Cytotoxic and immunosuppressive inflammatory cells predict regression and prognosis following neoadjuvant radiochemotherapy of oesophageal adenocarcinoma.

Göbel, Holger H; Büttner-Herold, Maike J; Fuhrich, Nicole; Aigner, Thomas; Grabenbauer, Gerhard G; Distel, Luitpold V R · Radiother Oncol · 2020

retrospective_cohort · Level III

Where this comes from

Abstract

Tumour infiltrating lymphocytes (TIL) and tumour associated macrophages (TAM) play a key role in anticancer immunosurveillance. We studied their influence on response to neoadjuvant radiochemotherapy (RCT) and prognosis in patients with oesophageal adenocarcinoma (OAC). Between 10/2004 and 06/2018, pre-RCT biopsy-specimens were available from 76 patients with locally advanced, non-metastatic OAC scheduled for trimodality therapy. We evaluated intra- and peritumoural expression of FoxP3+-, CD8+-TIL and CD68+-, CD163+-TAM, contemplating cell density, cell ratios and cell-to-cell distances to determine a possible influence on tumour regression grade (TRG) and survival. Median follow-up time for all patients was 18 months (IQR 9-43), and 54 months (25-97) for surviving patients. Data were analysed using risk analysis, logrank test and Cox regression. Poor tumour regression was detected for cN+ (RR 0.77 [95% CI 0.66-0.90], p = 0.001), low intratumoural FoxP3+/CD8+ ratio (RR 0.75 [0.60-0.96], p = 0.020), high peritumoural CD163+/CD68+ ratio (RR 0.77 [0.60-0.99], p = 0.045) and high intratumoural TAM density (RD -0.44 [-0.82 to -0.06], p = 0.023). Apart from poor resection quality and TRG, pretherapeutic high peritumoural CD8+ infiltration (HR 2.36 [1.21-4.61], p = 0.012) and short intratumoural FoxP3+ to CD8+ cell-to-cell distances in middle ranged CD8+ density (HR 2.55 [1.00-6.52], p = 0.050) were significant unfavourable prognostic factors in multivariate analysis. Immunologic parameters, such as CD8+-, FoxP3+-TIL and CD68+-, CD163+-TAM, were identified to be of independent predictive and prognostic value in patients with OAC. Further and independent validation of these biomarkers by a large size dataset may urgently be contemplated.

Medical subject headings