı A.
        Tumour budding in preoperative biopsy specimens is a useful prognostic index for identifying high-risk patients in early-stage (pN0) colon cancer.
        Turkish J Med Sci. 2020; 50: 375-385
        • Jesinghaus M.
        • Brühl F.
        • Steiger K.
        • et al.
        Cellular dissociation grading based on the parameters tumor budding and cell nest size in pretherapeutic biopsy specimens allows for prognostic patient stratification in esophageal squamous cell carcinoma independent from clinical staging.
        Am J Surg Pathol. 2019; 43: 618-627
        • Jesinghaus M.
        • Boxberg M.
        • Wilhelm D.
        • et al.
        Post-neoadjuvant cellular dissociation grading based on tumour budding and cell nest size is associated with therapy response and survival in oesophageal squamous cell scarcinoma.
        Br J Cancer. 2019; 121: 1050-1057
        • WHO Classification of Tumours Editorial Board
        Digestive System Tumours.
        International Agency for Research on Cancer, Lyon, France2019 (WHO Classification of Tumours Series, 5th Ed.; Vol. 1)
        • Oncology Guideline Program (German Cancer Society, German Cancer Aid, AWMF)
        S3 Guideline diagnosis and therapy of squamous cell carcinoma and adenocarcinoma of the esophagus, long version 2.0, 2018, AWMF registration number: 021/023OL.
        • Jesinghaus M.
        • Strehl J.
        • Boxberg M.
        • et al.
        Introducing a novel highly prognostic grading scheme based on tumour budding and cell nest size for squamous cell carcinoma of the uterine cervix.
        J Pathol Clin Res. 2018; 4: 93-102
        • Elseragy A.
        • Salo T.
        • Coletta R.D.
        • et al.
        A proposal to revise the histopathologic grading system of early oral tongue cancer incorporating tumor budding.
        Am J Surg Pathol. 2019; 43: 703-709
        • Shi H.
        • Ye L.
        • Lu W.
        • Lu B.
        Grading of endocervical adenocarcinoma: a novel prognostic system based on tumor budding and cell cluster size.
        Mod Pathol. 2022; 35: 524-532
        • Brown M.
        • Sillah K.
        • Griffiths E.A.
        • et al.
        Tumour budding and a low host inflammatory response are associated with a poor prognosis in oesophageal and gastro-oesophageal junction cancers.
        Histopathology. 2010; 56: 893-899