Advancing Artificial Intelligence Integration Into the Pathology Workflow: Exploring Opportunities in Gastrointestinal Tract Biopsies.

This review aims to present a comprehensive overview of the current landscape of artificial intelligence (AI) applications in the analysis of tubular gastrointestinal biopsies. These publications cover a spectrum of conditions, ranging from inflammatory ailments to malignancies. Moving beyond the conventional diagnosis based on hematoxylin and eosin-stained whole-slide images, the review explores additional implications of AI, including its involvement in interpreting immunohistochemical results, molecular subtyping, and the identification of cellular spatial biomarkers. Furthermore, the review examines how AI can contribute to enhancing the quality and control of diagnostic processes, introducing new workflow options, and addressing the limitations and caveats associated with current AI platforms in this context.

Artificial Intelligence Enables Quantitative Assessment of Ulcerative Colitis Histology.

Ulcerative colitis is a chronic inflammatory bowel disease that is characterized by a relapsing and remitting course. Assessment of disease activity critically informs treatment decisions. In addition to endoscopic remission, histologic remission is emerging as a treatment target and a key factor in the evaluation of disease activity and therapeutic efficacy. However, manual pathologist evaluation is semiquantitative and limited in granularity. Machine learning approaches are increasingly being developed to aid pathologists in accurate and reproducible scoring of histology, enabling precise quantitation of clinically relevant features. Here, we report the development and validation of convolutional neural network models that quantify histologic features pertinent to ulcerative colitis disease activity, directly from hematoxylin and eosin-stained whole slide images. Tissue and cell model predictions were used to generate quantitative human-interpretable features to fully characterize the histology samples. Tissue and cell predictions showed comparable agreement to pathologist annotations, and the extracted slide-level human-interpretable features demonstrated strong correlations with disease severity and pathologist-assigned Nancy histological index scores. Moreover, using a random forest classifier based on 13 human-interpretable features derived from the tissue and cell models, we were able to accurately predict Nancy histological index scores, with a weighted kappa (κ = 0.91) and Spearman correlation (⍴ = 0.89, P < .001) when compared with pathologist consensus Nancy histological index scores. We were also able to predict histologic remission, based on the absence of neutrophil extravasation, with a high accuracy of 0.97. This work demonstrates the potential of computer vision to enable a standardized and robust assessment of ulcerative colitis histopathology for translational research and improved evaluation of disease activity and prognosis.

Trastuzumab with trimodality treatment for oesophageal adenocarcinoma with HER2 overexpression (NRG Oncology/RTOG 1010): a multicentre, randomised, phase 3 trial.

BACKGROUND: Trastuzumab is a monoclonal antibody against HER2 (also known as ERBB2). The primary objective of the NRG Oncology/RTOG-1010 trial was to establish whether trastuzumab improves disease-free survival when combined with trimodality treatment (paclitaxel plus carboplatin and radiotherapy, followed by surgery) for patients with untreated HER2-overexpressing oesophageal adenocarcinoma. METHODS: NRG Oncology/RTOG-1010 was an open label, randomised, phase 3 trial for which patients were accrued from 111 NRG-affiliated institutions in the USA. Eligible patients were adults (aged ≥18 years) with newly diagnosed pathologically confirmed oesophageal adenocarcinoma, American Joint Committee on Cancer 7th edition T1N1-2 or T2-3N0-2 stage disease, and a Zubrod performance status of 0-2. Patients were stratified by adenopathy (no vs yes [coeliac absent] vs yes [coeliac present ≤2 cm]) and randomly assigned (1:1) to receive weekly intravenous paclitaxel (50 mg/m2 intravenously over 1 h) and carboplatin (area under the curve 2, intravenously over 30-60 min) for 6 weeks with radiotherapy 50·4 Gy in 28 fractions (chemoradiotherapy) followed by surgery, with or without intravenous trastuzumab (4 mg/kg in week one, 2 mg/kg per week for 5 weeks during chemoradiotherapy, 6 mg/kg once presurgery, and 6 mg/kg every 3 weeks for 13 treatments starting 21-56 days after surgery). The primary endpoint, disease-free survival, was defined as the time from randomisation to death or first of locoregional disease persistence or recurrence, distant metastases, or second primary malignancy. Analyses were done by modified intention to treat. This study is registered with Clinicaltrials.gov, NCT01196390; it is now closed and in follow-up. FINDINGS: 606 patients were entered for HER2 assessment from Dec 30, 2010 to Nov 10, 2015, and 203 eligible patients who were HER2-positive were enrolled and randomly assigned to chemoradiotherapy plus trastuzumab (n=102) or chemoradiotherapy alone (n=101). Median duration of follow-up was 2·8 years (IQR 1·4-5·7). Median disease-free survival was 19·6 months (95% CI 13·5-26·2) with chemoradiotherapy plus trastuzumab compared with 14·2 months (10·5-23·0) for chemoradiotherapy alone (hazard ratio 0·99 [95% CI 0·71-1·39], log-rank p=0·97). Grade 3 treatment-related adverse events occurred in 41 (43%) of 95 patients in the chemoradiotherapy plus trastuzumab group versus 52 (54%) of 96 in the chemoradiotherapy group and grade 4 events occurred in 20 (21%) versus 21 (22%). The most common grade 3 or worse treatment-related adverse events for both groups were haematological (53 [56%] of 95 patients in the chemoradiotherapy plus trastuzumab group vs 55 [57%] of 96 patients in the chemotherapy group) or gastrointestinal disorders (28 [29%] vs 20 [21 %]). 34 (36%) of 95 patients in the chemoradiotherapy plus trastuzumab group and 27 (28%) of 96 patients in the chemoradiotherapy only group had treatment-related serious adverse events. There were eight treatment-related deaths: five (5%) of 95 patients in the chemoradiotherapy plus trastuzumab group (bronchopleural fistula, oesophageal anastomotic leak, lung infection, sudden death, and death not otherwise specified), and three (3%) of 96 in the chemoradiotherapy group (two multiorgan failure and one sepsis). INTERPRETATION: The addition of trastuzumab to neoadjuvant chemoradiotherapy for HER2-overexpressing oesophageal cancer was not effective. Trastuzumab did not lead to increased toxicities, suggesting that future studies combining it with or using other agents targeting HER2 in oesophageal cancer are warranted. FUNDING: National Cancer Institute and Genentech.

A Machine Learning Approach to Liver Histological Evaluation Predicts Clinically Significant Portal Hypertension in NASH Cirrhosis.

BACKGROUND AND AIMS: The hepatic venous pressure gradient (HVPG) is the standard for estimating portal pressure but requires expertise for interpretation. We hypothesized that HVPG could be extrapolated from liver histology using a machine learning (ML) algorithm. APPROACH AND RESULTS: Patients with NASH with compensated cirrhosis from a phase 2b trial were included. HVPG and biopsies from baseline and weeks 48 and 96 were reviewed centrally, and biopsies evaluated with a convolutional neural network (PathAI, Boston, MA). Using trichrome-stained biopsies in the training set (n = 130), an ML model was developed to recognize fibrosis patterns associated with HVPG, and the resultant ML HVPG score was validated in a held-out test set (n = 88). Associations between the ML HVPG score with measured HVPG and liver-related events, and performance of the ML HVPG score for clinically significant portal hypertension (CSPH) (HVPG ≥ 10 mm Hg), were determined. The ML-HVPG score was more strongly correlated with HVPG than hepatic collagen by morphometry (ρ = 0.47 vs. ρ = 0.28; P < 0.001). The ML HVPG score differentiated patients with normal (0-5 mm Hg) and elevated (5.5-9.5 mm Hg) HVPG and CSPH (median: 1.51 vs. 1.93 vs. 2.60; all P < 0.05). The areas under receiver operating characteristic curve (AUROCs) (95% CI) of the ML-HVPG score for CSPH were 0.85 (0.80, 0.90) and 0.76 (0.68, 0.85) in the training and test sets, respectively. Discrimination of the ML-HVPG score for CSPH improved with the addition of a ML parameter for nodularity, Enhanced Liver Fibrosis, platelets, aspartate aminotransferase (AST), and bilirubin (AUROC in test set: 0.85; 95% CI: 0.78, 0.92). Although baseline ML-HVPG score was not prognostic, changes were predictive of clinical events (HR: 2.13; 95% CI: 1.26, 3.59) and associated with hemodynamic response and fibrosis improvement. CONCLUSIONS: An ML model based on trichrome-stained liver biopsy slides can predict CSPH in patients with NASH with cirrhosis.

A Machine Learning Approach Enables Quantitative Measurement of Liver Histology and Disease Monitoring in NASH.

BACKGROUND AND AIMS: Manual histological assessment is currently the accepted standard for diagnosing and monitoring disease progression in NASH, but is limited by variability in interpretation and insensitivity to change. Thus, there is a critical need for improved tools to assess liver pathology in order to risk stratify NASH patients and monitor treatment response. APPROACH AND RESULTS: Here, we describe a machine learning (ML)-based approach to liver histology assessment, which accurately characterizes disease severity and heterogeneity, and sensitively quantifies treatment response in NASH. We use samples from three randomized controlled trials to build and then validate deep convolutional neural networks to measure key histological features in NASH, including steatosis, inflammation, hepatocellular ballooning, and fibrosis. The ML-based predictions showed strong correlations with expert pathologists and were prognostic of progression to cirrhosis and liver-related clinical events. We developed a heterogeneity-sensitive metric of fibrosis response, the Deep Learning Treatment Assessment Liver Fibrosis score, which measured antifibrotic treatment effects that went undetected by manual pathological staging and was concordant with histological disease progression. CONCLUSIONS: Our ML method has shown reproducibility and sensitivity and was prognostic for disease progression, demonstrating the power of ML to advance our understanding of disease heterogeneity in NASH, risk stratify affected patients, and facilitate the development of therapies.

Combination Therapies Including Cilofexor and Firsocostat for Bridging Fibrosis and Cirrhosis Attributable to NASH.

BACKGROUND AND AIMS: Advanced fibrosis attributable to NASH is a leading cause of end-stage liver disease. APPROACH AND RESULTS: In this phase 2b trial, 392 patients with bridging fibrosis or compensated cirrhosis (F3-F4) were randomized to receive placebo, selonsertib 18 mg, cilofexor 30 mg, or firsocostat 20 mg, alone or in two-drug combinations, once-daily for 48 weeks. The primary endpoint was a ≥1-stage improvement in fibrosis without worsening of NASH between baseline and 48 weeks based on central pathologist review. Exploratory endpoints included changes in NAFLD Activity Score (NAS), liver histology assessed using a machine learning (ML) approach, liver biochemistry, and noninvasive markers. The majority had cirrhosis (56%) and NAS ≥5 (83%). The primary endpoint was achieved in 11% of placebo-treated patients versus cilofexor/firsocostat (21%; P = 0.17), cilofexor/selonsertib (19%; P = 0.26), firsocostat/selonsertib (15%; P = 0.62), firsocostat (12%; P = 0.94), and cilofexor (12%; P = 0.96). Changes in hepatic collagen by morphometry were not significant, but cilofexor/firsocostat led to a significant decrease in ML NASH CRN fibrosis score (P = 0.040) and a shift in biopsy area from F3-F4 to ≤F2 fibrosis patterns. Compared to placebo, significantly higher proportions of cilofexor/firsocostat patients had a ≥2-point NAS reduction; reductions in steatosis, lobular inflammation, and ballooning; and significant improvements in alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, bile acids, cytokeratin-18, insulin, estimated glomerular filtration rate, ELF score, and liver stiffness by transient elastography (all P ≤ 0.05). Pruritus occurred in 20%-29% of cilofexor versus 15% of placebo-treated patients. CONCLUSIONS: In patients with bridging fibrosis and cirrhosis, 48 weeks of cilofexor/firsocostat was well tolerated, led to improvements in NASH activity, and may have an antifibrotic effect. This combination offers potential for fibrosis regression with longer-term therapy in patients with advanced fibrosis attributable to NASH.

Somatic mutations in collagens are associated with a distinct tumor environment and overall survival in gastric cancer.

BACKGROUND: Gastric cancer is a heterogeneous disease with poorly understood genetic and microenvironmental factors. Mutations in collagen genes are associated with genetic diseases that compromise tissue integrity, but their role in tumor progression has not been extensively reported. Aberrant collagen expression has been long associated with malignant tumor growth, invasion, chemoresistance, and patient outcomes. We hypothesized that somatic mutations in collagens could functionally alter the tumor extracellular matrix. METHODS: We used publicly available datasets including The Tumor Cancer Genome Atlas (TCGA) to interrogate somatic mutations in collagens in stomach adenocarcinomas. To demonstrate that collagens were significantly mutated above background mutation rates, we used a moderated Kolmogorov-Smirnov test along with combination analysis with a bootstrap approach to define the background accounting for mutation rates. Association between mutations and clinicopathological features was evaluated by Fisher or chi-squared tests. Association with overall survival was assessed by Kaplan-Meier and the Cox-Proportional Hazards Model. Gene Set Enrichment Analysis was used to interrogate pathways. Immunohistochemistry and in situ hybridization tested expression of COL7A1 in stomach tumors. RESULTS: In stomach adenocarcinomas, we identified individual collagen genes and sets of collagen genes harboring somatic mutations at a high frequency compared to background in both microsatellite stable, and microsatellite instable tumors in TCGA. Many of the missense mutations resemble the same types of loss of function mutations in collagenopathies that disrupt tissue formation and destabilize cells providing guidance to interpret the somatic mutations. We identified combinations of somatic mutations in collagens associated with overall survival, with a distinctive tumor microenvironment marked by lower matrisome expression and immune cell signatures. Truncation mutations were strongly associated with improved outcomes suggesting that loss of expression of secreted collagens impact tumor progression and treatment response. Germline collagenopathy variants guided interpretation of impactful somatic mutations on tumors. CONCLUSIONS: These observations highlight that many collagens, expressed in non-physiologically relevant conditions in tumors, harbor impactful somatic mutations in tumors, suggesting new approaches for classification and therapy development in stomach cancer. In sum, these findings demonstrate how classification of tumors by collagen mutations identified strong links between specific genotypes and the tumor environment.

A Multi-institutional Study of Peritoneal Recurrence Following Resection of Low-grade Appendiceal Mucinous Neoplasms.

BACKGROUND: Peritoneal dissemination of low-grade appendiceal mucinous neoplasms (LAMNs), sometimes referred to as pseudomyxoma peritonei, can result in significant morbidity and mortality. Little is known about the natural history of localized (non-disseminated) LAMNs. OBJECTIVE: The goal of this study was to evaluate the risk of peritoneal recurrence in patients with localized LAMNs. METHODS: We performed a multi-institutional retrospective review of patients with pathologically confirmed localized LAMNs. Baseline characteristics, pathology, and follow-up data were collected. The primary endpoint was the rate of peritoneal recurrence. RESULTS: We identified 217 patients with localized LAMNs. Median age was 59 years (11-95) and 131 (60%) patients were female. Surgical management included appendectomy for 124 (57.1%) patients, appendectomy with partial cecectomy for 26 (12.0%) patients, and colectomy for 67 (30.9%) patients. Pathology revealed perforation in 46 patients (37.7% of 122 patients with perforation status mentioned in the report), extra-appendiceal acellular mucin (EAM) in 49 (22.6%) patients, and extra-appendiceal neoplastic cells (EAC) in 13 (6.0%) patients. Median follow-up was 51.1 months (0-271). Seven (3.2%) patients developed a peritoneal recurrence, with a median time to recurrence of 14.4 months (2.5-47.0). Seven (15.2%) patients with histologic evidence of perforation had recurrence, versus no patients (0%) without perforation (p < 0.001); five (10.2%) patients with EAM versus two (1.2%) patients without EAM (p = 0.007), and one (7.7%) patient with EAC versus six (2.9%) patients without EAC (p = 0.355) had recurrence. CONCLUSIONS: This multi-institutional study represents the largest reported series of patients with localized LAMNs. In the absence of perforation or extra-appendiceal mucin or cells, recurrence was extremely rare; however, patients with any of these pathologic findings require careful follow-up.

Smooth muscle tumors of the gastrointestinal tract: an analysis of prognostic features in 407 cases.

Smooth muscle tumors represent the second most common mural mesenchymal neoplasm in the gastrointestinal tract, but established criteria for prognostic assessment of these tumors are lacking. A large cohort of surgically resected intramural gastrointestinal smooth muscle tumors from 31 institutions was analyzed to identify potential prognostic features. Pathologic features were assessed by expert gastrointestinal and/or soft tissue pathologists at each center. Immunohistochemical confirmation was required. A total of 407 cases from the esophagus (n = 97, 24%), stomach (n = 180, 44%), small bowel (n = 74, 18%), and colorectum (n = 56, 14%) were identified. Patients ranged in age from 19 to 92 years (mean 55 years), with a slight female predominance (57%). Mean tumor size was 5.4 cm, with the largest tumor measuring 29 cm. Disease progression following surgery, defined as local recurrence, metastasis, or disease-related death, occurred in 56 patients (14%). Colorectal tumors were most likely to progress, followed by small bowel and gastric tumors. None of the esophageal tumors in this series progressed. Receiver operator characteristic analysis identified optimal cutoffs of 9.8 cm and 3 mitoses/5 mm2 for discriminating between progressive and non-progressive tumors. Histologic features strongly associated with progression by univariate analysis included moderate-to-severe atypia, high cellularity, abnormal differentiation (defined as differentiation not closely resembling that of normal smooth muscle), tumor necrosis, mucosal ulceration, lamina propria involvement, and serosal involvement (P < 0.0001 for all features). Age, sex, and margin status were not significantly associated with progression (P = 0.23, 0.82, and 0.07, respectively). A risk assessment table was created based on tumor site, size, and mitotic count, and Kaplan-Meier plots of progression-free survival for each subgroup revealed progression-based tiers. Based on our findings, it appears that nonesophageal gastrointestinal smooth muscle tumors measuring >10 cm and/or showing ≥3 mitoses/5 mm2 may behave aggressively, and therefore close clinical follow-up is recommended in these cases.

Elastin in the Tumor Microenvironment.

Elastic fibers are found in the extracellular matrix (ECM) of tissues requiring resilience and depend on elasticity. Elastin and its degradation products have multiple roles in the oncologic process. In many malignancies, the remodeled ECM expresses high levels of the elastin protein which may have either positive or negative effects on tumor growth. Elastin cross-linking with other ECM components and the enzymes governing this process all have effects on tumorigenesis. Elastases, and specifically neutrophil elastase, are key drivers of invasion and metastasis and therefore are important targets for inhibition. Elastin degradation leads to the generation of bioactive fragments and elastin-derived peptides that further modulate tumor growth and spread. Interestingly, elastin-like peptides (ELP) and elastin-derived peptides (EDP) may also be utilized as nano-carriers to combat tumor growth. EDPs drive tumor development in a variety of ways, and specifically targeting EDPs and their binding proteins are major objectives for ongoing and future anti-cancer therapies. Research on both the direct anti-cancer activity and the drug delivery capabilities of ELPs is another area likely to result in novel therapeutic agents in the near future.

Cytokeratin 7-negative and GATA binding protein 3-negative breast cancers: Clinicopathological features and prognostic significance.

BACKGROUND: Cytokeratin 7 (CK7) and GATA binding protein 3 (GATA3) are considered as immunohistochemical hallmarks of breast cancers; however, there are breast tumors lacking these markers. Clinicopathological characterization of CK7 negative breast cancer has not been addressed previously and similar studies on GATA3 negative tumors are limited. METHODS: This study included 196 consecutive cases of Nottingham Grade 3 breast cancers with 159 cases of Grade 1 and Grade 2 tumors for comparison. CK7 and GATA3 expression was correlated with patient's age, histological type, pathological grade and stage, hormone receptor status, molecular subtype and overall survival. RESULTS: CK7 negativity was seen in 13% of Grade 3, 9% of Grade 2, and 2% of Grade 1 cases (P = 0.0457). Similarly, 28% of Grade 3, 5% of Grade 2 and 2% of Grade 1 cases were GATA3 negative (P < 0.0001). CK7 negative tumors did not show association with other clinicopathological parameters. GATA3 negative tumors were enriched in the basal-like molecular subgroup and were associated with negative estrogen receptor (ER) and negative progesterone receptor (PR) statuses. Both CK7 and GATA3 expression showed no association with overall survival in patients with Grade 3 tumor. CONCLUSIONS: This is the first study to characterize CK7 negative breast tumors in the context of clinicopathology. Profiling the CK7 negative and GATA3 negative breast cancers helps to understand the biology of these specific tumor subgroups and may aid in their diagnosis.

Stromal ColXα1 expression correlates with tumor-infiltrating lymphocytes and predicts adjuvant therapy outcome in ER-positive/HER2-positive breast cancer.

BACKGROUND: The breast cancer microenvironment contributes to tumor progression and response to chemotherapy. Previously, we reported that increased stromal Type X collagen α1 (ColXα1) and low TILs correlated with poor pathologic response to neoadjuvant therapy in estrogen receptor and HER2-positive (ER+/HER2+) breast cancer. Here, we investigate the relationship of ColXα1 and long-term outcome of ER+/HER2+ breast cancer patients in an adjuvant setting. METHODS: A total of 164 cases with at least 5-year follow-up were included. Immunohistochemistry for ColXα1 was performed on whole tumor sections. Associations between ColXα1expression, clinical pathological features, and outcomes were analyzed. RESULTS: ColXα1 expression was directly proportional to the amount of tumor associated stroma (p = 0.024) and inversely proportional to TILs. Increased ColXα1 was significantly associated with shorter disease free survival and overall survival by univariate analysis. In multivariate analysis, OS was lower in ColXα1 expressing (HR = 2.1; 95% CI = 1.2-3.9) tumors of older patients (> = 58 years) (HR = 5.3; 95% CI = 1.7-17) with higher stage (HR = 2.6; 95% CI = 1.3-5.2). Similarly, DFS was lower in ColXα1 expressing (HR = 1.8; 95% CI = 1.6-5.7) tumors of older patients (HR = 3.2; 95% CI = 1.3-7.8) with higher stage (HR = 2.7; 95% CI = 1.6-5.7) and low TILs. In low PR+ tumors, higher ColXα1 expression was associated with poorer prognosis. CONCLUSION: ColXα1 expression is associated with poor disease free survival and overall survival in ER+/HER2+ breast cancer. This study provides further support for the prognostic utility of ColXα1 as a breast cancer associated stromal factor that predicts response to chemotherapy.

Retrospective Analysis Confirms Tetracycline Quadruple as Best Helicobacter pylori Regimen in the USA.

BACKGROUND: Declining Helicobacter pylori (H. pylori) eradication rates have prompted a switch in first-line therapy from standard triple (PPI, clarithromycin, and amoxicillin) to bismuth-based quadruple therapy. A caveat of the ACG 2017 H. pylori treatment guidelines was a paucity of recent US eradication data. AIM: To determine Rhode Island H. pylori eradication data, in the largest US study from the last two decades. METHODS: Electronic records were queried for patients with H. pylori infection diagnosed by pathology, urea breath test, or stool antigen from 2015 to 2017. Demographics, diagnostic test, treatment regimen, and test of cure were extracted. Eradication rates were calculated, and treatment regimens were compared. RESULTS: A total of 1710 patients were identified (64% female): 825 (46%) diagnosed by breath test, 755 (42%) by biopsy, and 191 (12%) by stool antigen. Full data were obtained on 1101 patients. Seven regimens were used: quadruple (64%), triple (25%), doxycycline quadruple (5%), and miscellaneous (6%). Quadruple was superior to triple: (85% vs. 75%, P = 0.002), quadruple 14 days versus triple 14 days (87% vs. 79%, P = 0.0052), quadruple 10 days versus triple 10 days (77% vs. 67%, P = 0.33). Increased therapy length improved eradication (quadruple 14 days  vs. 10 days, 87% vs. 77%, P = 0.002; triple 14 days  versus 10 days 79% vs. 67%, P = 0.13). Finally, substituting doxycycline for tetracycline yielded lower eradication (85% vs. 67%, P = 0.006). CONCLUSION: Quadruple therapy is superior to triple therapy within the Rhode Island population. Fourteen-day therapy achieves superior eradication compared to 10-day therapy, and doxycycline is inferior to tetracycline for quadruple therapy. Our findings support adherence to ACG and international guidelines advising 14-day quadruple therapy.

Role of immune microenvironment in gastrointestinal stromal tumours.

AIMS: The immune microenvironment is a prognostic factor for various malignancies. The significance of key players of this immune microenvironment, including tumour-infiltrating lymphocytes (TILs) and expression of programmed death-ligand 1 (PD-L1), indoleamine 2,3-dioxygenase (IDO) and tryptophanyl-tRNA synthetase (WARS) in gastrointestinal stromal tumours (GISTs) is largely unknown. METHODS AND RESULTS: Tissue microarrays were constructed from pathology files, 1996-2016. Immunohistochemistry for PD-L1, IDO and WARS was correlated with tumour size, mitoses and outcomes. TILs expressing CD3, CD4, CD8, FoxP3 and GBP5 were counted. A total of 129 GISTs were analysed. Mean patient age was 62.5 years; 52.0% were male. Tumour location included 89 stomach (69.0%), 33 small bowel (25.6%) and seven other (5.4%). Mean tumour size was 5.6 cm; mean mitoses were 7.2 per 50 high-power field. Nineteen patients (15.0%) developed disease progression, to abdominal wall (n = 8), liver (n = 6) and elsewhere (n = 5). Median progression-free survival was 56.6 months; five patients died of disease. PD-L1 was positive in 88 of 127 tumour samples (69.0%), 114 of 127 tumours were IDO-positive (89.8%) and 60 of 127 were positive for WARS (47.2%). PD-L1 was associated with increased size (P = 0.01), necrosis (P = 0.018) and mitoses (P = 0.006). Disease progression was not associated with PD-L1 (P = 0.44), IDO (P = 0.14) or WARS (P = 0.36) expression. PD-L1-positive GISTs with CD8+ or CD3+ TILs were significantly smaller than tumours with CD8+ or CD3+ TILs. CONCLUSIONS: PD-L1 expression was associated with increased size and mitoses. High CD8+ or CD3+ TIL counts were associated with decreased PD-L1/IDO+ GIST size. PD-L1 and IDO could be significant in GIST tumour biology, which invites consideration of immunotherapy as a potential treatment option.

Fatty acid-binding protein 1 is preferentially lost in microsatellite instable colorectal carcinomas and is immune modulated via the interferon γ pathway.

Fatty acid-binding protein 1 (FABP1) is an intracellular protein responsible for the transportation of long chain fatty acids. Aside from its functions in lipid metabolism and cellular differentiation, FABP1 also plays a role in inflammation through its interaction with peroxisome proliferator-activated receptors (PPARs). Previously, we compared expression of colonic epithelium genes in a subset of microsatellite instable (MSI) colorectal carcinomas (medullary carcinomas) to normal colonic mucosa and found that FABP1 expression was markedly decreased in the tumors. Further analysis of RNA expression in the colorectal subtypes and The Cancer Genome Atlas data set found that FABP1 expression is decreased in the CMS1 subset of colorectal carcinomas, which is characterized by microsatellite instability. As MSI colorectal carcinomas are known for their robust immune response, we then aimed to link FABP1 to the immune microenvironment of MSI carcinomas. To confirm the gene expression results, we performed immunohistochemical analysis of a cohort of colorectal carcinomas. FABP1 was preferentially lost in MSI carcinomas (123/133, 93%) compared with microsatellite stable carcinomas (240/562, 43%, P<0.0001). In addition, higher numbers of tumor-infiltrating lymphocytes were present in tumors with loss of FABP1 (P<0.0001). Decreased expression of the fatty acid storage and glucose regulator, PPARγ, was associated with the loss of FABP1 (P<0.0001). Colorectal cancer cell lines treated with interferon γ exhibited decreased expression of FABP1. FABP1 expression was partially recovered with the treatment of the cell lines with rosiglitazone, a PPARγ agonist. This study demonstrated that the loss of FABP1 expression is associated with MSI carcinomas and that interferon γ stimulation plays a role in this process via its interaction with PPARγ.

Medullary carcinoma of the colon: a distinct morphology reveals a distinctive immunoregulatory microenvironment.

Medullary carcinoma of the colon is a unique histologic subtype of microsatellite unstable colorectal carcinoma but little is known regarding its tumor-immunoregulatory microenvironment. The aims of this study were to characterize the immune environment of medullary carcinoma and compare it with other microsatellite unstable and microsatellite stable colorectal carcinomas. An initial gene expression microarray analysis of six cases of medullary carcinoma was used to detect potentially differentially expressed genes. We extended this analysis utilizing genomic data from the Cancer Genome Atlas to compare eight cases of medullary carcinoma with other microsatellite unstable and stable carcinomas. Finally, we evaluated expression of key immune pathway proteins and lymphocyte subsets via immunohistochemistry of a large group of medullary carcinomas (n=105) and compared these findings with three other groups: poorly differentiated, microsatellite unstable well-differentiated and microsatellite stable well-differentiated carcinomas. Microarray and the Cancer Genome Atlas data analysis identified significant upregulation of several immunoregulatory genes induced by IFNγ including IDO-1, WARS (tRNA(trp)), GBP1, GBP4, GBP5, PDCD1 (PD-1), and CD274 (PD-L1) in medullary carcinoma compared with other microsatellite unstable and microsatellite stable tumors. By immunohistochemistry, IDO-1 was expressed in 64% of medullary carcinomas compared with 19% (9/47) of poorly differentiated carcinomas, 14% (3/22) of microsatellite unstable, and 7% (2/30) of the microsatellite stable well-differentiated carcinomas (P<0.0001). tRNA(trp) was overexpressed in 81% (84/104) of medullary carcinomas, 19% (9/47) of poorly differentiated, 32% (7/22) of microsatellite unstable, and 3% (1/30) of microsatellite stable well-differentiated carcinomas (P<0.0001). Medullary carcinoma had higher mean CD8+ and PD-L1+ tumor-infiltrating lymphocytes compared with all other groups (P<0.0001). This study demonstrates overexpression of several immunoregulatory genes in microsatellite unstable colorectal carcinomas and that expression of these genes and proteins is more prevalent in the medullary carcinoma subtype, which may be of use both diagnostically and therapeutically.

Identification of stromal ColXα1 and tumor-infiltrating lymphocytes as putative predictive markers of neoadjuvant therapy in estrogen receptor-positive/HER2-positive breast cancer.

BACKGROUND: The influence of the tumor microenvironment and tumor-stromal interactions on the heterogeneity of response within breast cancer subtypes have just begun to be explored. This study focuses on patients with estrogen receptor-positive/human epidermal growth factor receptor 2-positive (ER+/HER2+) breast cancer receiving neoadjuvant chemotherapy and HER2-targeted therapy (NAC+H), and was designed to identify novel predictive biomarkers by combining gene expression analysis and immunohistochemistry with pathologic response. METHODS: We performed gene expression profiling on pre-NAC+H tumor samples from responding (no or minimal residual disease at surgery) and non-responding patients. Gene set enrichment analysis identified potentially relevant pathways, and immunohistochemical staining of pre-treatment biopsies was used to measure protein levels of those pathways, which were correlated with pathologic response in both univariate and multivariate analysis. RESULTS: Increased expression of genes encoding for stromal collagens, including Col10A1, and reduced expression of immune-associated genes, reflecting lower levels of total tumor-infiltrating lymphocytes (TILs), were strongly associated with poor pathologic response. Lower TILs in tumor biopsies correlated with reduced likelihood of achieving an optimal pathologic response, but increased expression of the Col10A1 gene product, colXα1, had greater predictive value than stromal abundance for poor response (OR = 18.9, p = 0.003), and the combination of increased colXα1 expression and low TILs was significantly associated with poor response in multivariate analysis. ROC analysis suggests strong specificity and sensitivity for this combination in predicting treatment response. CONCLUSIONS: Increased expression of stromal colXα1 and low TILs correlate with poor pathologic response in ER+/HER2+ breast tumors. Further studies are needed to confirm their predictive value and impact on long-term outcomes, and to determine whether this collagen exerts a protective effect on the cancer cells or simply reflects other factors within the tumor microenvironment.

Glycogen synthase kinase 3 beta inhibits microRNA-183-96-182 cluster via the ß-Catenin/TCF/LEF-1 pathway in gastric cancer cells.

Glycogen synthase kinase 3 beta (GSK3ß) is a critical protein kinase that phosphorylates numerous proteins in cells and thereby impacts multiple pathways including the ß-Catenin/TCF/LEF-1 pathway. MicroRNAs (miRs) are a class of noncoding small RNAs of ∼22 nucleotides in length. Both GSK3ß and miR play myriad roles in cell functions including stem cell development, apoptosis, embryogenesis and tumorigenesis. Here we show that GSK3ß inhibits the expression of miR-96, miR-182 and miR-183 through the ß-Catenin/TCF/LEF-1 pathway. Knockout of GSK3ß in mouse embryonic fibroblast cells increases expression of miR-96, miR-182 and miR-183, coinciding with increases in the protein level and nuclear translocation of ß-Catenin. In addition, overexpression of ß-Catenin enhances the expression of miR-96, miR-182 and miR-183 in human gastric cancer AGS cells. GSK3ß protein levels are decreased in human gastric cancer tissue compared with surrounding normal gastric tissue, coinciding with increases of ß-Catenin protein, miR-96, miR-182, miR-183 and primary miR-183-96-182 cluster (pri-miR-183). Furthermore, suppression of miR-183-96-182 cluster with miRCURY LNA miR inhibitors decreases the proliferation and migration of AGS cells. Knockdown of GSK3ß with siRNA increases the proliferation of AGS cells. Mechanistically, we show that ß-Catenin/TCF/LEF-1 binds to the promoter of miR-183-96-182 cluster gene and thereby activates the transcription of the cluster. In summary, our findings identify a novel role for GSK3ß in the regulation of miR-183-96-182 biogenesis through ß-Catenin/TCF/LEF-1 pathway in gastric cancer cells.

Expression of cadherin 17 in well-differentiated neuroendocrine tumours.

AIMS: Cadherin 17 (CDH17) is expressed primarily in normal intestinal epithelium and digestive tract tumours, and has limited expression in other neoplasms. The aims of this study were to examine CDH17 expression in well-differentiated neuroendocrine tumours (WDNETs) from various primary sites, representing the foregut, midgut, and hindgut, and tumours metastasizing to the liver, and to correlate the differences between the expression of CDH17, CDX2, and thyroid transcription factor 1 (TTF1). METHODS AND RESULTS: We investigated CDH17 immunohistochemical expression in 150 primary WDNETs from eight anatomical sites, including 68 from the foregut, 70 from the midgut, and 12 from the hindgut, and 15 metastases. CDH17 immunoreactivity increased significantly from foregut to hindgut WDNETs (P < 0.0001). Pancreatic WDNETs expressed CDH17 at a significantly higher frequency than other foregut tumours. Within the midgut, appendiceal and small-intestinal WDNETs were more frequently positive for CDH17 than for CDX2. All hindgut WDNETs expressed CDH17, in contrast to CDX2 (positive in one rectal case). CDH17 expression in liver metastases was similar to that of the primary tumours. CONCLUSIONS: This study is the first to comprehensively examine CDH17 expression in WDNETs from different sites. CDH17 is a sensitive marker for midgut WDNETs, and the CDH17+/CDX2-/TTF1- phenotype was found to be sensitive (92%) and specific (91%) for hindgut WDNETs.