AQP5 enriches for stem cells and cancer origins in the distal stomach.

LGR5 marks resident adult epithelial stem cells at the gland base in the mouse pyloric stomach1, but the identity of the equivalent human stem cell population remains unknown owing to a lack of surface markers that facilitate its prospective isolation and validation. In mouse models of intestinal cancer, LGR5+ intestinal stem cells are major sources of cancer following hyperactivation of the WNT pathway2. However, the contribution of pyloric LGR5+ stem cells to gastric cancer following dysregulation of the WNT pathway-a frequent event in gastric cancer in humans3-is unknown. Here we use comparative profiling of LGR5+ stem cell populations along the mouse gastrointestinal tract to identify, and then functionally validate, the membrane protein AQP5 as a marker that enriches for mouse and human adult pyloric stem cells. We show that stem cells within the AQP5+ compartment are a source of WNT-driven, invasive gastric cancer in vivo, using newly generated Aqp5-creERT2 mouse models. Additionally, tumour-resident AQP5+ cells can selectively initiate organoid growth in vitro, which indicates that this population contains potential cancer stem cells. In humans, AQP5 is frequently expressed in primary intestinal and diffuse subtypes of gastric cancer (and in metastases of these subtypes), and often displays altered cellular localization compared with healthy tissue. These newly identified markers and mouse models will be an invaluable resource for deciphering the early formation of gastric cancer, and for isolating and characterizing human-stomach stem cells as a prerequisite for harnessing the regenerative-medicine potential of these cells in the clinic.

Stroma AReactive Invasion Front Areas (SARIFA) improves prognostic risk stratification of perioperative chemotherapy treated oesophagogastric cancer patients from the MAGIC and the ST03 trial.

BACKGROUND: Tumour-associated fat cells without desmoplastic stroma reaction at the invasion front (Stroma AReactive Invasion Front Areas (SARIFA)) is a prognostic biomarker in gastric and colon cancer. The clinical utility of the SARIFA status in oesophagogastric cancer patients treated with perioperative chemotherapy is currently unknown. METHODS: The SARIFA status was determined in tissue sections from patients recruited into the MAGIC (n = 292) or ST03 (n = 693) trials treated with surgery alone (S, MAGIC) or perioperative chemotherapy (MAGIC, ST03). The relationship between SARIFA status, clinicopathological factors, overall survival (OS) and treatment was analysed. RESULTS: The SARIFA status was positive in 42% MAGIC trial S patients, 28% MAGIC and 48% ST03 patients after pre-operative chemotherapy. SARIFA status was related to OS in MAGIC trial S patients and was an independent prognostic biomarker in ST03 trial patients (HR 1.974, 95% CI 1.555-2.507, p < 0.001). ST03 patients with lymph node metastasis (ypN + ) and SARIFA-positive tumours had poorer OS than patients with ypN+ and SARIFA-negative tumours (plogrank < 0.001). CONCLUSIONS: The SARIFA status has clinical utility as prognostic biomarker in oesophagogastric cancer patients irrespective of treatment modality. Whilst underlying biological mechanisms warrant further investigation, the SARIFA status might be used to identify new drug targets, potentially enabling repurposing of existing drugs targeting lipid metabolism.

Automated curation of large-scale cancer histopathology image datasets using deep learning.

BACKGROUND: Artificial intelligence (AI) has numerous applications in pathology, supporting diagnosis and prognostication in cancer. However, most AI models are trained on highly selected data, typically one tissue slide per patient. In reality, especially for large surgical resection specimens, dozens of slides can be available for each patient. Manually sorting and labelling whole-slide images (WSIs) is a very time-consuming process, hindering the direct application of AI on the collected tissue samples from large cohorts. In this study we addressed this issue by developing a deep-learning (DL)-based method for automatic curation of large pathology datasets with several slides per patient. METHODS: We collected multiple large multicentric datasets of colorectal cancer histopathological slides from the United Kingdom (FOXTROT, N = 21,384 slides; CR07, N = 7985 slides) and Germany (DACHS, N = 3606 slides). These datasets contained multiple types of tissue slides, including bowel resection specimens, endoscopic biopsies, lymph node resections, immunohistochemistry-stained slides, and tissue microarrays. We developed, trained, and tested a deep convolutional neural network model to predict the type of slide from the slide overview (thumbnail) image. The primary statistical endpoint was the macro-averaged area under the receiver operating curve (AUROCs) for detection of the type of slide. RESULTS: In the primary dataset (FOXTROT), with an AUROC of 0.995 [95% confidence interval [CI]: 0.994-0.996] the algorithm achieved a high classification performance and was able to accurately predict the type of slide from the thumbnail image alone. In the two external test cohorts (CR07, DACHS) AUROCs of 0.982 [95% CI: 0.979-0.985] and 0.875 [95% CI: 0.864-0.887] were observed, which indicates the generalizability of the trained model on unseen datasets. With a confidence threshold of 0.95, the model reached an accuracy of 94.6% (7331 classified cases) in CR07 and 85.1% (2752 classified cases) for the DACHS cohort. CONCLUSION: Our findings show that using the low-resolution thumbnail image is sufficient to accurately classify the type of slide in digital pathology. This can support researchers to make the vast resource of existing pathology archives accessible to modern AI models with only minimal manual annotations.

Pathological regression of primary tumour and metastatic lymph nodes following chemotherapy in resectable OG cancer: pooled analysis of two trials.

BACKGROUND: No definitive largescale data exist evaluating the role of pathologically defined regression changes within the primary tumour and lymph nodes (LN) of resected oesophagogastric (OG) adenocarcinoma following neoadjuvant chemotherapy and the impact on survival. METHODS: Data and samples from two large prospective randomised trials (UK MRC OE05 and ST03) were pooled. Stained slides were available for central pathology review from 1619 patients. Mandard tumour regression grade (TRG) and regression of tumour within LNs (LNR: scored as present/absent) were assessed and correlated with overall survival (OS) using a Cox regression model. An exploratory analysis to define subgroups with distinct prognoses was conducted using a classification and regression tree (CART) analysis. RESULTS: Neither trial demonstrated a relationship between TRG score and the presence or absence of LNR. In univariable analysis, lower TRG, lower ypN stage, lower ypT stage, presence of LNR, presence of well/moderate tumour differentiation, and absence of tumour at resection margin were all associated with better OS. However, the multivariable analysis demonstrated that only ypN, ypT, grade of differentiation and resection margin (R0) were independent indicators of prognosis. Exploratory CART analysis identified six subgroups with 3-year OS ranging from 83% to 22%; with ypN stage being the most important single prognostic variable. CONCLUSIONS: Pathological LN stage within the resection specimen was the single most important determiner of survival. Our results suggest that the assessment of regression changes within the primary tumour or LNs may not be necessary to define the prognosis further.

Tumour infiltrating lymphocytes and survival after adjuvant chemotherapy in patients with gastric cancer: post-hoc analysis of the CLASSIC trial.

BACKGROUND: Only a subset of gastric cancer (GC) patients with stage II-III benefits from chemotherapy after surgery. Tumour infiltrating lymphocytes per area (TIL density) has been suggested as a potential predictive biomarker of chemotherapy benefit. METHODS: We quantified TIL density in digital images of haematoxylin-eosin (HE) stained tissue using deep learning in 307 GC patients of the Yonsei Cancer Center (YCC) (193 surgery+adjuvant chemotherapy [S + C], 114 surgery alone [S]) and 629 CLASSIC trial GC patients (325 S + C and 304 S). The relationship between TIL density, disease-free survival (DFS) and clinicopathological variables was analysed. RESULTS: YCC S patients and CLASSIC S patients with high TIL density had longer DFS than S patients with low TIL density (P = 0.007 and P = 0.013, respectively). Furthermore, CLASSIC patients with low TIL density had longer DFS if treated with S + C compared to S (P = 0.003). No significant relationship of TIL density with other clinicopathological variables was found. CONCLUSION: This is the first study to suggest TIL density automatically quantified in routine HE stained tissue sections as a novel, clinically useful biomarker to identify stage II-III GC patients deriving benefit from adjuvant chemotherapy. Validation of our results in a prospective study is warranted.

Prognostic Significance of Negative Lymph Node Long Axis in Esophageal Cancer: Results From the Randomized Controlled UK MRC OE02 Trial.

OBJECTIVE: To analyze the relationship between negative lymph node (LNneg) size as a possible surrogate marker of the host antitumor immune response and overall survival (OS) in esophageal cancer (EC) patients. BACKGROUND: Lymph node (LN) status is a well-established prognostic factor in EC patients. An increased number of LNnegs is related to better survival in EC. Follicular hyperplasia in LNneg is associated with better survival in cancer-bearing mice and might explain increased LN size. METHODS: The long axis of 304 LNnegs was measured in hematoxylin-eosin stained sections from resection specimens of 367 OE02 trial patients (188 treated with surgery alone (S), 179 with neoadjuvant chemotherapy plus surgery (C+S)) as a surrogate of LN size. The relationship between LNneg size, LNneg microarchitecture, clinicopathological variables, and OS was analyzed. RESULTS: Large LNneg size was related to lower pN category ( P = 0.01) and lower frequency of lymphatic invasion ( P = 0.02) in S patients only. Irrespective of treatment, (y)pN0 patients with large LNneg had the best OS. (y)pN1 patients had the poorest OS irrespective of LNneg size ( P < 0.001). Large LNneg contained less lymphocytes ( P = 0.02) and had a higher germinal centers/lymphocyte ratio ( P = 0.05). CONCLUSIONS: This is the first study to investigate LNneg size in EC patients randomized to neoadjuvant chemotherapy followed by surgery or surgery alone. Our pilot study suggests that LNneg size is a surrogate marker of the host antitumor immune response and a potentially clinically useful new prognostic biomarker for (y)pN0 EC patients. Future studies need to confirm our results and explore underlying biological mechanisms.

The clinical importance of the host anti-tumour reaction patterns in regional tumour draining lymph nodes in patients with locally advanced resectable gastric cancer: a systematic review and meta-analysis.

BACKGROUND: The status of regional tumour draining lymph nodes (LN) is crucial for prognostic evaluation in gastric cancer (GaC) patients. Changes in lymph node microarchitecture, such as follicular hyperplasia (FH), sinus histiocytosis (SH), or paracortical hyperplasia (PH), may be triggered by the anti-tumour immune response. However, the prognostic value of these changes in GaC patients is unclear. METHODS: A systematic search in multiple databases was conducted to identify studies on the prognostic value of microarchitecture changes in regional tumour-negative and tumour-positive LNs measured on histopathological slides. Since the number of GaC publications was very limited, the search was subsequently expanded to include junctional and oesophageal cancer (OeC). RESULTS: A total of 28 articles (17 gastric cancer, 11 oesophageal cancer) met the inclusion criteria, analyzing 26,503 lymph nodes from 3711 GaC and 1912 OeC patients. The studies described eight different types of lymph node microarchitecture changes, categorized into three patterns: hyperplasia (SH, FH, PH), cell-specific infiltration (dendritic cells, T cells, neutrophils, macrophages), and differential gene expression. Meta-analysis of five GaC studies showed a positive association between SH in tumour-negative lymph nodes and better 5-year overall survival. Pooled risk ratios for all LNs showed increased 5-year overall survival for the presence of SH and PH. CONCLUSIONS: This systematic review suggests that sinus histiocytosis and paracortical hyperplasia in regional tumour-negative lymph nodes may provide additional prognostic information for gastric and oesophageal cancer patients. Further studies are needed to better understand the lymph node reaction patterns and explore their impact of chemotherapy treatment and immunotherapy efficacy.

Deep learning-based subtyping of gastric cancer histology predicts clinical outcome: a multi-institutional retrospective study.

INTRODUCTION: The Laurén classification is widely used for Gastric Cancer (GC) histology subtyping. However, this classification is prone to interobserver variability and its prognostic value remains controversial. Deep Learning (DL)-based assessment of hematoxylin and eosin (H&E) stained slides is a potentially useful tool to provide an additional layer of clinically relevant information, but has not been systematically assessed in GC. OBJECTIVE: We aimed to train, test and externally validate a deep learning-based classifier for GC histology subtyping using routine H&E stained tissue sections from gastric adenocarcinomas and to assess its potential prognostic utility. METHODS: We trained a binary classifier on intestinal and diffuse type GC whole slide images for a subset of the TCGA cohort (N = 166) using attention-based multiple instance learning. The ground truth of 166 GC was obtained by two expert pathologists. We deployed the model on two external GC patient cohorts, one from Europe (N = 322) and one from Japan (N = 243). We assessed classification performance using the Area Under the Receiver Operating Characteristic Curve (AUROC) and prognostic value (overall, cancer specific and disease free survival) of the DL-based classifier with uni- and multivariate Cox proportional hazard models and Kaplan-Meier curves with log-rank test statistics. RESULTS: Internal validation using the TCGA GC cohort using five-fold cross-validation achieved a mean AUROC of 0.93 ± 0.07. External validation showed that the DL-based classifier can better stratify GC patients' 5-year survival compared to pathologist-based Laurén classification for all survival endpoints, despite frequently divergent model-pathologist classifications. Univariate overall survival Hazard Ratios (HRs) of pathologist-based Laurén classification (diffuse type versus intestinal type) were 1.14 (95% Confidence Interval (CI) 0.66-1.44, p-value = 0.51) and 1.23 (95% CI 0.96-1.43, p-value = 0.09) in the Japanese and European cohorts, respectively. DL-based histology classification resulted in HR of 1.46 (95% CI 1.18-1.65, p-value < 0.005) and 1.41 (95% CI 1.20-1.57, p-value < 0.005), in the Japanese and European cohorts, respectively. In diffuse type GC (as defined by the pathologist), classifying patients using the DL diffuse and intestinal classifications provided a superior survival stratification, and demonstrated statistically significant survival stratification when combined with pathologist classification for both the Asian (overall survival log-rank test p-value < 0.005, HR 1.43 (95% CI 1.05-1.66, p-value = 0.03) and European cohorts (overall survival log-rank test p-value < 0.005, HR 1.56 (95% CI 1.16-1.76, p-value < 0.005)). CONCLUSION: Our study shows that gastric adenocarcinoma subtyping using pathologist's Laurén classification as ground truth can be performed using current state of the art DL techniques. Patient survival stratification seems to be better by DL-based histology typing compared with expert pathologist histology typing. DL-based GC histology typing has potential as an aid in subtyping. Further investigations are warranted to fully understand the underlying biological mechanisms for the improved survival stratification despite apparent imperfect classification by the DL algorithm.

Direct prediction of genetic aberrations from pathology images in gastric cancer with swarm learning.

BACKGROUND: Computational pathology uses deep learning (DL) to extract biomarkers from routine pathology slides. Large multicentric datasets improve performance, but such datasets are scarce for gastric cancer. This limitation could be overcome by Swarm Learning (SL). METHODS: Here, we report the results of a multicentric retrospective study of SL for prediction of molecular biomarkers in gastric cancer. We collected tissue samples with known microsatellite instability (MSI) and Epstein-Barr Virus (EBV) status from four patient cohorts from Switzerland, Germany, the UK and the USA, storing each dataset on a physically separate computer. RESULTS: On an external validation cohort, the SL-based classifier reached an area under the receiver operating curve (AUROC) of 0.8092 (± 0.0132) for MSI prediction and 0.8372 (± 0.0179) for EBV prediction. The centralized model, which was trained on all datasets on a single computer, reached a similar performance. CONCLUSIONS: Our findings demonstrate the feasibility of SL-based molecular biomarkers in gastric cancer. In the future, SL could be used for collaborative training and, thus, improve the performance of these biomarkers. This may ultimately result in clinical-grade performance and generalizability.

Weakly supervised annotation-free cancer detection and prediction of genotype in routine histopathology.

Deep learning is a powerful tool in computational pathology: it can be used for tumor detection and for predicting genetic alterations based on histopathology images alone. Conventionally, tumor detection and prediction of genetic alterations are two separate workflows. Newer methods have combined them, but require complex, manually engineered computational pipelines, restricting reproducibility and robustness. To address these issues, we present a new method for simultaneous tumor detection and prediction of genetic alterations: The Slide-Level Assessment Model (SLAM) uses a single off-the-shelf neural network to predict molecular alterations directly from routine pathology slides without any manual annotations, improving upon previous methods by automatically excluding normal and non-informative tissue regions. SLAM requires only standard programming libraries and is conceptually simpler than previous approaches. We have extensively validated SLAM for clinically relevant tasks using two large multicentric cohorts of colorectal cancer patients, Darmkrebs: Chancen der Verhütung durch Screening (DACHS) from Germany and Yorkshire Cancer Research Bowel Cancer Improvement Programme (YCR-BCIP) from the UK. We show that SLAM yields reliable slide-level classification of tumor presence with an area under the receiver operating curve (AUROC) of 0.980 (confidence interval 0.975, 0.984; n = 2,297 tumor and n = 1,281 normal slides). In addition, SLAM can detect microsatellite instability (MSI)/mismatch repair deficiency (dMMR) or microsatellite stability/mismatch repair proficiency with an AUROC of 0.909 (0.888, 0.929; n = 2,039 patients) and BRAF mutational status with an AUROC of 0.821 (0.786, 0.852; n = 2,075 patients). The improvement with respect to previous methods was validated in a large external testing cohort in which MSI/dMMR status was detected with an AUROC of 0.900 (0.864, 0.931; n = 805 patients). In addition, SLAM provides human-interpretable visualization maps, enabling the analysis of multiplexed network predictions by human experts. In summary, SLAM is a new simple and powerful method for computational pathology that could be applied to multiple disease contexts. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

Machine-learning model derived gene signature predictive of paclitaxel survival benefit in gastric cancer: results from the randomised phase III SAMIT trial.

OBJECTIVE: To date, there are no predictive biomarkers to guide selection of patients with gastric cancer (GC) who benefit from paclitaxel. Stomach cancer Adjuvant Multi-Institutional group Trial (SAMIT) was a 2×2 factorial randomised phase III study in which patients with GC were randomised to Pac-S-1 (paclitaxel +S-1), Pac-UFT (paclitaxel +UFT), S-1 alone or UFT alone after curative surgery. DESIGN: The primary objective of this study was to identify a gene signature that predicts survival benefit from paclitaxel chemotherapy in GC patients. SAMIT GC samples were profiled using a customised 476 gene NanoString panel. A random forest machine-learning model was applied on the NanoString profiles to develop a gene signature. An independent cohort of metastatic patients with GC treated with paclitaxel and ramucirumab (Pac-Ram) served as an external validation cohort. RESULTS: From the SAMIT trial 499 samples were analysed in this study. From the Pac-S-1 training cohort, the random forest model generated a 19-gene signature assigning patients to two groups: Pac-Sensitive and Pac-Resistant. In the Pac-UFT validation cohort, Pac-Sensitive patients exhibited a significant improvement in disease free survival (DFS): 3-year DFS 66% vs 40% (HR 0.44, p=0.0029). There was no survival difference between Pac-Sensitive and Pac-Resistant in the UFT or S-1 alone arms, test of interaction p<0.001. In the external Pac-Ram validation cohort, the signature predicted benefit for Pac-Sensitive (median PFS 147 days vs 112 days, HR 0.48, p=0.022). CONCLUSION: Using machine-learning techniques on one of the largest GC trials (SAMIT), we identify a gene signature representing the first predictive biomarker for paclitaxel benefit. TRIAL REGISTRATION NUMBER: UMIN Clinical Trials Registry: C000000082 (SAMIT); ClinicalTrials.gov identifier, 02628951 (South Korean trial).

Energy balance-related factors in childhood and adolescence and risk of colorectal cancer expressing different levels of proteins involved in the Warburg-effect.

Early-life (childhood to adolescence) energy balance-related factors (height, energy restriction, BMI) have been associated with adult colorectal cancer (CRC) risk. Warburg-effect activation via PI3K/Akt-signaling might explain this link. We investigated whether early-life energy balance-related factors were associated with risk of Warburg-subtypes in CRC. We used immunohistochemistry for six proteins involved in the Warburg-effect (LDHA, GLUT1, MCT4, PKM2, P53, and PTEN) on tissue microarrays of 2399 incident CRC cases from the prospective Netherlands Cohort Study (NLCS). Expression levels of all proteins were combined into a pathway-based sum score and categorized into three Warburg-subtypes (Warburg-low/-moderate/-high). Multivariable Cox-regression analyses were used to estimate associations of height, energy restriction proxies (exposure to Dutch Hunger Winter; Second World War [WWII]; Economic Depression) and adolescent BMI with Warburg-subtypes in CRC. Height was positively associated with colon cancer in men, regardless of Warburg-subtypes, and with Warburg-low colon and Warburg-moderate rectal cancer in women. Energy restriction during the Dutch Hunger Winter was inversely associated with colon cancer in men, regardless of Warburg-subtypes. In women, energy restriction during the Hunger Winter and WWII was inversely associated with Warburg-low colon cancer, whereas energy restriction during the Economic Depression was positively associated with Warburg-high colon cancer. Adolescent BMI was positively associated with Warburg-high colon cancer in men, and Warburg-moderate rectal cancer in women. In conclusion, the Warburg-effect seems to be involved in associations of adolescent BMI with colon cancer in men, and of energy restriction during the Economic Depression with colon cancer in women. Further research is needed to validate these results.

Energy balance-related factors in childhood and adolescence and risk of colorectal cancer based on KRAS, PIK3CA, and BRAF mutations and MMR status.

KRAS mutations (KRASmut ), PIK3CAmut , BRAFmut , and deficient DNA mismatch repair (dMMR) have been associated with the Warburg effect. We previously reported differential associations between early-life energy balance-related factors (height, energy restriction, body mass index [BMI]) and colorectal cancer (CRC) subtypes based on the Warburg effect. We now investigated associations of early-life energy balance-related factors and the risk of CRC subgroups based on mutation and MMR status. Data from the Netherlands Cohort Study was used. KRASmut , PIK3CAmut, BRAFmut, and MMR status were available for 2349 CRC cases, and complete covariate data for 1934 cases and 3911 subcohort members. Multivariable-adjusted Cox regression was used to estimate associations of height, energy restriction proxies (exposure to Dutch Hunger Winter, Second World War, Economic Depression), and early adult BMI (age 20 years) with risk of CRC based on individual molecular features and combinations thereof (all-wild-type+MMR-proficient [pMMR]; any-mutation/dMMR). Height was positively associated with any-mutation/dMMR CRC but not all-wild-type+pMMR CRC, with the exception of rectal cancer in men, and with heterogeneity in associations observed for colon cancer in men (p-heterogeneity = 0.049) and rectal cancer in women (p-heterogeneity = 0.014). Results on early-life energy restriction proxies in relation to the risk of CRC subgroups did not show clear patterns. Early adult BMI was positively, but not significantly, associated with KRASmut colon cancer in men and with BRAFmut and dMMR colon cancer in women. Our results suggest a role of KRASmut , PIK3CAmut , BRAFmut , and dMMR in the etiological pathway between height and CRC risk. KRASmut might potentially play a role in associations of early adult BMI with colon cancer risk in men, and BRAFmut and dMMR in women.

Increasing frequency of gene copy number aberrations is associated with immunosuppression and predicts poor prognosis in gastric adenocarcinoma.

BACKGROUND: Patients with Epstein-Barr virus-positive gastric cancers or those with microsatellite instability appear to have a favourable prognosis. However, the prognostic value of the chromosomal status (chromosome-stable (CS) versus chromosomal instable (CIN)) remains unclear in gastric cancer. METHODS: Gene copy number aberrations (CNAs) were determined in 16 CIN-associated genes in a retrospective study including test and validation cohorts of patients with gastric cancer. Patients were stratified into CS (no CNA), CINlow (1-2 CNAs) or CINhigh (3 or more CNAs). The relationship between chromosomal status, clinicopathological variables, and overall survival (OS) was analysed. The relationship between chromosomal status, p53 expression, and tumour infiltrating immune cells was also assessed and validated externally. RESULTS: The test and validation cohorts included 206 and 748 patients, respectively. CINlow and CINhigh were seen in 35.0 and 15.0 per cent of patients, respectively, in the test cohort, and 48.5 and 20.7 per cent in the validation cohort. Patients with CINhigh gastric cancer had the poorest OS in the test and validation cohorts. In multivariable analysis, CINlow, CINhigh and pTNM stage III-IV (P < 0.001) were independently associated with poor OS. CIN was associated with high p53 expression and low immune cell infiltration. CONCLUSION: CIN may be a potential new prognostic biomarker independent of pTNM stage in gastric cancer. Patients with gastric cancer demonstrating CIN appear to be immunosuppressed, which might represent one of the underlying mechanisms explaining the poor survival and may help guide future therapeutic decisions.

Deep learning detects genetic alterations in cancer histology generated by adversarial networks.

Deep learning can detect microsatellite instability (MSI) from routine histology images in colorectal cancer (CRC). However, ethical and legal barriers impede sharing of images and genetic data, hampering development of new algorithms for detection of MSI and other biomarkers. We hypothesized that histology images synthesized by conditional generative adversarial networks (CGANs) retain information about genetic alterations. To test this, we developed a 'histology CGAN' which was trained on 256 patients (training cohort 1) and 1457 patients (training cohort 2). The CGAN synthesized 10 000 synthetic MSI and non-MSI images which contained a range of tissue types and were deemed realistic by trained observers in a blinded study. Subsequently, we trained a deep learning detector of MSI on real or synthetic images and evaluated the performance of MSI detection in a held-out set of 142 patients. When trained on real images from training cohort 1, this system achieved an area under the receiver operating curve (AUROC) of 0.742 [0.681, 0.854]. Training on the larger cohort 2 only marginally improved the AUROC to 0.757 [0.707, 0.869]. Training on purely synthetic data resulted in an AUROC of 0.743 [0.658, 0.801]. Training on both real and synthetic data further increased AUROC to 0.777 [0.715, 0.821]. We conclude that synthetic histology images retain information reflecting underlying genetic alterations in colorectal cancer. Using synthetic instead of real images to train deep learning systems yields non-inferior classifiers. This approach can be used to create large shareable data sets or to augment small data sets with rare molecular features. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

Spatial profiling of gastric cancer patient-matched primary and locoregional metastases reveals principles of tumour dissemination.

OBJECTIVE: Endoscopic mucosal biopsies of primary gastric cancers (GCs) are used to guide diagnosis, biomarker testing and treatment. Spatial intratumoural heterogeneity (ITH) may influence biopsy-derived information. We aimed to study ITH of primary GCs and matched lymph node metastasis (LNmet). DESIGN: GC resection samples were annotated to identify primary tumour superficial (PTsup), primary tumour deep (PTdeep) and LNmet subregions. For each subregion, we determined (1) transcriptomic profiles (NanoString 'PanCancer Progression Panel', 770 genes); (2) next-generation sequencing (NGS, 225 gastrointestinal cancer-related genes); (3) DNA copy number profiles by multiplex ligation-dependent probe amplification (MLPA, 16 genes); and (4) histomorphological phenotypes. RESULTS: NanoString profiling of 64 GCs revealed no differences between PTsup1 and PTsup2, while 43% of genes were differentially expressed between PTsup versus PTdeep and 38% in PTsup versus LNmet. Only 16% of genes were differently expressed between PTdeep and LNmet. Several genes with therapeutic potential (eg IGF1, PIK3CD and TGFB1) were overexpressed in LNmet and PTdeep compared with PTsup. NGS data revealed orthogonal support of NanoString results with 40% mutations present in PTdeep and/or LNmet, but not in PTsup. Conversely, only 6% of mutations were present in PTsup and were absent in PTdeep and LNmet. MLPA demonstrated significant ITH between subregions and progressive genomic changes from PTsup to PTdeep/LNmet. CONCLUSION: In GC, regional lymph node metastases are likely to originate from deeper subregions of the primary tumour. Future clinical trials of novel targeted therapies must consider assessment of deeper subregions of the primary tumour and/or metastases as several therapeutically relevant genes are only mutated, overexpressed or amplified in these regions.

Real-time lipid patterns to classify viable and necrotic liver tumors.

Real-time tissue classifiers based on molecular patterns are emerging tools for fast tumor diagnosis. Here, we used rapid evaporative ionization mass spectrometry (REIMS) and multivariate statistical analysis (principal component analysis-linear discriminant analysis) to classify tissues with subsequent comparison to gold standard histopathology. We explored whether REIMS lipid patterns can identify human liver tumors and improve the rapid characterization of their underlying metabolic features. REIMS-based classification of liver parenchyma (LP), hepatocellular carcinoma (HCC), and metastatic adenocarcinoma (MAC) reached an accuracy of 98.3%. Lipid patterns of LP were more similar to those of HCC than to those of MAC and allowed clear distinction between primary and metastatic liver tumors. HCC lipid patterns were more heterogeneous than those of MAC, which is consistent with the variation seen in the histopathological phenotype. A common ceramide pattern discriminated necrotic from viable tumor in MAC with 92.9% accuracy and in other human tumors. Targeted analysis of ceramide and related sphingolipid mass features in necrotic tissues may provide a new classification of tumor cell death based on metabolic shifts. Real-time lipid patterns may have a role in future clinical decision-making in cancer precision medicine.

Gastric cancer.

Gastric cancer is the fifth most common cancer and the third most common cause of cancer death globally. Risk factors for the condition include Helicobacter pylori infection, age, high salt intake, and diets low in fruit and vegetables. Gastric cancer is diagnosed histologically after endoscopic biopsy and staged using CT, endoscopic ultrasound, PET, and laparoscopy. It is a molecularly and phenotypically highly heterogeneous disease. The main treatment for early gastric cancer is endoscopic resection. Non-early operable gastric cancer is treated with surgery, which should include D2 lymphadenectomy (including lymph node stations in the perigastric mesentery and along the celiac arterial branches). Perioperative or adjuvant chemotherapy improves survival in patients with stage 1B or higher cancers. Advanced gastric cancer is treated with sequential lines of chemotherapy, starting with a platinum and fluoropyrimidine doublet in the first line; median survival is less than 1 year. Targeted therapies licensed to treat gastric cancer include trastuzumab (HER2-positive patients first line), ramucirumab (anti-angiogenic second line), and nivolumab or pembrolizumab (anti-PD-1 third line).

Lymph node response to chemoradiotherapy in oesophageal cancer patients: relationship with radiotherapy fields.

BACKGROUND: The presence of lymph node metastasis (LNmets) is a poor prognostic factor in oesophageal cancer (OeC) patients treated with neoadjuvant chemoradiotherapy (nCRT) followed by surgery. Tumour regression grade (TRG) in LNmets has been suggested as a predictor for survival. The aim of this study was to investigate whether TRG in LNmets is related to their location within the radiotherapy (RT) field. METHODS: Histopathological TRG was retrospectively classified in 2565 lymph nodes (LNs) from 117 OeC patients treated with nCRT and surgery as: (A) no tumour, no signs of regression; (B) tumour without regression; (C) viable tumour and regression; and (D) complete response. Multivariate survival analysis was used to investigate the relationship between LN location within the RT field, pathological TRG of the LN and TRG of the primary tumour. RESULTS: In 63 (54%) patients, viable tumour cells or signs of regression were seen in 264 (10.2%) LNs which were classified as TRG-B (n = 56), C (n = 104) or D (n = 104) LNs. 73% of B, C and D LNs were located within the RT field. There was a trend towards a relationship between LN response and anatomical LN location with respect to the RT field (p = 0.052). Multivariate analysis showed that only the presence of LNmets within the RT field with TRG-B is related to poor overall survival. CONCLUSION: Patients have the best survival if all LNmets show tumour regression, even if LNmets are located outside the RT field. Response in LNmets to nCRT is heterogeneous which warrants further studies to better understand underlying mechanisms.

A national study to assess outcomes of definitive chemoradiation regimens in proximal esophageal cancer.

Background: Proximal esophageal cancer (EC) is commonly treated with definitive chemoradiation (CRT). The radiation dose and type of chemotherapy backbone are still under debate. The objective of this study was to compare the treatment outcomes of contemporary CRT regimens.Material and Methods: In this retrospective observational cohort study, we included patients with locally advanced squamous cell cancer of the proximal esophagus, from 11 centers in the Netherlands, treated with definitive CRT between 2004 and 2014. Each center had a preferential CRT regimen, based on cisplatin (Cis) or carboplatin-paclitaxel (CP) combined with low (≤50.4 Gy) or high (>50.4 Gy) dose radiotherapy (RT). Differences in overall survival (OS) between CRT regimens were assessed using a fully adjusted Cox proportional hazards and propensity score (PS) weighted model. Safety profiles were compared using a multilevel logistic regression model.Results: Two hundred patients were included. Fifty-four, 39, 95, and 12 patients were treated with Cis-low-dose RT, Cis-high-dose RT, CP-low-dose RT, and CP-high-dose RT, respectively. Median follow-up was 62.6 months (95% CI: 47.9-77.2 months). Median OS (21.9 months; 95% CI: 16.9-27.0 months) was comparable between treatment groups (logrank p = .88), confirmed in the fully adjusted and PS weighted model (p > .05). Grades 3-5 acute adverse events were less frequent in patients treated with CP-low-dose RT versus Cis-high-dose RT (OR 3.78; 95% CI: 1.31-10.87; p = .01). The occurrence of grades 3-5 late toxicities was not different between treatment groups.Conclusion: Our study was unable to demonstrate a difference in OS between the CRT regimens, probably related to the relatively small sample size. Based on the superior safety profile, carboplatin and paclitaxel-based CRT regimens are preferred in patients with locally advanced proximal EC.