High intratumoural but not peritumoural inflammatory host response is associated with better prognosis in primary resected oesophageal adenocarcinomas.

The host inflammatory response plays an important role in many solid malignancies. Studies on oesophageal adenocarcinomas (EACs) point towards a beneficial role of pronounced immunoreaction, however, congruent results have yet to be obtained. We analysed 111 primary resected EAC using a tissue microarray containing three cores of the tumour centre and the periphery per case. Overall inflammation was assessed by histomorphology. Tumour infiltrating lymphocytes (TILs) were characterised by immunohistochemistry for CD3, CD8 and FoxP3, and evaluated by image analysis (Aperio ImageScope). High levels of inflammation in the tumour centre, but not the periphery were associated with better patient survival (p = 0.001), similar to high counts of intratumoural FoxP3+, CD3+, CD8+ TILs (p = 0.001; p = 0.027; p = 0.038) and a combination of CD3+/CD8+/FoxP3+ TILs, the latter displaying three different prognostic groups (triple high/mixed/triple low; p=0.003). Intratumoural inflammation [hazard ratio (HR) = 0.432; p = 0.030], FoxP3+ TIL counts (HR = 0.411; p = 0.033) and the combination CD3+/CD8+/FoxP3+ TILs (HR = 0.173; p = 0.006) were also independent prognostic parameters. In summary, both high grade total inflammation and high TIL counts in the tumour centre, but not the tumour periphery, show a beneficial prognostic impact on EAC. This may be a target for novel therapeutic options but also serves as prognostic indicator in these tumours.

Impact of peritumoral and intratumoral budding in esophageal adenocarcinomas.

Tumor budding has prognostic significance in many carcinomas and is defined as the presence of detached isolated single cells or small cell clusters up to 5 cells at the invasion front (peritumoral budding [PTB]) or within the tumor (intratumoral budding [ITB]). For esophageal adenocarcinomas (EACs), there are currently only few data about the impact of this morphological feature. We investigated tumor budding in a large collective of 200 primarily resected EACs. Pancytokeratin staining was demonstrated to be superior to hematoxylin and eosin staining for the detection of buds with substantial to excellent interobserver agreement and used for subsequent analysis. PTB and ITB were scored across 10 high-power fields (HPFs). The median count of tumor buds was 130/10 HPFs for PTB (range, 2-593) and 80/10 HPFs for ITB (range, 1-656). PTB and ITB correlated significantly with each other (r = 0.9; P < .001). High PTB and ITB rates were seen in more advanced tumor categories (P < .001 each); tumors with lymph node metastases (P < .001/P = .002); and lymphatic, vascular, and perineural invasion and higher tumor grading (P < .001 each). Survival analysis showed an association with worse survival for high-grade ITB (P = .029) but not PTB (P = .385). However, in multivariate analysis, lymph node and resection status, but not ITB, were independent prognostic parameters. In conclusion, PTB and ITB can be observed in EAC to various degrees. High-grade budding is associated with aggressive tumor phenotype. Assessment of tumor budding, especially ITB, may provide additional prognostic information about tumor behavior and may be useful in specific cases for risk stratification of EAC patients.