Deep learning identifies inflamed fat as a risk factor for lymph node metastasis in early colorectal cancer.

The spread of early-stage (T1 and T2) adenocarcinomas to locoregional lymph nodes is a key event in disease progression of colorectal cancer (CRC). The cellular mechanisms behind this event are not completely understood and existing predictive biomarkers are imperfect. Here, we used an end-to-end deep learning algorithm to identify risk factors for lymph node metastasis (LNM) status in digitized histopathology slides of the primary CRC and its surrounding tissue. In two large population-based cohorts, we show that this system can predict the presence of more than one LNM in pT2 CRC patients with an area under the receiver operating curve (AUROC) of 0.733 (0.67-0.758) and patients with any LNM with an AUROC of 0.711 (0.597-0.797). Similarly, in pT1 CRC patients, the presence of more than one LNM or any LNM was predictable with an AUROC of 0.733 (0.644-0.778) and 0.567 (0.542-0.597), respectively. Based on these findings, we used the deep learning system to guide human pathology experts towards highly predictive regions for LNM in the whole slide images. This hybrid human observer and deep learning approach identified inflamed adipose tissue as the highest predictive feature for LNM presence. Our study is a first proof of concept that artificial intelligence (AI) systems may be able to discover potentially new biological mechanisms in cancer progression. Our deep learning algorithm is publicly available and can be used for biomarker discovery in any disease setting. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Unique expansion of IL-21+ Tfh and Tph cells under control of ICOS identifies Sjögren's syndrome with ectopic germinal centres and MALT lymphoma.

OBJECTIVES: To explore the relevance of T-follicular-helper (Tfh) and pathogenic peripheral-helper T-cells (Tph) in promoting ectopic lymphoid structures (ELS) and B-cell mucosa-associated lymphoid tissue (MALT) lymphomas (MALT-L) in Sjögren's syndrome (SS) patients. METHODS: Salivary gland (SG) biopsies with matched peripheral blood were collected from four centres across the European Union. Transcriptomic (microarray and quantitative PCR) analysis, FACS T-cell immunophenotyping with intracellular cytokine detection, multicolor immune-fluorescence microscopy and in situ hybridisation were performed to characterise lesional and circulating Tfh and Tph-cells. SG-organ cultures were used to investigate functionally the blockade of T-cell costimulatory pathways on key proinflammatory cytokine production. RESULTS: Transcriptomic analysis in SG identified Tfh-signature, interleukin-21 (IL-21) and the inducible T-cell co-stimulator (ICOS) costimulatory pathway as the most upregulated genes in ELS+SS patients, with parotid MALT-L displaying a 400-folds increase in IL-21 mRNA. Peripheral CD4+CXC-motif chemokine receptor 5 (CXCR5)+programmed cell death protein 1 (PD1)+ICOS+ Tfh-like cells were significantly expanded in ELS+SS patients, were the main producers of IL-21, and closely correlated with circulating IgG and reduced complement C4. In the SG, lesional CD4+CD45RO+ICOS+PD1+ cells selectively infiltrated ELS+ tissues and were aberrantly expanded in parotid MALT-L. In ELS+SG and MALT-L parotids, conventional CXCR5+CD4+PD1+ICOS+Foxp3- Tfh-cells and a uniquely expanded population of CXCR5-CD4+PD1hiICOS+Foxp3- Tph-cells displayed frequent IL-21/interferon-γ double-production but poor IL-17 expression. Finally, ICOS blockade in ex vivo SG-organ cultures significantly reduced the production of IL-21 and inflammatory cytokines IL-6, IL-8 and tumour necrosis factor-α (TNF-α). CONCLUSIONS: Overall, these findings highlight Tfh and Tph-cells, IL-21 and the ICOS costimulatory pathway as key pathogenic players in SS immunopathology and exploitable therapeutic targets in SS.