Prognostic and predictive impact of sex in locally advanced microsatellite instability high gastric or gastroesophageal junction cancer: An individual patient data pooled analysis of randomized clinical trials.

BACKGROUND: Surgery plus peri-operative/adjuvant chemotherapy is the standard of care for locally advanced GC/GEJC, though with unsatisfactory results. dMMR/MSI-high tumors have better prognosis and scant benefit from chemotherapy as compared to pMMR/MSS ones. The differential outcome of therapies in terms of safety and efficacy according to sex is still debated in GC/GEJC patients. METHODS: We previously performed an individual patient data pooled analysis of MAGIC, CLASSIC, ITACA-S, and ARTIST trials including GC/GEJC patients treated with surgery alone or surgery plus peri-operative/adjuvant chemotherapy to assess the value of MSI status. We performed a secondary analysis investigating the prognostic and predictive role of sex (female versus male) in the pooled analysis dataset in the overall population and patients stratified for MSI status (MSI-high versus MSS/MSI-low). Disease-free (DFS) and overall survival (OS) were calculated. RESULTS: Patients with MSI-high tumors had improved survival as compared to MSS/MSI-low ones irrespective of sex, whereas in those with MSS/MSI-low tumors, females had numerically longer OS and DFS (5-year OS was 63.2% versus 57.6%, HR 0.842; p = 0.058, and 5-year DFS was 55.8% versus 50.8%, HR 0.850; p = 0.0504 in female versus male patients). The numerical difference for the detrimental effect of chemotherapy in MSI-high GC was higher in females than males, while the significant benefit of chemotherapy over surgery alone was confirmed in MSS/MSI-low GC irrespective of sex. CONCLUSIONS: This pooled analysis including four randomized trials highlights a relevant impact of sex in the prognosis and treatment efficacy of MSI-high and MSS/MSI-low non-metastatic GC/GEJC.

International consensus on the management of metastatic gastric cancer: step by step in the foggy landscape : Bertinoro Workshop, November 2022.

BACKGROUND: Many gastric cancer patients in Western countries are diagnosed as metastatic with a median overall survival of less than twelve months using standard chemotherapy. Innovative treatments, like targeted therapy or immunotherapy, have recently proved to ameliorate prognosis, but a general agreement on managing oligometastatic disease has yet to be achieved. An international multi-disciplinary workshop was held in Bertinoro, Italy, in November 2022 to verify whether achieving a consensus on at least some topics was possible. METHODS: A two-round Delphi process was carried out, where participants were asked to answer 32 multiple-choice questions about CT, laparoscopic staging and biomarkers, systemic treatment for different localization, role and indication of palliative care. Consensus was established with at least a 67% agreement. RESULTS: The assembly agreed to define oligometastases as a "dynamic" disease which either regresses or remains stable in response to systemic treatment. In addition, the definition of oligometastases was restricted to the following sites: para-aortic nodal stations, liver, lung, and peritoneum, excluding bones. In detail, the following conditions should be considered as oligometastases: involvement of para-aortic stations, in particular 16a2 or 16b1; up to three technically resectable liver metastases; three unilateral or two bilateral lung metastases; peritoneal carcinomatosis with PCI ≤ 6. No consensus was achieved on how to classify positive cytology, which was considered as oligometastatic by 55% of participants only if converted to negative after chemotherapy. CONCLUSION: As assessed at the time of diagnosis, surgical treatment of oligometastases should aim at R0 curativity on the entire disease volume, including both the primary tumor and its metastases. Conversion surgery was defined as surgery on the residual volume of disease, which was initially not resectable for technical and/or oncological reasons but nevertheless responded to first-line treatment.

Computational methods for metastasis detection in lymph nodes and characterization of the metastasis-free lymph node microarchitecture: A systematic-narrative hybrid review.

Background: Histological examination of tumor draining lymph nodes (LNs) plays a vital role in cancer staging and prognostication. However, as soon as a LN is classed as metastasis-free, no further investigation will be performed and thus, potentially clinically relevant information detectable in tumor-free LNs is currently not captured. Objective: To systematically study and critically assess methods for the analysis of digitized histological LN images described in published research. Methods: A systematic search was conducted in several public databases up to December 2023 using relevant search terms. Studies using brightfield light microscopy images of hematoxylin and eosin or immunohistochemically stained LN tissue sections aiming to detect and/or segment LNs, their compartments or metastatic tumor using artificial intelligence (AI) were included. Dataset, AI methodology, cancer type, and study objective were compared between articles. Results: A total of 7201 articles were collected and 73 articles remained for detailed analyses after article screening. Of the remaining articles, 86% aimed at LN metastasis identification, 8% aimed at LN compartment segmentation, and remaining focused on LN contouring. Furthermore, 78% of articles used patch classification and 22% used pixel segmentation models for analyses. Five out of six studies (83%) of metastasis-free LNs were performed on publicly unavailable datasets, making quantitative article comparison impossible. Conclusions: Multi-scale models mimicking multiple microscopy zooms show promise for computational LN analysis. Large-scale datasets are needed to establish the clinical relevance of analyzing metastasis-free LN in detail. Further research is needed to identify clinically interpretable metrics for LN compartment characterization.

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.

Germinal centres within tumour positive sentinel lymph nodes are positively associated with tumour infiltrating lymphocytes and tertiary lymphoid structures in breast cancer.

BACKGROUND: Stromal tumour infiltrating lymphocytes (sTILs) and presence of tertiary lymphoid structures have been proposed as indicators of tumour-related immune response in breast cancer. An increased number of germinal centres (GCs) in lymph nodes is considered a sign of humoral immune reactivity. AIMS: It is unclear whether a relationship exists between number and size of GCs within tumour positive sentinel lymph nodes (SLNpos), sTILs and tertiary lymphoid structures within matched primary breast cancer and breast cancer subtype. METHODS: Axillary SLNpos from 175 patients with breast cancer were manually contoured in digitized haematoxylin and eosin stained sections. Total SLN area, GC number and GC area were measured in SLNpos with the largest metastatic area. To correct for SLN size, GC number and GC area were divided by SLN area. sTILs and presence of tertiary lymphoid structures were assessed in the primary breast cancer. RESULTS: A higher GC number and larger GC area were found in patients with high sTILs (≥2%) (both P < 0.001) and in patients with presence of tertiary lymphoid structures (PGC number = 0.034 and PGC area = 0.016). Triple negative and HER2-positive (N = 45) breast cancer subtypes had a higher GC number and higher sTILs compared to hormone receptor positive, HER2-negative breast cancer (N = 130) (PGC number < 0.001 and PsTILs= 0.001). CONCLUSION: This study suggests GCs measured within SLNpos might be useful indicators of the humoral anti-tumour immune response in breast cancer. Future studies are needed investigating underlying biological mechanisms and prognostic value of GCs in SLNs.

End-to-end prognostication in colorectal cancer by deep learning: a retrospective, multicentre study.

BACKGROUND: Precise prognosis prediction in patients with colorectal cancer (ie, forecasting survival) is pivotal for individualised treatment and care. Histopathological tissue slides of colorectal cancer specimens contain rich prognostically relevant information. However, existing studies do not have multicentre external validation with real-world sample processing protocols, and algorithms are not yet widely used in clinical routine. METHODS: In this retrospective, multicentre study, we collected tissue samples from four groups of patients with resected colorectal cancer from Australia, Germany, and the USA. We developed and externally validated a deep learning-based prognostic-stratification system for automatic prediction of overall and cancer-specific survival in patients with resected colorectal cancer. We used the model-predicted risk scores to stratify patients into different risk groups and compared survival outcomes between these groups. Additionally, we evaluated the prognostic value of these risk groups after adjusting for established prognostic variables. FINDINGS: We trained and validated our model on a total of 4428 patients. We found that patients could be divided into high-risk and low-risk groups on the basis of the deep learning-based risk score. On the internal test set, the group with a high-risk score had a worse prognosis than the group with a low-risk score, as reflected by a hazard ratio (HR) of 4·50 (95% CI 3·33-6·09) for overall survival and 8·35 (5·06-13·78) for disease-specific survival (DSS). We found consistent performance across three large external test sets. In a test set of 1395 patients, the high-risk group had a lower DSS than the low-risk group, with an HR of 3·08 (2·44-3·89). In two additional test sets, the HRs for DSS were 2·23 (1·23-4·04) and 3·07 (1·78-5·3). We showed that the prognostic value of the deep learning-based risk score is independent of established clinical risk factors. INTERPRETATION: Our findings indicate that attention-based self-supervised deep learning can robustly offer a prognosis on clinical outcomes in patients with colorectal cancer, generalising across different populations and serving as a potentially new prognostic tool in clinical decision making for colorectal cancer management. We release all source codes and trained models under an open-source licence, allowing other researchers to reuse and build upon our work. FUNDING: The German Federal Ministry of Health, the Max-Eder-Programme of German Cancer Aid, the German Federal Ministry of Education and Research, the German Academic Exchange Service, and the EU.

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.

The Role of Nonmetastatic Lymph Nodes in the Survival of Colorectal Cancer: A Systematic Review.

Objective: In this review, we aim to provide an overview of literature on lymph node (LN) histomorphological features and their relationship with the prognosis in colorectal cancer (CRC). Background: Lymph nodes play a crucial role in the treatment and prognosis of CRC. The presence of LN metastases considerably worsens the prognosis in CRC patients. Literature has shown that the total number of LNs and the number negative LNs (LNnegs) has prognostic value in CRC patients. In esophageal carcinoma, LN size seems to be surrogate of the host antitumor response and a potentially clinically useful new prognostic biomarker for (y)pN0 esophageal carcinoma. Methods: A comprehensive search was performed in Pubmed, Embase, Medline, CINAHL, and the Cochrane library in March 2021. The PRISMA guidelines were followed. Only studies focusing on histomorphological features and LN size and their relation to overall survival were selected. Results: A total of 9 unique articles met all inclusion criteria and were therefore included in this systematic review. Six of these studies investigated HMF (eg, paracortical hyperplasia, germinal center predominance, and sinus histiocytosis) and 4 studies LNneg size and their relationship with overall survival. The presence of paracortical hyperplasia and an increased number of large LNnegs is related to a more favorable prognosis in CRC. Conclusion: The results of this systematic review seem to support the hypothesis that there is a relationship between the host antitumor response reflected in different histomorphological reaction patterns visible in LNnegs and LNneg size related to survival in CRC patients.

Clinical Features of Gastric Signet Ring Cell Cancer: Results from a Systematic Review and Meta-Analysis.

BACKGROUND: Conflicting results about the prognostic relevance of signet ring cell histology in gastric cancer have been reported. We aimed to perform a meta-analysis focusing on the clinicopathological features and prognosis of this subgroup of cancer compared with other histologies. METHODS: A systematic literature search in the PubMed database was conducted, including all publications up to 1 October 2021. A meta-analysis comparing the results of the studies was performed. RESULTS: A total of 2062 studies referring to gastric cancer with signet ring cell histology were identified, of which 262 studies reported on its relationship with clinical information. Of these, 74 were suitable to be included in the meta-analysis. A slightly lower risk of developing nodal metastases in signet ring cell tumours compared to other histotypes was found (especially to undifferentiated/poorly differentiated/mucinous and mixed histotypes); the lower risk was more evident in early and slightly increased in advanced gastric cancer. Survival tended to be better in early stage signet ring cell cancer compared to other histotypes; no differences were shown in advanced stages, and survival was poorer in metastatic patients. In the subgroup analysis, survival in signet ring cell cancer was slightly worse compared to non-signet ring cell cancer and differentiated/well-to-moderately differentiated adenocarcinoma. CONCLUSIONS: Most of the conflicting results in signet ring cell gastric cancer literature could be derived from the lack of standardisation in their classification and the comparison with the different subtypes of gastric cancer. There is a critical need to strive for a standardised classification system for gastric cancer, fostering clarity and coherence in the forthcoming research and clinical applications.

The prognostic value of the histological shape of tumor negative sentinel nodes in breast cancer.

Introduction: Sentinel lymph node (SLN) metastasis is an important predictor of prognosis in breast cancer (BC) patients, guiding treatment decisions. However, patients with the same BC subtype and tumor negative SLN (SLNneg) can have different survival outcomes. We hypothesized that the host anti-tumor immune reaction in SLNneg is important and results in morphometrically measurable changes in SLN size or shape which are related to patient prognosis. Methods: Surface area, circumference, long axis and short axis were histologically measured in 694 SLNneg from 356 cases of invasive BC and 67 ductal carcinoma in situ cases. The area occupied by fat was categorized as less or more than 50%. The long to short axis (L/S) ratio was calculated. The relationship between SLNneg morphometries and clinicopathological variables like tumor-infiltrating lymphocytes (TILs) within the primary tumor, as well as prognosis at 10 years follow up were analyzed. Results: The mean SLNneg surface area was 78.7mm2, circumference 40.3mm, long axis 13.1mm, short axis 8.2mm and L/S ratio 1.7. Larger surface area, long axis and short axis, including age >55 years were associated with higher body mass index (BMI) and SLN fat over 50% (p<0.003). In invasive BC, a high SLNneg L/S ratio (≥1.9) was related to poorer disease-free (HR=1.805, 95%CI 1.182-2.755, p=0.006) and overall (HR=2.389, 95%CI 1.481-3.851, p<0.001) survival. A low SLNneg L/S ratio (<1.9) was associated with high TILs in the primary BC (≥10%) (p=0.005). However a high TIL count was not of prognostic relevance. Conclusions: This is the first study to suggest that morphometric characteristics of axillary SLNneg, like L/S ratio, could be used to predict prognosis in patients with SLNneg invasive BC of all subtypes. The association between low L/S ratio and high TILs suggest that SLN shape is related to immunological functioning of the SLN and could be used in addition to TIL evaluation. Regarding the dubious role of TILs in hormone receptor positive breast cancer, SLNneg morphometry to gain information about host immune status could especially be of benefit in this subtype. Further studies are warranted to better understand the underlying biological mechanisms.

Deep learning trained on lymph node status predicts outcome from gastric cancer histopathology: a retrospective multicentric study.

AIM: Gastric cancer (GC) is a tumour entity with highly variant outcomes. Lymph node metastasis is a prognostically adverse biomarker. We hypothesised that GC primary tissue contains information that is predictive of lymph node status and patient prognosis and that this information can be extracted using deep learning (DL). METHODS: Using three patient cohorts comprising 1146 patients, we trained and validated a DL system to predict lymph node status directly from haematoxylin and eosin-stained GC tissue sections. We investigated the concordance between the DL-based prediction from the primary tumour slides (aiN score) and the histopathological lymph node status (pN). Furthermore, we assessed the prognostic value of the aiN score alone and when combined with the pN status. RESULTS: The aiN score predicted the pN status reaching area under the receiver operating characteristic curves of 0.71 in the training cohort and 0.69 and 0.65 in the two test cohorts. In a multivariate Cox analysis, the aiN score was an independent predictor of patient survival with hazard ratios of 1.5 in the training cohort and of 1.3 and 2.2 in the two test cohorts. A combination of the aiN score and the pN status prognostically stratified patients by survival with p-values <0.05 in logrank tests. CONCLUSION: GC primary tumour tissue contains additional prognostic information that is accessible using the aiN score. In combination with the pN status, this can be used for personalised management of GC patients after prospective validation.

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.

Transformer-based biomarker prediction from colorectal cancer histology: A large-scale multicentric study.

Deep learning (DL) can accelerate the prediction of prognostic biomarkers from routine pathology slides in colorectal cancer (CRC). However, current approaches rely on convolutional neural networks (CNNs) and have mostly been validated on small patient cohorts. Here, we develop a new transformer-based pipeline for end-to-end biomarker prediction from pathology slides by combining a pre-trained transformer encoder with a transformer network for patch aggregation. Our transformer-based approach substantially improves the performance, generalizability, data efficiency, and interpretability as compared with current state-of-the-art algorithms. After training and evaluating on a large multicenter cohort of over 13,000 patients from 16 colorectal cancer cohorts, we achieve a sensitivity of 0.99 with a negative predictive value of over 0.99 for prediction of microsatellite instability (MSI) on surgical resection specimens. We demonstrate that resection specimen-only training reaches clinical-grade performance on endoscopic biopsy tissue, solving a long-standing diagnostic problem.

Programmed death-ligand 1 (PD-L1) expression in primary gastric adenocarcinoma and matched metastases.

BACKGROUND: Combination of immunotherapy and chemotherapy is recommended for first line treatment of gastric adenocarcinoma (GC) patients with locally advanced unresectable disease or metastatic disease. However, data regarding the concordance rate between PD-L1 combined positive score (CPS) in primary GC and matched regional lymph node metastasis (LNmet) or matched distant metastasis (Dmet) is limited. METHODS: Tissue microarray sections from primary resected GC, LNmet and Dmet were immunohistochemically stained with anti-PD-L1 (clone SP263). PD-L1 expression was scored separately in tumour cells and immune cells and compared between matched primary GC, LNmet and/or Dmet. CPS was calculated and results for CPS cut-offs 1 and 5 were compared between matched samples. RESULTS: 275 PD-L1 stained GC were analysed. 189 primary GC had matched LNmet. CPS cut-off 1 concordance rate between primary GC and LNmet was 77%. 23 primary GC had matched Dmet but no matched LNmet, CPS cut-off 1 concordance rate was 70%. 63 primary GC had both matched LNmet and matched Dmet, CPS cut-off 1 concordance rate of 67%. CPS cut-off 5 results were similar. The proportion of PD-L1 positive tumour cells increased from primary GC (26%) to LNmet (42%) and was highest in Dmet (75%). CONCLUSION: Our study showed up to 33% discordance of PD-L1 CPS between primary GC and LNmet and/or Dmet suggesting that multiple biopsies of primary GC and metastatic sites might need to be tested before considering treatment options. Moreover, this is the first study that seems to suggest that tumour cells acquire PD-L1 expression during disease progression.

Pathologic Lymph Node Regression After Neoadjuvant Chemotherapy Predicts Recurrence and Survival in Esophageal Adenocarcinoma: A Multicenter Study in the United Kingdom.

PURPOSE: There is limited evidence regarding the prognostic effects of pathologic lymph node (LN) regression after neoadjuvant chemotherapy for esophageal adenocarcinoma, and a definition of LN response is lacking. This study aimed to evaluate how LN regression influences survival after surgery for esophageal adenocarcinoma. METHODS: Multicenter cohort study of patients with esophageal adenocarcinoma treated with neoadjuvant chemotherapy followed by surgical resection at five high-volume centers in the United Kingdom. LNs retrieved at esophagectomy were examined for chemotherapy response and given a LN regression score (LNRS)-LNRS 1, complete response; 2, <10% residual tumor; 3, 10%-50% residual tumor; 4, >50% residual tumor; and 5, no response. Survival analysis was performed using Cox regression adjusting for confounders including primary tumor regression. The discriminatory ability of different LN response classifications to predict survival was evaluated using Akaike information criterion and Harrell C-index. RESULTS: In total, 17,930 LNs from 763 patients were examined. LN response classified as complete LN response (LNRS 1 ≥1 LN, no residual tumor in any LN; n = 62, 8.1%), partial LN response (LNRS 1-3 ≥1 LN, residual tumor ≥1 LN; n = 155, 20.3%), poor/no LN response (LNRS 4-5; n = 303, 39.7%), or LN negative (no tumor/regression; n = 243, 31.8%) demonstrated superior discriminatory ability. Mortality was reduced in patients with complete LN response (hazard ratio [HR], 0.35; 95% CI, 0.22 to 0.56), partial LN response (HR, 0.72; 95% CI, 0.57 to 0.93) or negative LNs (HR, 0.32; 95% CI, 0.25 to 0.42) compared with those with poor/no LN response. Primary tumor regression and LN regression were discordant in 165 patients (21.9%). CONCLUSION: Pathologic LN regression after neoadjuvant chemotherapy was a strong prognostic factor and provides important information beyond pathologic TNM staging and primary tumor regression grading. LN regression should be included as standard in the pathologic reporting of esophagectomy specimens.

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.

Dissecting the genetic heterogeneity of gastric cancer.

BACKGROUND: Gastric cancer (GC) is clinically heterogenous according to location (cardia/non-cardia) and histopathology (diffuse/intestinal). We aimed to characterize the genetic risk architecture of GC according to its subtypes. Another aim was to examine whether cardia GC and oesophageal adenocarcinoma (OAC) and its precursor lesion Barrett's oesophagus (BO), which are all located at the gastro-oesophageal junction (GOJ), share polygenic risk architecture. METHODS: We did a meta-analysis of ten European genome-wide association studies (GWAS) of GC and its subtypes. All patients had a histopathologically confirmed diagnosis of gastric adenocarcinoma. For the identification of risk genes among GWAS loci we did a transcriptome-wide association study (TWAS) and expression quantitative trait locus (eQTL) study from gastric corpus and antrum mucosa. To test whether cardia GC and OAC/BO share genetic aetiology we also used a European GWAS sample with OAC/BO. FINDINGS: Our GWAS consisting of 5816 patients and 10,999 controls highlights the genetic heterogeneity of GC according to its subtypes. We newly identified two and replicated five GC risk loci, all of them with subtype-specific association. The gastric transcriptome data consisting of 361 corpus and 342 antrum mucosa samples revealed that an upregulated expression of MUC1, ANKRD50, PTGER4, and PSCA are plausible GC-pathomechanisms at four GWAS loci. At another risk locus, we found that the blood-group 0 exerts protective effects for non-cardia and diffuse GC, while blood-group A increases risk for both GC subtypes. Furthermore, our GWAS on cardia GC and OAC/BO (10,279 patients, 16,527 controls) showed that both cancer entities share genetic aetiology at the polygenic level and identified two new risk loci on the single-marker level. INTERPRETATION: Our findings show that the pathophysiology of GC is genetically heterogenous according to location and histopathology. Moreover, our findings point to common molecular mechanisms underlying cardia GC and OAC/BO. FUNDING: German Research Foundation (DFG).